Dual N-Back FAQ

A compendium of DNB, WM, IQ information up to 2015
DNB, psychology, experiments, R, survey, Bayes, IQ
2009-03-252019-12-05 in progress certainty: unlikely importance: 7

Dual N-Back is a kind of in­tended to ex­pand your (WM), and hope­fully your in­tel­li­gence (IQ1).

The the­ory orig­i­nally went that novel2 cog­ni­tive processes tend to over­lap and seem to go through . As it hap­pens, WM pre­dicts and cor­re­lates with IQ3 and may use the same neural net­works4, sug­gest­ing that WM might be IQ5. WM is known to be train­able, and so im­prov­ing WM would hope­fully im­prove IQ. And N-back is a fam­ily of tasks which stress at­ten­tion and WM.

Later re­search found that per­for­mance and im­prove­ment on N-back seems to cor­re­late bet­ter with IQ rather than clas­sic mea­sures of WM like recit­ing lists of num­bers, rais­ing the ques­tion of whether N-back works via in­creas­ing WM or by im­prov­ing self­-con­trol or im­prov­ing ma­nip­u­la­tion of WM con­tents (rather than WM’s size) or some­how train­ing IQ di­rect­ly.6 Per­for­mance on DNB has com­pli­cated cor­re­la­tions with per­for­mance on other tests of work­ing mem­ory or IQ, so it’s not clear what it is tap­ping in­to. (And the link be­tween WM and per­for­mance on IQ tests has been dis­put­ed; high WM as mea­sured by OSPAN does not cor­re­late well with per­for­mance on hard Raven’s ques­tions7 and the va­lid­ity of sin­gle tests of WM train­ing has been ques­tioned8.)

Brain Work­shop offers many mod­es, some far more elab­o­rate than sim­ple Dual N-back; no re­search has been done on them, so lit­tle can be said about what they are good for or what they train or what im­prove­ments they may offer; Jaeggi 2010 seemed to find Sin­gle N-back bet­ter than Dual N-back. Some of the more elab­o­rate modes seem to fo­cus heav­ily on shift­ing the cor­rect re­sponse among var­i­ous modal­i­ties - not just sound, but left­/right, eg. - and so stress ; there are re­sults that task switch­ing can be trained and that it trans­fers9, but how use­ful this is and how well the BW modes train this are un­known.

The Argument

for learn­ing and also just gen­eral in­tel­li­gence.10 It’s not too hard to see why work­ing mem­ory could be so im­por­tant. Work­ing mem­ory boils down to ‘how much stuff you can think about at the same time’.

Imag­ine a poor pro­gram­mer who has suffered brain dam­age and has only enough work­ing mem­ory for 1 de­fi­n­i­tion at a time. How could he write any­thing? To write a cor­rect pro­gram, he needs to know si­mul­ta­ne­ously 2 things - what a vari­able, say, con­tains, and what is valid in­put for a pro­gram. But un­for­tu­nate­ly, our pro­gram­mer can know that the vari­able foo con­tains a string with the in­put, or he can know that the func­tion processInput uses a string, but he can’t re­mem­ber these 2 things si­mul­ta­ne­ous­ly! He will dead­lock forever, un­sure ei­ther what to do with this foo, or un­sure what ex­actly processInput was sup­posed to work on.

More se­ri­ous­ly, work­ing mem­ory can be use­ful since it al­lows one to grasp more of the struc­ture of some­thing at any one time. Com­men­ta­tors on pro­gram­ming often write that one of the great chal­lenges of pro­gram­ming (be­sides the chal­lenge of ac­cept­ing & deal­ing with the re­al­ity that a com­puter re­ally is just a mind­less rule-fol­low­ing ma­chine), is that pro­gram­ming re­quires one to keep in mind dozens of things and cir­cum­stances - any one of which could com­pletely bol­lix things up. Fo­cus is ab­solutely es­sen­tial. One of the char­ac­ter­is­tics of great pro­gram­mers is their ap­par­ent om­ni­science. Ob­ses­sion grants them this abil­ity to know what they are ac­tu­ally do­ing:

“With pro­gram­mers, it’s es­pe­cially hard. Pro­duc­tiv­ity de­pends on be­ing able to jug­gle a lot of lit­tle de­tails in short term mem­ory all at once. Any kind of in­ter­rup­tion can cause these de­tails to come crash­ing down. When you re­sume work, you can’t re­mem­ber any of the de­tails (like lo­cal vari­able names you were us­ing, or where you were up to in im­ple­ment­ing that search al­go­rithm) and you have to keep look­ing these things up, which slows you down a lot un­til you get back up to speed.” –, “Where do These Peo­ple Get Their (Uno­rig­i­nal) Ideas?”

“Sev­eral friends men­tioned hack­ers’ abil­ity to con­cen­trate - their abil­i­ty, as one put it, to ‘tune out every­thing out­side their own heads.’ I’ve cer­tainly no­ticed this. And I’ve heard sev­eral hack­ers say that after drink­ing even half a beer they can’t pro­gram at all. So maybe hack­ing does re­quire some spe­cial abil­ity to fo­cus. Per­haps great hack­ers can load a large amount of con­text into their head, so that when they look at a line of code, they see not just that line but the whole pro­gram around it. John McPhee wrote that Bill Bradley’s suc­cess as a bas­ket­ball player was due partly to his ex­tra­or­di­nary pe­riph­eral vi­sion. ‘Per­fect’ eye­sight means about 47° of ver­ti­cal pe­riph­eral vi­sion. Bill Bradley had 70; he could see the bas­ket when he was look­ing at the floor. Maybe great hack­ers have some sim­i­lar in­born abil­i­ty. (I cheat by us­ing a very dense lan­guage, which shrinks the court.) This could ex­plain the dis­con­nect over cu­bi­cles. Maybe the peo­ple in charge of fa­cil­i­ties, not hav­ing any con­cen­tra­tion to shat­ter, have no idea that work­ing in a cu­bi­cle feels to a hacker like hav­ing one’s brain in a blender.” –, “Great Hack­ers”

It’s sur­pris­ing, but bugs have a close re­la­tion­ship to num­ber of lines of code - no mat­ter whether the lan­guage is as low-level as as­sem­bler or high­-level as Haskell (hu­mor­ous­ly, Nor­ris’ num­ber); is this be­cause each line takes up a sim­i­lar amount of work­ing and short­-term mem­ory and there’s only so much mem­ory to go around?11

The Silver Bullet

It’s not all that ob­vi­ous, but just about every pro­duc­tiv­ity in­no­va­tion in com­put­ing is about ei­ther cut­ting down on how much a pro­gram­mer needs to know (eg. ), or mak­ing it eas­ier for him to shuffle things in and out of his ‘short term mem­ory’. Why are some com­men­ta­tors like so fo­cused12 on hav­ing mul­ti­ple ? For that mat­ter, why are there real stud­ies show­ing sur­pris­ingly large pro­duc­tiv­ity boosts by sim­ply adding a sec­ond mon­i­tor?13 It’s not like the per­son is any differ­ent after­wards. And ar­guably mul­ti­ple or larger mon­i­tors come with dam­ag­ing over­heads14.

Or, why does think touch-typ­ing is one of the few skills pro­gram­mers must know (a­long with read­ing)?15 Why is Unix guru one re­gret not learn­ing typ­ing?16 Typ­ing hardly seems very im­por­tant - it’s what you say, not how you say it. The com­piler does­n’t care if you typed the source code in at 30WPM or 120WPM, after all.

I love be­ing able to type that with­out look­ing! It’s em­pow­er­ing, be­ing able to type al­most as fast as you can think. Why would you want it any other way?

The thing is, mul­ti­ple mon­i­tors, touch-typ­ing, speed-read­ing17 - they’re all about mak­ing the ex­ter­nal world part of your mind. What’s the real differ­ence be­tween hav­ing a in your short­-term mem­ory or promi­nently dis­played in your sec­ond mon­i­tor? What’s the real differ­ence be­tween writ­ing a in your mind or touch-typ­ing it as fast as you cre­ate it?

WM prob­lems

Just some speed. Just some time. And the more vis­i­ble that type sig­na­ture is, the faster you can type out that com­ment, the larger your ‘mem­ory’ gets. And the larger your mem­ory is, the more in­tel­li­gen­t/pro­duc­tive you can be. (Think of this as the the­sis as ap­plied to pro­gram­ming!) Great pro­gram­mers often1819 talk vaguely about ‘keep­ing a sys­tem in your head’ or ‘hav­ing a model’, and hate dis­trac­tions20, say­ing they de­stroy one’s care­fully de­vel­oped thoughts; I think what they are talk­ing about is try­ing to store all the rel­e­vant de­tails in­side their short­-term or work­ing mem­o­ry. Learn­ing pro­gram­ming has a cor­re­la­tion with WM.21 (Once you start look­ing, you see this every­where. Games, for ex­am­ple.22) Or in bug rates - WM has been pro­posed as the rea­son why small or large chunks of pro­grams have more pro­por­tional er­rors than medium sized chunks23. It re­mains to be seen whether pro­gram­ming tools de­signed with an eye to mem­ory will be help­ful, though.

But as great as things like garbage col­lec­tion & touch-typ­ing & mul­ti­ple mon­i­tors are (I am a fan & user of the fore­go­ing), they are still im­per­fect sub­sti­tutes. Would­n’t it be bet­ter if one could just im­prove one’s short­-ter­m/­work­ing mem­ory di­rect­ly? It might be more effec­tive, and cer­tainly would be more portable!


Un­for­tu­nate­ly, in gen­er­al, IQ/ and mem­ory don’t seem to be train­able. Many ap­par­ent effects are swamped by ex­er­cise or nu­tri­tion or by sim­ple prac­tice. And when prac­tice does re­sult in gains on tasks or ex­pen­sive games24, said ben­e­fits often do not ; many pop­u­lar ‘brain games’ & ex­er­cises fail this cri­te­rion or at least have not been shown to trans­fer252627, even brainy skilled ex­er­cises like mu­sic28 or chess29 or mem­ory com­pe­ti­tions30. summed it up:

…Gen­eral Dree­dle wants his [pi­lots] to spend as much time on the skeet-shoot­ing range as the fa­cil­i­ties and their flight sched­ule would al­low. eight hours a month was ex­cel­lent train­ing for them. It trained them to shoot skeet.

In­deed, the gen­eral his­tory of at­tempts to in­crease IQ in any chil­dren or adults re­mains es­sen­tially what it was when wrote his 1969 pa­per —a his­tory of fail­ure. The ex­cep­tions prove the rule by ei­ther ap­ply­ing to nar­row groups with spe­cific deficits or work only be­fore birth, like . (See also : if there were an easy fit­ness-in­creas­ing way to make us smarter, evo­lu­tion would have al­ready used it.)

But hope springs eter­nal, and there are pos­si­ble ex­cep­tions. The one this FAQ fo­cuses on is Dual , and it’s a vari­ant on an old work­ing-mem­ory test.

One of the nice things about N-back is that while it may or may not im­prove your IQ, it may help you in other ways. WM train­ing helps al­co­holics re­duce their con­sump­tion31 and in­creases pa­tience in re­cov­er­ing stim­u­lant ad­dicts (co­caine & metham­phet­a­mine)32. The self­-dis­ci­pline or willpower of stu­dents cor­re­lates bet­ter with grades than even IQ33, WM cor­re­lates with grades and lower be­hav­ioral prob­lems34 & WM out­-pre­dicts grades 6 years later in 5-year olds & 2 years later in older chil­dren35. WM train­ing has been shown to help chil­dren with ADHD36 and also preschool­ers with­out ADHD37; Lu­cas 2008 found be­hav­ior im­prove­ments at a sum­mer camp. An­other in­ter­ven­tion us­ing a mis­cel­lany of ‘rea­son­ing’ games with young (7-9 years old) poor chil­dren found a For­wards Digit Span (but not Back­wards) and IQ gains, with no gain to the sub­jects play­ing games re­quir­ing “rapid vi­sual de­tec­tion and rapid mo­tor re­sponses”38, but it’s worth re­mem­ber­ing that IQ scores are un­re­li­able in child­hood39 or per­haps, as an ado­les­cent brain imag­ing study in­di­cates40, they sim­ply are much more mal­leable at that point. (WM train­ing in teenagers does­n’t seem much stud­ied but given their is­sues, may help; see “Beau­ti­ful Brains” or “The Trou­ble With Teens”.)

There are many kinds of WM train­ing. One re­view worth read­ing is “Does work­ing mem­ory train­ing work? The promise and chal­lenges of en­hanc­ing cog­ni­tion by train­ing work­ing mem­ory” (Mor­ri­son & Chein 2011); “Is Work­ing Mem­ory Train­ing Effec­tive?” (Ship­stead, Redick, & En­gle 2012) dis­cusses the mul­ti­ple method­olog­i­cal diffi­cul­ties of de­sign­ing WM train­ing ex­per­i­ments (at least, they are diffi­cult if you want to show gen­uine im­prove­ments which trans­fer to non-WM skill­s).


The orig­i­nal N-back test sim­ply asked that you re­mem­ber a sin­gle stream of let­ters, and sig­nal if any let­ters were pre­cise­ly, say, 2 po­si­tions apart. ‘A S S R’ would­n’t merit a sig­nal, but ‘A S A R’ would since there are ‘A’ char­ac­ters ex­actly 2 po­si­tions away from each oth­er. The pro­gram would give you an­other let­ter, you would sig­nal or not, and so on. This is sim­ple enough once you un­der­stand it, but is a lit­tle hard to ex­plain. It may be best to read the Brain Work­shop tu­to­r­ial, or watch a video.

Dual N-back

In 2003, Su­san Jaeggi and her team be­gan stud­ies us­ing a vari­ant of N-back which tried to in­crease the bur­den on each turn - re­mem­ber­ing mul­ti­ple things in­stead of just 1. The ab­stract de­scribes the rea­son why:

With ref­er­ence to sin­gle tasks, ac­ti­va­tion in the pre­frontal cor­tex (PFC) com­monly in­creases with in­cre­men­tal mem­ory load, whereas for dual tasks it has been hy­poth­e­sized pre­vi­ously that ac­tiv­ity in the PFC de­creases in the face of ex­ces­sive pro­cess­ing de­mands, i.e., if the ca­pac­ity of the work­ing mem­o­ry’s cen­tral ex­ec­u­tive sys­tem is ex­ceed­ed. How­ev­er, our re­sults show that dur­ing both sin­gle and dual tasks, pre­frontal ac­ti­va­tion in­creases con­tin­u­ously as a func­tion of mem­ory load. An in­crease of pre­frontal ac­ti­va­tion was ob­served in the dual tasks even though pro­cess­ing de­mands were ex­ces­sive in the case of the most diffi­cult con­di­tion, as in­di­cated by be­hav­ioral ac­cu­racy mea­sures. The hy­poth­e­sis con­cern­ing the de­crease in pre­frontal ac­ti­va­tion could not be sup­ported and was dis­cussed in terms of mo­ti­va­tion fac­tors.41

In this ver­sion, called “dual N-back” (to dis­tin­guish it from the clas­sic sin­gle N-back), one is still play­ing a turn-based game. In the Brain Work­shop ver­sion, you are pre­sented with a 3x3 grid in which every turn, a block ap­pears in 1 of the 9 spaces and a let­ter is spo­ken aloud. (There are any num­ber of vari­ants: the NATO pho­netic al­pha­bet, pi­ano keys, etc. And Brain Work­shop has any num­ber of mod­es, like ‘Arith­metic N-back’ or ‘Quin­tu­ple N-back’.)


In 1-back, the task is to cor­rectly an­swer whether the let­ter is the same as the pre­vi­ous round, and whether the po­si­tion is the same as the pre­vi­ous round. It can be both, mak­ing 4 pos­si­ble re­sponses (po­si­tion, sound, po­si­tion+­sound, & nei­ther).

This stresses work­ing mem­ory since you need to keep in mind 4 things si­mul­ta­ne­ous­ly: the po­si­tion and let­ter of the pre­vi­ous turn, and the po­si­tion and let­ter of the cur­rent turn (so you can com­pare the cur­rent let­ter with the old let­ter and the cur­rent po­si­tion with the old po­si­tion). Then on the next turn you need to im­me­di­ately for­get the old po­si­tion & let­ter (which are now use­less) and re­mem­ber the new po­si­tion and let­ter. So you are con­stantly re­mem­ber­ing and for­get­ting and com­par­ing.


But 1-back is pretty easy. The turns come fast enough that you could eas­ily keep the let­ters in your and lighten the load on your work­ing mem­o­ry. In­deed, after 10 rounds or so of 1-back, I mas­tered it - I now get 100%, un­less I for­get for a sec­ond that it’s 1-back and not 2-back (or I sim­ply lose my con­cen­tra­tion com­plete­ly). Most peo­ple find 1-back very easy to learn, al­though a bit chal­leng­ing at first since the pres­sure is con­stant (games and tests usu­ally have some slack or rest pe­ri­od­s).

The next step up is a doozy: 2-back. In 2-back, you do the same thing as 1-back but as the name sug­gests, you are in­stead match­ing against 2 turns ago. So be­fore you would be look­ing for re­peated let­ters - ‘AA’ - but now you need to look for sep­a­rated let­ters - ‘ABA’. And of course, you can’t for­get so quick­ly, since you still need to match against some­thing like ‘ABABA’.

2-back stresses your work­ing mem­ory even more, as now you are re­mem­ber­ing 6 things, not 4: 2 turns ago, the pre­vi­ous turn, and the cur­rent turn - all of which have 2 salient fea­tures. At 6 items, we’re also in the mid-range of es­ti­mates for :

Work­ing mem­ory is gen­er­ally con­sid­ered to have lim­ited ca­pac­i­ty. The ear­li­est quan­tifi­ca­tion of the ca­pac­ity limit as­so­ci­ated with short­-term mem­ory was the in­tro­duced by Miller (1956). He no­ticed that the mem­ory span of young adults was around seven el­e­ments, called chunks, re­gard­less whether the el­e­ments were dig­its, let­ters, words, or other units. Later re­search re­vealed that span does de­pend on the cat­e­gory of chunks used (e.g., span is around seven for dig­its, around six for let­ters, and around five for word­s), and even on fea­tures of the chunks within a cat­e­go­ry….­Sev­eral other fac­tors also affect a per­son’s mea­sured span, and there­fore it is diffi­cult to pin down the ca­pac­ity of short­-term or work­ing mem­ory to a num­ber of chunks. Nonethe­less, Cowan (2001) has pro­posed that work­ing mem­ory has a ca­pac­ity of about four chunks in young adults (and fewer in chil­dren and old adult­s).

And even if there are only a few things to re­mem­ber, the num­ber of re­sponses you have to choose be­tween go up ex­po­nen­tially with how many ‘modes’ there are, so Triple N-back has not ⅓ more pos­si­ble re­sponses than Dual N-back, but more than twice as many: if m is the num­ber of mod­es, then the num­ber of pos­si­ble re­sponses is 2m-1 (the -1 is there be­cause one can noth­ing in every mode, but that’s bor­ing and re­quires no choice or thought), so DNB has 3 pos­si­ble re­sponses42, while TNB has 743, Quadru­ple N-back 1544, and Quin­tu­ple N-back 3145!

Worse, the tem­po­ral gap be­tween el­e­ments is deeply con­fus­ing. It’s par­tic­u­larly bad when there’s rep­e­ti­tion in­volved - if the same square is se­lected twice with the same let­ter, you might wind up for­get­ting both!

So 2-back is where the chal­lenge first re­ally man­i­fests. After about 20 games I started to get the hang of it. (It helped to play a few games fo­cus­ing only on one of the stim­uli, like the let­ters; this helps you get used to the ‘reach­ing back’ of 2-back.)

Personal reflection on results

Have I seen any ben­e­fits yet? Not re­al­ly. Thus far it’s like med­i­ta­tion: I haven’t seen any spe­cific im­prove­ments, but it’s been in­ter­est­ing just to ex­plore con­cen­tra­tion - I’ve learned that my abil­ity to fo­cus is much less than I thought it was! It is very sober­ing to get 30% scores on some­thing as triv­ial as 1-back and strain to reach D2B, and even more sober­ing to score 60% and min­utes later score 20%. Be­sides the in­trin­sic in­ter­est of chang­ing one’s brain through a sim­ple ex­er­cise - med­i­ta­tion is equally in­ter­est­ing for how one’s mind re­fuses to co­op­er­ate with the sim­ple work of med­i­tat­ing, and I un­der­stand that there are even vivid hal­lu­ci­na­tions at the higher lev­els - N-back might func­tion as a kind of men­tal cal­is­then­ics. Few peo­ple ex­er­cise and stretch be­cause they find the ac­tiv­i­ties in­trin­si­cally valu­able, but they serve to fur­ther some other goal; some peo­ple jog be­cause they just en­joy run­ning, but many more jog so they can play soc­cer bet­ter or live longer. I am young, and it’s good to ex­plore these sorts of cal­is­then­ics while one has a long life ahead of one; then one can reap the most ben­e­fits.


N-back train­ing is some­times re­ferred to sim­ply as ‘N-back­ing’, and par­tic­i­pants in such train­ing are called ‘N-back­ers’. Al­most every­one uses the Free, fea­ture­ful & portable pro­gram Brain Work­shop, ab­bre­vi­ated “BW” (but see the soft­ware sec­tion for al­ter­na­tives).

There are many vari­ants of N-back train­ing. A 3-let­ter acronym end­ing in ‘B’ spec­i­fies one of the pos­si­bil­i­ties. For ex­am­ple, ‘D2B’ and ‘D6B’ both re­fer to a dual N-back task, but in the for­mer the depth of re­call is 2 turns, while in the lat­ter one must re­mem­ber back 6 rounds; the ‘D’, for ‘Dual’, in­di­cates that each round presents 2 stim­uli (usu­ally the po­si­tion of the square, and a spo­ken let­ter).

But one can add fur­ther stim­uli: spo­ken let­ter, po­si­tion of square, and color of square. That would be ‘Triple N-back’, and so one might speak of how one is do­ing on ‘T4B’.

One can go fur­ther. Spo­ken let­ter, po­si­tion, col­or, and geo­met­ric shape. This would be ‘Quad N-back’, so one might dis­cuss one’s per­for­mance on ‘Q3B’. (It’s un­clear how to com­pare the var­i­ous mod­es, but it seems to be much harder to go from D2B to T3B than to go from D2B to D3B.)

Past QNB, there is Pen­tu­ple N-back (PNB) which was added in Brain Work­shop 4.7 (video demon­stra­tion). The 5th modal­ity is added by a sec­ond au­dio chan­nel - that is, now sounds are in stereo.

Other ab­bre­vi­a­tions are in com­mon use: ‘WM’ for ‘work­ing mem­ory’, ‘Gf’ for ‘’, and ‘g’ for the mea­sured by IQ tests.

Notes from the author

N-back in general

To those whose time is lim­it­ed: you may wish to stop read­ing here. If you seek to im­prove your life, and want the great­est ‘bang for the buck’, you are well-ad­vised to look else­where.

Med­i­ta­tion, for ex­am­ple, is eas­ier, faster, and ul­tra­-portable. Typ­ing train­ing will di­rectly im­prove your fa­cil­ity with a com­put­er, a valu­able skill for this mod­ern world. mem­o­riza­tion tech­niques offer un­par­al­leled ad­van­tages to stu­dents. are the epit­ome of ease (just swal­low!), and their effects are much more eas­ily as­sessed - one can even run dou­ble-blind ex­per­i­ments on one­self, im­pos­si­ble with dual N-back. Other sup­ple­ments like can de­liver ben­e­fits in­com­men­su­rable with DNB - what is the cog­ni­tive value of an­other num­ber in work­ing mem­ory thanks to DNB com­pared to a good night’s sleep thanks to mela­ton­in? Mod­est changes to one’s diet and en­vi­rons can fun­da­men­tally im­prove one’s well-be­ing. Even ba­sic train­ing in read­ing, with the crud­est tech­niques, can pay large div­i­dends if one is be­low a ba­sic level of read­ing like 200 & still sub­vo­cal­iz­ing. And all of these can start pay­ing off im­me­di­ate­ly.

DNB, on the other hand, re­quires a min­i­mum of 15 hours be­fore one can ex­pect gen­uine so­matic im­prove­ments. The task it­self is un­proven - the Jaeggi stud­ies are sug­ges­tive, not de­fin­i­tive (and there are con­trary re­sults). Pro­grams for DNB train­ing rely es­sen­tially on guess­work as they ex­plore the large de­sign-space; there are no data on what fea­tures are es­sen­tial, what sort of pre­sen­ta­tion op­ti­mal, or even how long or when to train for. The task it­self is un­en­joy­able. It can be weary­ing, diffi­cult & em­bar­rass­ing. It can be one too many daily tasks, a straw which breaks the camel’s back, and a dis­trac­tion from what­ever ac­tiv­ity has the great­est for one46 and one ought to be do­ing in­stead.

So why then do I per­se­vere with DNB?

I do it be­cause I find it fas­ci­nat­ing. Fas­ci­nat­ing that WM can be so large a part of IQ; fas­ci­nat­ing that it can be in­creased by an ap­par­ently triv­ial ex­er­cise. I’m fas­ci­nated that there are mea­sur­able gross changes in brain ac­tiv­ity & chem­istry & com­po­si­tion47 - that the effects are not purely ‘men­tal’ or place­bo. I’m fas­ci­nated by how the se­quence of po­si­tions and let­ters can at some times ap­pear in my mind with bound­less lu­cid­i­ty, yet at other times I grope con­fused in a men­tal murk un­sure of even what the last po­si­tion/let­ter was - even though I can rise from my com­puter and go about nor­mal ac­tiv­i­ties nor­mal­ly; or with how time can stretch and com­press dur­ing N-back­ing48. I’m fas­ci­nated by how a sin­gle in­crease in n-level can ren­der the task night­mar­ishly diffi­cult when I just fin­ished n-1 at 90 or 100%. I’m fas­ci­nated by how sac­cad­ing, an­other ap­par­ently triv­ial ex­er­cise, can re­li­ably boost my score by 10 or 20%, and how my mind seems to be fagged after just a few rounds but re­cov­ers within min­utes. I’m equally fas­ci­nated by the large lit­er­a­ture on WM: what it is, what’s it good for, how it can be ma­nip­u­lat­ed, etc.

I do not think that DNB is ter­ri­bly prac­ti­cal - but in­ter­est­ing? Very.

Reading this FAQ

Brian: “Look, you’ve got it all wrong! You don’t need to fol­low me, You don’t need to fol­low any­body! You’ve got to think for your selves! You’re all in­di­vid­u­als!”

The Crowd: “Yes! We’re all in­di­vid­u­als!”49

This FAQ is al­most solely my own work. I’ve striven to make it fair, to in­cor­po­rate most of the rel­e­vant re­search, and to not omit things. But in­evitably I will have made er­rors or im­por­tant omis­sions. You must read this skep­ti­cal­ly.

You must read this skep­ti­cally also be­cause the N-back com­mu­nity formed around the mail­ing list is a com­mu­ni­ty. That means it is prone to all the bi­ases and is­sues of a com­mu­ni­ty. One would ex­pect a com­mu­nity formed around a tech­nique or prac­tice to be made up only of peo­ple who find value in it; any ma­te­r­ial (like this FAQ or in­cluded tes­ti­mo­ni­als) is au­to­mat­i­cally sus­pect due to bi­ases such as the . Imag­ine if sci­en­tists pub­lished only pa­pers which showed new re­sults, and no pa­pers re­port­ing fail­ure to ! Why would any N-backer hang around who had dis­cov­ered that DNB was not use­ful or a fraud? Cer­tainly the fans would not thank him. ( has an ex­cel­lent es­say called “Evap­o­ra­tive Cool­ing of Group Be­liefs” on this top­ic; for­tu­nate­ly, the dam­age caused by a dual n-back would be lim­it­ed, in com­par­i­son to some other ex­am­ples of evap­o­ra­tive cool­ing like or mind-con­trol vic­tims.)

Fi­nal­ly, you must read skep­ti­cally be­cause this is about psy­chol­o­gy. Psy­chol­ogy is no­to­ri­ously for be­ing one of the hard­est sci­en­tific fields to get solid re­sults in, be­cause every­body is WEIRD and differ­ent. As one of my pro­fes­sors joked, “if you have 2 psy­chol­ogy pa­pers re­port­ing the same re­sult, one of them is wrong”; there are many is­sues with tak­ing a psy­chol­ogy study at face value (to which I have de­voted an ap­pen­dix, “Flaws in main­stream sci­ence (and psy­chol­o­gy)”). It’s very tempt­ing to en­gage in but you must­n’t. Every­body is differ­ent; your pos­i­tive (or neg­a­tive) re­sult could be due to a , it could be thanks to that re­cent shift in your sleep sched­ule for the bet­ter50, or that nap you took51, it could be the ex­er­cise you’re get­ting52, it could be a mild lift­ing (or set­ting in), it could be a cal­cium or 53 or io­dine de­fi­ciency, 545556, vari­a­tion in mo­ti­va­tion etc.

N-back training

Should I do multiple daily sessions, or just one?

Most users seem to go for one long N-back ses­sion, point­ing out that ex­er­cises one’s fo­cus. Oth­ers do one ses­sion in the morn­ing and one in the evening so they can fo­cus bet­ter on each one. There is some sci­en­tific sup­port for the idea that evening ses­sions are bet­ter than morn­ing ses­sions, though; see Kuriyama 2008 on how prac­tice be­fore bed­time was more effec­tive than after wak­ing up.

If you break up ses­sions into more than 2, you’re prob­a­bly wast­ing time due to over­head, and may not be get­ting enough ex­er­cise in each ses­sion to re­ally strain your­self like you need to.


The sim­plest men­tal strat­e­gy, and per­haps the most com­mon, is to men­tally think of a list, and for­get the last one each round, re­mem­ber­ing the newest in its place. This be­gins to break down on higher lev­els - if one is re­peat­ing the list men­tal­ly, the rep­e­ti­tion can just take too long.

Surcer writes up a list of strate­gies for differ­ent lev­els in his “My Sys­tem, let’s share strate­gies” thread.

Are strategies good or bad?

Peo­ple fre­quently ask and dis­cuss whether they should use some sort of strat­e­gy, and if so, what.

A num­ber of N-back­ers adopt an ‘in­tu­ition’ strat­e­gy. Rather than ex­plic­itly re­hears­ing se­quences of let­ters (‘f-up, h-mid­dle; f-up, h-mid­dle; g-down, f-up…’), they sim­ply think very hard and wait for a feel­ing that they should press ‘a’ (au­dio match), or ‘l’ (lo­ca­tion match). Some, like SwedishChef can be quite vo­cif­er­ous about it:

The chal­lenges are in help­ing peo­ple un­der­stand that du­al-n-back is NOT about re­mem­ber­ing n num­ber of vi­sual and au­di­tory stim­uli. It’s about de­vel­op­ing a new men­tal process that in­tu­itively rec­og­nizes when it has seen or heard a stim­uli n times ago.

Ini­tial­ly, most stu­dents of dual n-back want to re­mem­ber n items as fast as they can so they can con­quer the du­al-n-back hill. They use their own al­ready de­vel­oped tech­niques to help them re­mem­ber. They may try to hold the im­ages in their head men­tally and re­view them every time a new im­age is added and say the sounds out loud and re­view the sounds every time a new sound is added. This is NOT what we want. We want the brain to learn a new process that in­tu­itively rec­og­nizes if an item and sound was shown 3 back or 4 back. It’s sort of like play­ing a new type of mu­si­cal in­stru­ment.

I’ve helped some stu­dents on the site try to un­der­stand this. It’s not about how much you can re­mem­ber, it’s about learn­ing a new process. In the­o­ry, this new process trans­lates into a bet­ter work­ing mem­o­ry, which helps you make con­nec­tions bet­ter and faster.

Other N-back­ers think that in­tu­ition can’t work, or at least does­n’t very well:

I don’t be­lieve that much in the “in­tu­itive” method. I mean, sure, you can in­tu­itively re­mem­ber you heard the same let­ter or saw the square at the same po­si­tion a few times ago, but I fail to see how you can “feel” it was ex­actly 6 or 7 times ago with­out some kind of “ac­tive” re­mem­ber­ing. –Gaël DEEST

I to­tally agree with Gaël about the in­tu­itive method not hold­ing much wa­ter…­For me a lot of times the in­tu­itive method can be to­tally un­re­li­able. You’ll be do­ing 5-back one game and a few games later your fail­ing mis­er­ably at 3-back­..y­our score all over the place. Plus, in­tu­itive-wise, it’s best to play the same n-back level over and over be­cause then you train your in­tu­ition…and that does­n’t seem right. –MikeM (same thread)

Few N-back­ers have sys­tem­at­i­cally tracked in­tu­itive ver­sus strate­gic play­ing; Dark­Alrx re­ports on his blog the re­sults of his ex­per­i­ment, and while he con­sid­ers them pos­i­tive, oth­ers find them in­con­clu­sive, or like Pheonex­ia, even un­fa­vor­able for the in­tu­itive ap­proach:

Look­ing at your graphs and the over­all drop in your per­for­mance, I think it’s clear that in­tu­itive does­n’t work. On your score sheet, the first pic­ture, us­ing the in­tu­itive method over 38 days of TNB train­ing in 44 days your av­er­age n-back in­creased by less than .25. You were per­form­ing much bet­ter be­fore. With your neu­ro­ge­n­e­sis ex­per­i­ment, your av­er­age n-back ac­tu­ally de­creased.

Jaeggi her­self was more mod­er­ate in ~2008:

I would NOT rec­om­mend you [train the vi­sual and au­di­tory task sep­a­rate­ly] if you want to train the du­al-task (the one we used in our study). The rea­son is that the com­bi­na­tion of both modal­i­ties is an en­tirely differ­ent task than do­ing both sep­a­rate­ly! If you do the task sep­a­rate­ly, I as­sume you use some “re­hearsal strate­gies”, e.g. you re­peat the let­ters or po­si­tions for your­self. In the du­al-task ver­sion how­ev­er, these strate­gies might be more diffi­cult to ap­ply (s­ince you have to do 2 things si­mul­ta­ne­ous­ly…), and that is ex­actly what we want… We don’t want to train strate­gies, we want to train process­es. Processes that then might help you in the per­for­mance of oth­er, non-trained tasks (and that is our ul­ti­mate goal). So, it is not im­por­tant to reach a 7- or 8-back… It is im­por­tant to fully fo­cus your at­ten­tion on the task as well as pos­si­ble.

I can as­sure you, it is a very tough train­ing reg­i­men…. You can’t di­vert your at­ten­tion even 1 sec­ond (I’m sure you have no­ticed…). But even­tu­al­ly, you will see that you get bet­ter at it and maybe you no­tice that you are bet­ter able to con­cen­trate on cer­tain things, to re­mem­ber things more eas­i­ly, etc. (hope­ful­ly).

(Un­for­tu­nate­ly, doubt has been cast on this ad­vice by the ap­par­ent effec­tive­ness of sin­gle n-back in Jaeggi 2010. If sin­gle (vi­su­al/­po­si­tion) n-back is effec­tive in in­creas­ing IQ, then maybe train­ing just au­dio or just vi­sual is ac­tu­ally a good idea.)

this is a ques­tion i am be­ing asked a lot and un­for­tu­nate­ly, i don’t re­ally know whether i can help with that. i can only tell you what we tell (or rather not tell) our par­tic­i­pants and what they tell us. so, first of all, we don’t tell peo­ple at all what strat­egy to use - it is up to them. thing is, there are some peo­ple that tell us what you de­scribe above, i.e. some of them tell us that it works best if they don’t use a strat­egy at all and just “let the squares/let­ters flow by”. but of course, many par­tic­i­pants also use more con­scious strate­gies like re­hears­ing or group­ing items to­geth­er. but again - we let peo­ple chose their strate­gies them­selves! ref

But it may make no differ­ence. Even if you are en­gaged in a com­plex mnemon­ic-based strat­e­gy, you’re still work­ing your mem­o­ry. Strate­gies may not work; quot­ing from Jaeg­gi’s 2008 pa­per:

By this ac­count, one rea­son for hav­ing ob­tained trans­fer be­tween work­ing mem­ory and mea­sures of Gf is that our train­ing pro­ce­dure may have fa­cil­i­tated the abil­ity to con­trol at­ten­tion. This abil­ity would come about be­cause the con­stant up­dat­ing of mem­ory rep­re­sen­ta­tions with the pre­sen­ta­tion of each new stim­u­lus re­quires the en­gage­ment of mech­a­nisms to shift at­ten­tion. Al­so, our train­ing task dis­cour­ages the de­vel­op­ment of sim­ple task-spe­cific strate­gies that can pro­ceed in the ab­sence of con­trolled al­lo­ca­tion of at­ten­tion.

Even if they do, they may not be a good idea; quot­ing from Jaeggi 2010:

We also pro­posed that it is im­por­tant that par­tic­i­pants only min­i­mally learn task-spe­cific strate­gies in or­der to pre­vent spe­cific skill ac­qui­si­tion. We think that be­sides the trans­fer to ma­trix rea­son­ing, the im­prove­ment in the near trans­fer mea­sure pro­vides ad­di­tional ev­i­dence that the par­tic­i­pants trained on task-un­der­ly­ing processes rather than re­ly­ing on ma­te­ri­al-spe­cific strate­gies.

Hope­fully even if a trick lets you jump from 3-back to 5-back, Brain Work­shop will just keep es­ca­lat­ing the diffi­culty un­til you are chal­lenged again. It’s not the level you reach, but the work you do.

And the flashing right/wrong feedback?

A mat­ter of pref­er­ence, al­though those in fa­vor of dis­abling the vi­sual feed­back (SHOW_FEEDBACK = False) seem to be slightly more vo­cal or nu­mer­ous. Brain Twister ap­par­ently does­n’t give feed­back. Jaeggi says:

the gam­ing lit­er­a­ture also dis­agrees on this is­sue - there are differ­ent ways to think about this: whereas feed­back after each trial gives you im­me­di­ate feed­back whether you did right or wrong, it can also be dis­tract­ing as you are con­stantly mon­i­tor­ing (and eval­u­at­ing) your per­for­mance. we de­cided that we wanted peo­ple to fully and max­i­mally con­cen­trate on the task it­self and thus chose the ap­proach to only give feed­back at the end of the run. how­ev­er, we have newer ver­sions of the task for kids in which we give some sort of feed­back (points) for each tri­al. thus - i can’t tell you what the op­ti­mal way is - i guess there are in­terindi­vid­ual differ­ences and pref­er­ences as well.

Jonathan Toomim writes:

When I was do­ing vi­sual psy­chophysics re­search, I heard from my lab­mates that this ques­tion has been in­ves­ti­gated em­pir­i­cally (at least in the con­text of vi­sual psy­chophysic­s), and that the con­sen­sus in the field is that us­ing feed­back re­duces im­me­di­ate per­for­mance but im­proves learn­ing rates. I haven’t looked up the re­search to con­firm their opin­ion, but it sounds plau­si­ble to me. I would also ex­pect it to ap­ply to Brain Work­shop. The idea, as I see it, is that feed­back re­duces per­for­mance be­cause, when you get an an­swer wrong and you know it, your brain goes into an in­tro­spec­tive mode to an­a­lyze the rea­son for the er­ror and (hope­ful­ly) cor­rect it, but while in this mode your brain will be dis­tracted from the task at hand and will be more likely to miss sub­se­quent tri­als.

How can I do better on N-back?

Fo­cus hard­er. Play more. Sleep well, and eat health­ily. Use nat­ural light­ing57. Space out prac­tice. The less stressed you are, the bet­ter you can do.


Pen­ner et al 2012

This study com­pared a high in­ten­sity work­ing mem­ory train­ing (45 min­utes, 4 times per week for 4 weeks) with a dis­trib­uted train­ing (45 min­utes, 2 times per week for 8 weeks) in mid­dle-aged, healthy adult­s…Our re­sults in­di­cate that the dis­trib­uted train­ing led to in­creased per­for­mance in all cog­ni­tive do­mains when com­pared to the high in­ten­sity train­ing and the con­trol group with­out train­ing. The most sig­nifi­cant differ­ences re­vealed by in­ter­ac­tion con­trasts were found for ver­bal and vi­sual work­ing mem­o­ry, ver­bal short­-term mem­ory and men­tal speed.

This is rem­i­nis­cent of sleep’s in­volve­ment in other forms of mem­ory and cog­ni­tive change, and Kuriyama 2008.


Cur­tis War­ren has no­ticed that when he un­der­went a 4-day rou­tine of prac­tic­ing more than 4 hours a day, he jumped an en­tire level on even quad N-back58:

For ex­am­ple, over the past week I have been try­ing a new train­ing rou­tine. My goal was to in­crease my in­tel­li­gence as quickly as pos­si­ble. To that end, over the past 4 days I’ve done a to­tal of roughly 360 ses­sions @ 2 sec­onds per trial (≈360 min­utes of train­ing). I had to rest on Wednes­day, and I’m rest­ing again to­day (I only plan on do­ing about 40 tri­als to­day). But I in­tend to fin­ish off the week by do­ing 100 ses­sions on Sat­ur­day and an­other 100 on Sun­day. Or more, if I can man­age it.

But he cau­tions us that be­sides be­ing a con­sid­er­able time in­vest­ment, it may only work for him:

The point is, while I can say with­out a doubt that this sched­ule has been effec­tive for me, it might not be effec­tive for you. Are the ben­e­fits worth the amount of work need­ed? Will you even no­tice an im­prove­ment? Is this healthy? These are all fac­tors which de­pend en­tirely upon the in­di­vid­ual ac­tu­ally do­ing the train­ing.

Ra­man started DNB train­ing, and in his first 30 days, he “took breaks every 5 days or so, and was do­ing about 20-30 ses­sion each day and n-back wise I made good gains (from 2 to 7 touch­ing 9 on the way).”; he kept a jour­nal on the mail­ing list about the ex­pe­ri­ence with daily up­dates.

Alas, nei­ther Ra­man nor War­ren took an IQ or dig­it-s­pan test be­fore start­ing, so they can only re­port DNB level in­creases & sub­jec­tive as­sess­ments.

The re­search does sug­gest that di­min­ish­ing re­turns does not set in with train­ing regimes of 10 or 15 min­utes a day; for ex­am­ple, Nut­ley 2011 trained 4-year-olds in WM ex­er­cis­es, Gf (NVR) ex­er­cis­es, or both:

…These analy­ses took into ac­count that the groups differed in the amount of train­ing re­ceived, full dose for NVR or WM groups or half dose for the CB group (Table 3). Even though the pat­tern is not con­sis­tent across all tests (see Fig­ure 2), this is in­ter­preted as con­fir­ma­tion of the lin­ear dose effect that was ex­pected to be seen. Our re­sults sug­gest that the amount of trans­fer to non-trained tasks within the trained con­struct was roughly pro­por­tion­ate to the amount of train­ing on that con­struct. A sim­i­lar find­ing, with trans­fer pro­por­tional to amount of train­ing, was re­ported by Jaeggi et al. (2008). This has pos­si­ble im­pli­ca­tions for the de­sign of fu­ture cog­ni­tive train­ing par­a­digms and sug­gests that the train­ing should be in­ten­sive enough to lead to sig­nifi­cant trans­fer and that train­ing more than one con­struct does not en­tail any ad­van­tages in it­self. The train­ing effect pre­sum­ably reaches as­ymp­tote, but where this oc­curs is for fu­ture stud­ies to de­ter­mine. It is prob­a­bly im­por­tant to en­sure that par­tic­i­pants spend enough time on each task in or­der to see clin­i­cally sig­nifi­cant trans­fer, which may be diffi­cult when in­creas­ing the num­ber of tasks be­ing trained. This may be one of the ex­pla­na­tions for the lack of trans­fer seen in the (train­ing six tasks in 10 min­utes).

Plateauing, or, am I wasting time if I can’t get past 4-back?

Some peo­ple start n-back­ing with great vigor and rapidly as­cend lev­els un­til sud­denly they stop im­prov­ing and panic, won­der­ing if some­thing is wrong with them. Not at all! Reach­ing a high level is a good thing, and if one does so in just a few weeks, all the more im­pres­sive since most mem­bers take much longer than, say, 2 weeks to reach good scores on D4B. In fact, if you look at the re­ports in the Group sur­vey, most re­ports are of plateau­ing at D4B or D5B months in.

The cru­cial thing about N-back is just that you are stress­ing your work­ing mem­o­ry, that’s all. The ac­tual level does­n’t mat­ter very much, just whether you can barely man­age it; it is some­what like lift­ing weights, in that re­gard. From Jaeggi 2008:

The find­ing that the trans­fer to Gf re­mained even after tak­ing the spe­cific train­ing effect into ac­count seems to be coun­ter­in­tu­itive, es­pe­cially be­cause the spe­cific train­ing effect is also re­lated to train­ing time. The rea­son for this ca­pac­ity might be that par­tic­i­pants with a very high level of n at the end of the train­ing pe­riod may have de­vel­oped very task spe­cific strate­gies, which ob­vi­ously boosts n-back per­for­mance, but may pre­vent trans­fer be­cause these strate­gies re­main too task-spe­cific (5, 20). The av­er­aged n-back level in the last ses­sion is there­fore not crit­i­cal to pre­dict­ing a gain in Gf; rather, it seems that work­ing at the ca­pac­ity limit pro­motes trans­fer to Gf.

Mail­ing list mem­bers re­port ben­e­fits even if they have plateaued at 3 or 4-back; see the ben­e­fits sec­tion.

One com­monly re­ported tac­tic to break a plateau­ing is to de­lib­er­ately ad­vance a level (or in­crease modal­i­ties), and prac­tice hard on that ex­tra diffi­cult task, the idea be­ing that this will spur adap­ta­tion and make one ca­pa­ble of the lower lev­el.

Do breaks undo my work?

Some peo­ple have won­dered if not n-back­ing for a day/week/­month or other ex­tended pe­riod un­does all their hard work, and hence n-back­ing may not be use­ful in the long-term.

Mul­ti­ple group mem­bers have pointed to long gaps in their train­ing, some­times mul­ti­ple months up to a year, which did not change their scores sig­nifi­cantly (im­me­di­ately after the break, scores may dip a level or a few per­cent­age points in ac­cu­ra­cy, but quickly rises to the old lev­el). Some mem­bers have ceased n-back­ing for 2 or 3 years, and found their scores dropped by only 2-4 lev­els - far from 1 or 2-back. (Pon­tus Granström, on the other hand, took a break for sev­eral months and fell for a long pe­riod from D8B-D9B to D6B-D7B; he spec­u­lates it might re­flect a lack of mo­ti­va­tion.) huh­what/Nova fell 5 lev­els from D9B but re­cov­ered quick­ly:

I’ve been train­ing with n-back on and off, mostly off, for the past few years. I started about 3 years ago and was able to get up to 9-n back, but on av­er­age I would be do­ing around 6 or 7 n back. Then I took a break for a few years. Now after com­ing back, even though I have had my fair share of par­ty­ing, box­ing, light drugs, even polypha­sic sleep, on my first few tries I was able to get back up to 5-6, and a week into it I am back at get­ting up to 9 n back.

This anec­do­tal ev­i­dence is sup­ported by at least one WM-train­ing let­ter, Chrabaszcz 2010:

Fig­ure 1b il­lus­trates the de­gree to which train­ing trans­ferred to an os­ten­si­bly differ­ent (and un­trained) mea­sure of ver­bal work­ing mem­ory com­pared to a no-con­tact con­trol group. Not only did train­ing sig­nifi­cantly in­crease ver­bal work­ing mem­o­ry, but these gains per­sisted 3 months fol­low­ing the ces­sa­tion of train­ing!

Sim­i­lar­ly, Dahlin 2008 found WM train­ing gains which were durable over more than a year:

The au­thors in­ves­ti­gated im­me­di­ate train­ing gains, trans­fer effects, and 18-month main­te­nance after 5 weeks of com­put­er-based train­ing in up­dat­ing of in­for­ma­tion in work­ing mem­ory in young and older sub­jects. Trained young and older adults im­proved sig­nifi­cantly more than con­trols on the cri­te­rion task (let­ter mem­o­ry), and these gains were main­tained 18 months lat­er. Trans­fer effects were in gen­eral lim­ited and re­stricted to the young par­tic­i­pants, who showed trans­fer to an un­trained task that re­quired up­dat­ing (3-back)…

I heard 12-back is possible

Some users have re­ported be­ing able to go all the way up to 12-back; Ashirgo reg­u­larly plays at D13B, but the high­est at other modes seems to be T9B and Q6B.

Ashirgo offers up her 8-point scheme as to how to ac­com­plish such feats:

  1. ’Be fo­cused at all cost. The fluid in­tel­li­gence it­self is some­times called “the strength of fo­cus”.
  2. You had bet­ter not re­hearse the last po­si­tion/­sound . It will even­tu­ally de­crease your per­for­mance! I mean the re­hearsal “step by step”: it will slow you down and dis­tract. The only re­hearsal al­lowed should be nearly un­con­scious and “effort­less” (you will soon re­al­ize its mean­ing :)
  3. Both points 1 & 2 thus im­ply that you must be fo­cused on the most cur­rent stim­u­lus as strongly as you can. Nev­er­the­less, you can­not for­get about the pre­vi­ous stim­uli. How to do that? You should hold the im­age of them (im­age, pic­ture, draw­ing, what­ever you like) in your mind. No­tice that you still do not re­hearse any­thing that way.
  4. Con­sider di­vid­ing the stream of data (n) on smaller parts. 6-back will be then two 3-back, for in­stance.
  5. Fol­low square with your eyes as it changes its po­si­tion.
  6. Just turn on the Jaeggi mode with all the op­tions to en­sure your task is clos­est to the orig­i­nal ver­sion.
  7. Con­sider do­ing more than 20 tri­als. I am on my way to do no less than 30 to­day. It may also help.
  8. You may lower the diffi­culty by re­duc­ing the fal­l-back and ad­vance lev­els from >75 and =<90 to 70 and 85 re­spec­tively (for in­stance).’

What’s some relevant research?

Train­ing WM tasks has yielded a lit­er­a­ture of mixed re­sults - for every pos­i­tive, there’s a neg­a­tive, it seems. The fol­low­ing sec­tions of pos­i­tive and null re­sults il­lus­trate that, as do the pa­pers them­selves; from Nut­ley 2011:

How­ev­er, there are some stud­ies us­ing sev­eral WM tasks to train that have also shown trans­fer effects to rea­son­ing tasks (Kling­berg, Fer­nell, Ole­sen, John­son, Gustafs­son, Dahlstrçm, Gill­berg, Forss­berg & West­er­berg, 2005; Kling­berg, Forss­berg & West­er­berg, 2002), while other WM train­ing stud­ies have failed to show such trans­fer (Dahlin, Neely, Larsson, Back­man & Ny­berg, 2008; Holmes, Gath­er­cole, Place, Dun­ning, Hilton & El­liott, 2009; Thorell, Lindqvist, Bergman Nut­ley, Bohlin & Kling­berg, 2009). Thus, it is still un­clear un­der which con­di­tions effects of WM train­ing trans­fer to Gf.

Other in­ter­ven­tion stud­ies have in­cluded train­ing of at­ten­tion or ex­ec­u­tive func­tions. Rueda and col­leagues trained at­ten­tion in a sam­ple of 4- and 6-year-olds and found sig­nifi­cant gains in in­tel­li­gence (as mea­sured with the Kauf­man Brief In­tel­li­gence Test) in the 4-year-olds but only a ten­dency in the group of 6-year-olds (Rueda, Roth­bart, Mc­Can­dliss, Sac­co­manno & Pos­ner, 2005). A large train­ing study with 11,430 par­tic­i­pants re­vealed prac­ti­cally no trans­fer after a 6-week in­ter­ven­tion (10 min ⁄ day, 3 days a week) of a broader range of tasks in­clud­ing rea­son­ing and plan­ning or mem­o­ry, vi­suo-s­pa­tial skills, math­e­mat­ics and at­ten­tion (). How­ev­er, this study lacked con­trol in sam­ple se­lec­tion and com­pli­ance. In sum­ma­ry, it is still an open ques­tion to what ex­tent Gf can be im­proved by tar­geted train­ing.

Work­ing mem­ory train­ing in­clud­ing vari­ants on dual n-back has been shown to phys­i­cally change/in­crease the dis­tri­b­u­tion of white mat­ter in the brain59

Phys­i­cal changes have been linked to WM train­ing and n-back­ing. For ex­am­ple, Ole­sen PJ, West­er­berg H, Kling­berg T (2004) In­creased pre­frontal and pari­etal ac­tiv­ity after train­ing of work­ing mem­o­ry. Nat Neu­ro­science 7:75-79; about this study, Kuriyama writes:

“Ole­sen et al. (2004) pre­sented pro­gres­sive ev­i­dence ob­tained by func­tional mag­netic res­o­nance imag­ing that repet­i­tive train­ing im­proves spa­tial WM per­for­mance [both ac­cu­racy and re­sponse time (RT)] as­so­ci­ated with in­creased cor­ti­cal ac­tiv­ity in the mid­dle frontal gyrus and the su­pe­rior and in­fe­rior pari­etal cor­tices. Such a find­ing sug­gests that train­ing-in­duced im­prove­ment in WM per­for­mance could be based on neural plas­tic­i­ty, sim­i­lar to that for other skil­l-learn­ing char­ac­ter­is­tics.”

West­er­berg 2007, “Changes in cor­ti­cal ac­tiv­ity after train­ing of work­ing mem­o­ry—a sin­gle-sub­ject analy­sis”:

“…Prac­tice on the WM tasks grad­u­ally im­proved per­for­mance and this effect lasted sev­eral months. The effect of prac­tice also gen­er­al­ized to im­prove per­for­mance on a non-trained WM task and a rea­son­ing task. After train­ing, WM-re­lated brain ac­tiv­ity was sig­nifi­cantly in­creased in the mid­dle and in­fe­rior frontal gyrus. The changes in ac­tiv­ity were not due to ac­ti­va­tions of any ad­di­tional area that was not ac­ti­vated be­fore train­ing. In­stead, the changes could best be de­scribed by small in­creases in the ex­tent of the area of ac­ti­vated cor­tex. The effect of train­ing of WM is thus in sev­eral re­spects sim­i­lar to the changes in the func­tional map ob­served in pri­mate stud­ies of skill learn­ing, al­though the phys­i­o­log­i­cal effect in WM train­ing is lo­cated in the pre­frontal as­so­ci­a­tion cor­tex.”

Ex­ec­u­tive func­tions, in­clud­ing work­ing mem­ory and in­hi­bi­tion, are of cen­tral im­por­tance to much of hu­man be­hav­ior. In­ter­ven­tions in­tended to im­prove ex­ec­u­tive func­tions might there­fore serve an im­por­tant pur­pose. Pre­vi­ous stud­ies show that work­ing mem­ory can be im­proved by train­ing, but it is un­known if this also holds for in­hi­bi­tion, and whether it is pos­si­ble to train ex­ec­u­tive func­tions in preschool­ers. In the present study, preschool chil­dren re­ceived com­put­er­ized train­ing of ei­ther vi­suo-s­pa­tial work­ing mem­ory or in­hi­bi­tion for 5 weeks. An ac­tive con­trol group played com­mer­cially avail­able com­puter games, and a pas­sive con­trol group took part in only pre- and posttest­ing. Chil­dren trained on work­ing mem­ory im­proved sig­nifi­cantly on trained tasks; they showed train­ing effects on non-trained tests of spa­tial and ver­bal work­ing mem­o­ry, as well as trans­fer effects to at­ten­tion. Chil­dren trained on in­hi­bi­tion showed a sig­nifi­cant im­prove­ment over time on two out of three trained task par­a­digms, but no sig­nifi­cant im­prove­ments rel­a­tive to the con­trol groups on tasks mea­sur­ing work­ing mem­ory or at­ten­tion. In nei­ther of the two in­ter­ven­tions were there effects on non-trained in­hibitory tasks. The re­sults sug­gest that work­ing mem­ory train­ing can have sig­nifi­cant effects also among preschool chil­dren. The find­ing that in­hi­bi­tion could not be im­proved by ei­ther one of the two train­ing pro­grams might be due to the par­tic­u­lar train­ing pro­gram used in the present study or pos­si­bly in­di­cate that ex­ec­u­tive func­tions differ in how eas­ily they can be im­proved by train­ing, which in turn might re­late to differ­ences in their un­der­ly­ing psy­cho­log­i­cal and neural process­es.

A neural net­work un­der­ly­ing at­ten­tional con­trol in­volves the an­te­rior cin­gu­late in ad­di­tion to lat­eral pre­frontal ar­eas. An im­por­tant de­vel­op­ment of this net­work oc­curs be­tween 3 and 7 years of age. We have ex­am­ined the effi­ciency of at­ten­tional net­works across age and after 5 days of at­ten­tion train­ing (ex­per­i­men­tal group) com­pared with differ­ent types of no train­ing (con­trol groups) in 4-year-old and 6-year-old chil­dren. Strong im­prove­ment in ex­ec­u­tive at­ten­tion and in­tel­li­gence was found from ages 4 to 6 years. Both 4- and 6-year-olds showed more ma­ture per­for­mance after the train­ing than did the con­trol groups. This find­ing ap­plies to be­hav­ioral scores of the ex­ec­u­tive at­ten­tion net­work as mea­sured by the at­ten­tion net­work test, even­t-re­lated po­ten­tials recorded from the scalp dur­ing at­ten­tion net­work test per­for­mance, and in­tel­li­gence test scores. We also doc­u­mented the role of the tem­pera­men­tal fac­tor of effort­ful con­trol and the DAT1 gene in in­di­vid­ual differ­ences in at­ten­tion. Over­all, our data sug­gest that the ex­ec­u­tive at­ten­tion net­work ap­pears to de­velop un­der strong ge­netic con­trol, but that it is sub­ject to ed­u­ca­tional in­ter­ven­tions dur­ing de­vel­op­ment.

Be­hav­ioural find­ings in­di­cate that the core ex­ec­u­tive func­tions of in­hi­bi­tion and work­ing mem­ory are closely linked, and neu­roimag­ing stud­ies in­di­cate over­lap be­tween their neural cor­re­lates. There has not, how­ev­er, been a com­pre­hen­sive study, in­clud­ing sev­eral in­hi­bi­tion tasks and sev­eral work­ing mem­ory tasks, per­formed by the same sub­jects. In the present study, 11 healthy adult sub­jects com­pleted sep­a­rate blocks of 3 in­hi­bi­tion tasks (a stop task, a go/no-go task and a flanker task), and 2 work­ing mem­ory tasks (one spa­tial and one ver­bal). Ac­ti­va­tion com­mon to all 5 tasks was iden­ti­fied in the right in­fe­rior frontal gyrus, and, at a lower thresh­old, also the right mid­dle frontal gyrus and right pari­etal re­gions (BA 40 and BA 7). Left in­fe­rior frontal re­gions of in­ter­est (ROIs) showed a sig­nifi­cant con­junc­tion be­tween all tasks ex­cept the flanker task. The present study could not pin­point the spe­cific func­tion of each com­mon re­gion, but the pari­etal re­gion iden­ti­fied here has pre­vi­ously been con­sis­tently re­lated to work­ing mem­ory stor­age and the right in­fe­rior frontal gyrus has been as­so­ci­ated with in­hi­bi­tion in both le­sion and imag­ing stud­ies. These re­sults sup­port the no­tion that in­hibitory and work­ing mem­ory tasks in­volve com­mon neural com­po­nents, which may pro­vide a neural ba­sis for the in­ter­re­la­tion­ship be­tween the two sys­tems.

Re­cent func­tional neu­roimag­ing ev­i­dence sug­gests a bot­tle­neck be­tween learn­ing new in­for­ma­tion and re­mem­ber­ing old in­for­ma­tion. In two be­hav­ioral ex­per­i­ments and one func­tional MRI (fMRI) ex­per­i­ment, we tested the hy­poth­e­sis that learn­ing and re­mem­ber­ing com­pete when both processes hap­pen within a brief pe­riod of time. In the first be­hav­ioral ex­per­i­ment, par­tic­i­pants in­ten­tion­ally re­mem­bered old words dis­played in the fore­ground, while in­ci­den­tally learn­ing new scenes dis­played in the back­ground. In line with a mem­ory com­pe­ti­tion, we found that re­mem­ber­ing old in­for­ma­tion was as­so­ci­ated with im­paired learn­ing of new in­for­ma­tion. We repli­cated this find­ing in a sub­se­quent fMRI ex­per­i­ment, which showed that this be­hav­ioral effect was cou­pled with a sup­pres­sion of learn­ing-re­lated ac­tiv­ity in vi­sual and me­dial tem­po­ral ar­eas. More­over, the fMRI ex­per­i­ment pro­vided ev­i­dence that left mid-ven­tro­lat­eral pre­frontal cor­tex is in­volved in re­solv­ing the mem­ory com­pe­ti­tion, pos­si­bly by fa­cil­i­tat­ing rapid switch­ing be­tween learn­ing and re­mem­ber­ing. Crit­i­cal­ly, a fol­low-up be­hav­ioral ex­per­i­ment in which the back­ground scenes were re­placed with a vi­sual tar­get de­tec­tion task pro­vided in­di­ca­tions that the com­pe­ti­tion be­tween learn­ing and re­mem­ber­ing was not merely due to at­ten­tion. This study not only pro­vides novel in­sight into our ca­pac­ity to learn and re­mem­ber, but also clar­i­fies the neural mech­a­nisms un­der­ly­ing flex­i­ble be­hav­ior.

(There’s also a worth­while blog ar­ti­cle on this one: “Train­ing The Mind: Trans­fer Across Tasks Re­quir­ing In­ter­fer­ence Res­o­lu­tion”.)

“How dis­tractible are you? The an­swer may lie in your work­ing mem­ory ca­pac­ity”

  • Jen­nifer C. Mc­Vay, Michael J. Kane (2009). “Con­duct­ing the train of thought: Work­ing mem­ory ca­pac­i­ty, goal ne­glect, and mind wan­der­ing in an ex­ec­u­tive-con­trol task”. Jour­nal of Ex­per­i­men­tal Psy­chol­o­gy: Learn­ing, Mem­o­ry, and Cog­ni­tion, 35 (1), 196-204 DOI: 10.1037/a0014104:

On the ba­sis of the ex­ec­u­tive-at­ten­tion the­ory of work­ing mem­ory ca­pac­ity (WMC; e.g., M. J. Kane, A. R. A. Con­way, D. Z. Ham­brick, & R. W. En­gle, 2007), the au­thors tested the re­la­tions among WMC, mind wan­der­ing, and goal ne­glect in a sus­tained at­ten­tion to re­sponse task (SART; a go/no-go task). In 3 SART ver­sions, mak­ing con­cep­tual ver­sus per­cep­tual pro­cess­ing de­mands, sub­jects pe­ri­od­i­cally in­di­cated their thought con­tent when probed fol­low­ing rare no-go tar­gets. SART pro­cess­ing de­mands did not affect mind-wan­der­ing rates, but mind-wan­der­ing rates var­ied with WMC and pre­dicted goal-ne­glect er­rors in the task; fur­ther­more, mind-wan­der­ing rates par­tially me­di­ated the WMC-SART re­la­tion, in­di­cat­ing that WMC-related differ­ences in goal ne­glect were due, in part, to vari­a­tion in the con­trol of con­scious thought.

  • “Work­ing mem­ory ca­pac­ity and its re­la­tion to gen­eral in­tel­li­gence”; An­drew R.A. Con­way et al; TRENDS in Cog­ni­tive Sci­ences Vol.7 No.2003-12-12

    Sev­eral re­cent la­tent vari­able analy­ses sug­gest that (work­ing mem­ory ca­pac­i­ty) ac­counts for at least one-third and per­haps as much as one-half of the vari­ance in (in­tel­li­gence).What seems to be im­por­tant about WM span tasks is that they re­quire the ac­tive main­te­nance of in­for­ma­tion in the face of con­cur­rent pro­cess­ing and in­ter­fer­ence and there­fore re­cruit an ex­ec­u­tive at­ten­tion-con­trol mech­a­nism to com­bat in­ter­fer­ence. Fur­ther­more, this abil­ity seems to be me­di­ated by por­tions of the pre­frontal cor­tex.


Jaeggi 2005

“Ca­pac­ity Lim­i­ta­tions in Hu­man Cog­ni­tion: Be­hav­ioural and Bi­o­log­i­cal Con­tri­bu­tions”; Jaeggi the­sis:

…Ex­per­i­ment 6 and 7 fi­nally tackle the is­sue, whether ca­pac­ity lim­i­ta­tions are trait-like, i.e., fixed, or whether it is be pos­si­ble to ex­tend these lim­i­ta­tions with train­ing and whether gen­er­al­ized effects on other do­mains can be ob­served. In the last sec­tion, all the find­ings are in­te­grated and dis­cussed, and fur­ther is­sues re­main­ing to be in­ves­ti­gated are pointed out.

…In this ex­per­i­ment [6], the effects of a 10-day train­ing of an adap­tive ver­sion of an n-back dual task were stud­ied. The adap­tive ver­sion should be very di­rectly de­pend­ing on the ac­tual per­for­mance of the par­tic­i­pant: Not be­ing too easy, but also not too diffi­cult; al­ways pro­vid­ing a sense of achieve­ment in the par­tic­i­pant in or­der to keep the mo­ti­va­tion high. Com­par­ing pre and post mea­sures, effects on the task it­self were eval­u­at­ed, but also effects on other WM mea­sures, and on a mea­sure of fluid in­tel­li­gence.

…As stated be­fore, this study [7] was con­ducted in or­der to repli­cate and ex­tend the find­ings of Ex­per­i­ment 6: I was pri­mar­ily in­ter­ested to see whether an as­ymp­totic curve re­gard­ing per­for­mance would be reached after nearly twice of the train­ing ses­sions used in Ex­per­i­ment 6, and fur­ther, whether gen­er­al­ized and differ­en­tial effects on var­i­ous cog­ni­tive tasks could be ob­tained with this train­ing. There­fore, more tasks were in­cluded com­pared to Ex­per­i­ment 6, cov­er­ing many as­pects of WM (i.e., ver­bal tasks, vi­su­ospa­tial tasks), ex­ec­u­tive func­tions, as well as con­trol tasks not used in Ex­per­i­ment 6 in or­der to in­ves­ti­gate whether the WM train­ing has a se­lec­tive effect on tasks which are re­lated to the con­cept of WM and ex­ec­u­tive func­tions with no effect on these con­trol tasks. With re­spect to fluid in­tel­li­gence, a more ap­pro­pri­ate task than the APM, i.e., the ‘Bochumer Ma­trizen­test’ (BOMAT; Hossiep, Tur­ck, & Hasel­la, 1999) was used, which has the ad­van­tage that full par­al­lel-ver­sions are avail­able and that the task was ex­plic­itly de­vel­oped in or­der not to yield ceil­ing effects in stu­dent sam­ples. The ex­per­i­ment was car­ried out to­gether with Mar­tin Buschkuehl and Daniela Blaser; the lat­ter writ­ing her Mas­ter the­sis on the top­ic.

Jaeggi 2008

“Im­prov­ing fluid in­tel­li­gence with train­ing on work­ing mem­ory”, Jaeggi et al 2008 (sup­ple­ment; all the data in Jaeggi 2005 was used in this as well); this ar­ti­cle was widely cov­ered (eg. “Brain-Train­ing To Im­prove Mem­ory Boosts Fluid In­tel­li­gence” or ’s “For­get Brain Age: Re­searchers De­velop Soft­ware That Makes You Smarter”) and sparked most peo­ple’s in­ter­est in the top­ic. The ab­stract:

Fluid in­tel­li­gence (Gf) refers to the abil­ity to rea­son and to solve new prob­lems in­de­pen­dently of pre­vi­ously ac­quired knowl­edge. Gf is crit­i­cal for a wide va­ri­ety of cog­ni­tive tasks, and it is con­sid­ered one of the most im­por­tant fac­tors in learn­ing. More­over, Gf is closely re­lated to pro­fes­sional and ed­u­ca­tional suc­cess, es­pe­cially in com­plex and de­mand­ing en­vi­ron­ments. Al­though per­for­mance on tests of Gf can be im­proved through di­rect prac­tice on the tests them­selves, there is no ev­i­dence that train­ing on any other reg­i­men yields in­creased Gf in adults. Fur­ther­more, there is a long his­tory of re­search into cog­ni­tive train­ing show­ing that, al­though per­for­mance on trained tasks can in­crease dra­mat­i­cal­ly, trans­fer of this learn­ing to other tasks re­mains poor. Here, we present ev­i­dence for trans­fer from train­ing on a de­mand­ing work­ing mem­ory task to mea­sures of Gf. This trans­fer re­sults even though the trained task is en­tirely differ­ent from the in­tel­li­gence test it­self. Fur­ther­more, we demon­strate that the ex­tent of gain in in­tel­li­gence crit­i­cally de­pends on the amount of train­ing: the more train­ing, the more im­prove­ment in Gf. That is, the train­ing effect is dosage-de­pen­dent. Thus, in con­trast to many pre­vi­ous stud­ies, we con­clude that it is pos­si­ble to im­prove Gf with­out prac­tic­ing the test­ing tasks them­selves, open­ing a wide range of ap­pli­ca­tions.

Brain Work­shop in­cludes a spe­cial ‘Jaeggi mode’ which repli­cates al­most ex­actly the set­tings de­scribed for the “Brain Twister” soft­ware used in the study.

No study is de­fin­i­tive, of course, but Jaeggi 2008 is still one of the ma­jor stud­ies that must be cited in any DNB dis­cus­sion. There are some is­sues - not as many sub­jects as one would like, and the re­searchers (quoted in the Wired ar­ti­cle) ob­vi­ously don’t know if the WM or Gf gains are durable; more tech­ni­cal is­sues like the ad­min­is­tered Gf IQ tests be­ing speeded and thus pos­si­bly re­duced in va­lid­ity have been raised by Moody and oth­ers.

Qiu 2009

“Study on Im­prov­ing Fluid In­tel­li­gence through Cog­ni­tive Train­ing Sys­tem Based on Ga­bor Stim­u­lus”, 2009 First In­ter­na­tional Con­fer­ence on In­for­ma­tion Sci­ence and En­gi­neer­ing, ab­stract:

Gen­eral fluid in­tel­li­gence (Gf) is a hu­man abil­ity to rea­son and solve new prob­lems in­de­pen­dently of pre­vi­ously ac­quired knowl­edge and ex­pe­ri­ence. It is con­sid­ered one of the most im­por­tant fac­tors in learn­ing. One of the is­sues which aca­d­e­mic peo­ple con­cen­trates on is whether Gf of adults can be im­proved. Ac­cord­ing to the Dual N-back work­ing mem­ory the­ory and the char­ac­ter­is­tics of vi­sual per­cep­tual learn­ing, this pa­per put for­ward cog­ni­tive train­ing pat­tern based on Ga­bor stim­uli. A to­tal of 20 un­der­grad­u­ate stu­dents at 24 years old par­tic­i­pated in the ex­per­i­ment, with ten train­ing ses­sions for ten days. Through us­ing Raven’s Stan­dard Pro­gres­sive Ma­tri­ces as the eval­u­a­tion method to get and an­a­lyze the ex­per­i­men­tal re­sults, it was proved that train­ing pat­tern can im­prove fluid in­tel­li­gence of adults. This will pro­mote a wide range of ap­pli­ca­tions in the field of adult in­tel­lec­tual ed­u­ca­tion.

Dis­cus­sion and crit­i­cism of this Chi­nese6061 pa­per took place in 2 threads; the SPM was ad­min­is­ter in 25 min­utes, which while not as fast as Jaeggi 2008, is still not the nor­mal length. An ad­di­tional anom­aly is that ac­cord­ing to the fi­nal graph, the con­trol group’s IQ dropped mas­sively in the post-test (driv­ing much of the im­prove­men­t). As part of my , I tried to con­tact the 4 au­thors in May, June, July & Sep­tem­ber 2012; they even­tu­ally replied with da­ta.

polar (June 2009)

A group mem­ber, po­lar, con­ducted a small ex­per­i­ment at his uni­ver­sity where he was a stu­dent; his re­sults seemed to show an im­prove­ment. As po­lar would be the first to ad­mit, the at­tri­tion in sub­jects (few to be­gin with), rel­a­tively short time of train­ing and what­not make the power of his study weak.

Jaeggi 2010

“The re­la­tion­ship be­tween n-back per­for­mance and ma­trix rea­son­ing - im­pli­ca­tions for train­ing and trans­fer”, Jaeggi et al (coded as Jaeggi2 in meta-analy­sis); ab­stract:

…In the first study, we demon­strated that dual and sin­gle n-back task per­for­mances are ap­prox­i­mately equally cor­re­lated with per­for­mance on two differ­ent tasks mea­sur­ing Gf, whereas the cor­re­la­tion with a task as­sess­ing work­ing mem­ory ca­pac­ity was small­er. Based on these re­sults, the sec­ond study was aimed on test­ing the hy­poth­e­sis that train­ing on a sin­gle n-back task yields the same im­prove­ment in Gf as train­ing on a dual n-back task, but that there should be less trans­fer to work­ing mem­ory ca­pac­i­ty. We trained two groups of stu­dents for four weeks with ei­ther a sin­gle or a dual n-back in­ter­ven­tion. We in­ves­ti­gated trans­fer effects on work­ing mem­ory ca­pac­ity and Gf com­par­ing the two train­ing groups’ per­for­mance to con­trols who re­ceived no train­ing of any kind. Our re­sults showed that both train­ing groups im­proved more on Gf than con­trols, thereby repli­cat­ing and ex­tend­ing our prior re­sults.

The 2 stud­ies mea­sured Gf us­ing Raven’s APM and the BOMAT. In both stud­ies, the tests were ad­min­is­tered speeded to 10 or 15 min­utes as in Jaeggi 2008. The ex­per­i­men­tal groups saw av­er­age gains of 1 or 2 ad­di­tional cor­rect an­swers on the BOMAT and APM. It’s worth not­ing that the Sin­gle N-Back was done with a vi­sual modal­ity (and the DNB with the stan­dard vi­sual & au­dio).

Fol­lowup work:

  • Schnei­ders et al 2012 trained au­dio WM and found no trans­fer to vi­sual WM tasks; un­for­tu­nate­ly, they did not mea­sure any far trans­fer tasks like RAPM/BOMAT.
  • Beavon 2012 re­ports n = 47, ex­per­i­men­tals trained on sin­gle n-back & con­trols on “com­bined ver­bal tasks De­fine­time and Who wants to be a mil­lion­aire (Mil­lion­aire)”; no im­prove­ments on “STM span and at­ten­tion, short term au­di­tory mem­ory span and di­vided at­ten­tion, and WM as op­er­a­tionalised through the Wood­cock­-John­son III: Tests of cog­ni­tive abil­i­ties (WJ-III)”.

Studer-Luethi 2012

The sec­ond study’s data was reused for a Big Five per­son­al­ity fac­tor analy­sis in Stud­er-Luethi, Jaeg­gi, et al 2012, “In­flu­ence of neu­roti­cism and con­sci­en­tious­ness on work­ing mem­ory train­ing out­come”.62

The lack of n-back score cor­re­la­tion with WM score seems in line with an ear­lier study; “Work­ing Mem­o­ry, At­ten­tion Con­trol, and the N-Back Task: A Ques­tion of Con­struct Va­lid­ity”:

…Par­tic­i­pants also com­pleted a ver­bal WM span task (op­er­a­tion span task) and a marker test of gen­eral fluid in­tel­li­gence (Gf; Ravens Ad­vanced Pro­gres­sive Ma­tri­ces Test; J. C. Raven, J. E. Raven, & J. H. Court, 1998). N-back and WM span cor­re­lated weak­ly, sug­gest­ing they do not re­flect pri­mar­ily a sin­gle con­struct; more­over, both ac­counted for in­de­pen­dent vari­ance in Gf. N-back has face va­lid­ity as a WM task, but it does not demon­strate con­ver­gent va­lid­ity with at least 1 es­tab­lished WM mea­sure.

Stephenson 2010

“Does train­ing to in­crease work­ing mem­ory ca­pac­ity im­prove fluid in­tel­li­gence?”:

The cur­rent study was suc­cess­ful in repli­cat­ing Jaeggi et al.’s (2008) re­sults. How­ev­er, the cur­rent study also ob­served im­prove­ments in scores on the Raven’s Ad­vanced Pro­gres­sive Ma­tri­ces for par­tic­i­pants who com­pleted a vari­a­tion of the dual n-back task or a short­-term mem­ory task train­ing pro­gram. Par­tic­i­pants’ scores im­proved sig­nifi­cantly for only two of the four tests of GJ, which raises the is­sue of whether the tests mea­sure the con­struct Gf ex­clu­sive­ly, as de­fined by Cat­tell (1963), or whether they may be sen­si­tive to other fac­tors. The con­cern is whether the train­ing is ac­tu­ally im­prov­ing Gf or if the train­ing is im­prov­ing at­ten­tional con­trol and/or vi­su­ospa­tial skills, which im­proves per­for­mance on spe­cific tests of Gf. The find­ings are dis­cussed in terms of im­pli­ca­tions for con­cep­tu­al­iz­ing and as­sess­ing Gf.

136 par­tic­i­pants split over 25-28 sub­jects in ex­per­i­men­tal groups and the con­trol group. Vi­sual n-back im­proved more than au­dio n-back; the con­trol group was a pas­sive con­trol group (they did noth­ing but served as con­trols for test-retest effect­s).

Stephenson & Halpern 2013

“Im­proved ma­trix rea­son­ing is lim­ited to train­ing on tasks with a vi­su­ospa­tial com­po­nent”, Stephen­son & Halpern 2013:

Re­cent stud­ies (e.g., Jaeggi et al., 2008, 2010) have pro­vided ev­i­dence that scores on tests of fluid in­tel­li­gence can be im­proved by hav­ing par­tic­i­pants com­plete a four week train­ing pro­gram us­ing the dual n-back task. The dual n-back task is a work­ing mem­ory task that presents au­di­tory and vi­sual stim­uli si­mul­ta­ne­ous­ly. The pri­mary goal of our study was to de­ter­mine whether a vi­su­ospa­tial com­po­nent is re­quired in the train­ing pro­gram for par­tic­i­pants to ex­pe­ri­ence gains in tests of fluid in­tel­li­gence. We had par­tic­i­pants com­plete vari­a­tions of the dual n-back task or a short­-term mem­ory task as train­ing. Par­tic­i­pants were as­sessed with four tests of fluid in­tel­li­gence and four cog­ni­tive tests. We were suc­cess­ful in cor­rob­o­rat­ing Jaeggi et al.’s re­sults, how­ev­er, im­prove­ments in scores were ob­served on only two out of four tests of fluid in­tel­li­gence for par­tic­i­pants who com­pleted the dual n-back task, the vi­sual n-back task, or a short­-term mem­ory task train­ing pro­gram. Our re­sults raise the is­sue of whether the tests mea­sure the con­struct of fluid in­tel­li­gence ex­clu­sive­ly, or whether they may be sen­si­tive to other fac­tors. The find­ings are dis­cussed in terms of im­pli­ca­tions for con­cep­tu­al­iz­ing and as­sess­ing fluid in­tel­li­gence…The data in the cur­rent pa­per was part of Clay­ton Stephen­son’s doc­toral dis­ser­ta­tion.

Jaeggi 2011

Jaeg­gi, Buschkuehl, Jonides & Shah 2011 “Short- and long-term ben­e­fits of cog­ni­tive train­ing” (coded as Jaeggi3 in the meta-analy­sis); the ab­stract:

We trained el­e­men­tary and mid­dle school chil­dren by means of a videogame-like work­ing mem­ory task. We found that only chil­dren who con­sid­er­ably im­proved on the train­ing task showed a per­for­mance in­crease on un­trained fluid in­tel­li­gence tasks. This im­prove­ment was larger than the im­prove­ment of a con­trol group who trained on a knowl­edge-based task that did not en­gage work­ing mem­o­ry; fur­ther, this differ­en­tial pat­tern re­mained in­tact even after a 3-mo hia­tus from train­ing. We con­clude that cog­ni­tive train­ing can be effec­tive and long-last­ing, but that there are lim­it­ing fac­tors that must be con­sid­ered to eval­u­ate the effects of this train­ing, one of which is in­di­vid­ual differ­ences in train­ing per­for­mance. We pro­pose that fu­ture re­search should not in­ves­ti­gate whether cog­ni­tive train­ing works, but rather should de­ter­mine what train­ing reg­i­mens and what train­ing con­di­tions re­sult in the best trans­fer effects, in­ves­ti­gate the un­der­ly­ing neural and cog­ni­tive mech­a­nisms, and fi­nal­ly, in­ves­ti­gate for whom cog­ni­tive train­ing is most use­ful.

(This pa­per is not to be con­fused with the 2011 poster, “Work­ing Mem­ory Train­ing and Trans­fer to Gf: Ev­i­dence for Do­main Speci­fici­ty?”, Jaeggi et al 2011.)

It is worth not­ing that the study used Sin­gle N-back (vi­su­al). Un­like Jaeggi 2008, “de­spite the ex­per­i­men­tal group’s clear train­ing effect, we ob­served no sig­nifi­cant group × test ses­sion in­ter­ac­tion on trans­fer to the mea­sures of Gf” (so per­haps the train­ing was long enough for sub­jects to hit their ceil­ings). The group which did n-back could be split, based on fi­nal IQ & n-back scores, into 2 groups; in­ter­est­ingly “In­spec­tion of n-back train­ing per­for­mance re­vealed that there were no group differ­ences in the first 3 wk of train­ing; thus, it seems that group differ­ences emerge more clearly over time [first 3 wk: t(30) < 1; P = ns; last week: t(16) = 3.00; P < 0.01] (Fig. 3).” 3 weeks is ~21 days, or >19 days (the longest pe­riod in Jaeggi 2008). It’s also worth not­ing that Jaeggi 2011 seems to avoid Moody’s most co­gent crit­i­cism, the speed­ing of the IQ tests; from the pa­per’s “Ma­te­r­ial and Meth­ods” sec­tion;

We as­sessed ma­trix rea­son­ing with two differ­ent tasks, the Test of Non­ver­bal In­tel­li­gence (TONI) (23) and Raven’s Stan­dard Pro­gres­sive Ma­tri­ces (SPM) (24). Par­al­lel ver­sions were used for the pre, post-, and fol­low-up test ses­sions in coun­ter­bal­anced or­der. For the TONI, we used the stan­dard pro­ce­dure (45 items, five prac­tice items; un­timed), whereas for the SPM, we used a short­ened ver­sion (s­plit into odd and even items; 29 items per ver­sion; two prac­tice items; timed to 10 min after com­ple­tion of the prac­tice items. Note that vir­tu­ally all of the chil­dren com­pleted this task within the given time­frame).

The IQ re­sults were, specifi­cal­ly, the con­trol group av­er­aged 15.33/16.20 (be­fore/after) cor­rect an­swers on the SPM and 20.87/22.50 on the TONI; the n-back group av­er­aged 15.44/16.94 SPM and 20.41/22.03 TONI. 1.5 more right ques­tions rather than ~1 may not seem like much, but the split groups look quite differ­ent - the ‘small train­ing gain’ n-back­ing group ac­tu­ally fell on its sec­ond SPM and im­proved by <0.2 ques­tions on the TONI, while the ‘large train­ing gain’ in­creased >3 ques­tions on the SPM and TONI. The differ­ence is not so dra­matic in the fol­lowup 3 months lat­er: the small group is now 17.43/23.43 (SPM/TONI), and the large group 15.67/24.67. Strangely in the fol­lowup, the con­trol group has a higher SPM than the large group (but not the small group), and a higher TONI than ei­ther group; the con­trol group has higher IQ scores on both TONI & SPM in the fol­lowup than the ag­gre­gate n-back group. (The split­ting of groups is also un­ortho­dox63.)

UoM pro­duced a video with Jonides; Jaeggi 2011 has also been dis­cussed in main­stream me­dia. From the Wall Street Jour­nal’s “Boot Camp for Boost­ing IQ”:

…when sev­eral dozen el­e­men­tary- and mid­dle-school kids from the De­troit area used this ex­er­cise for 15 min­utes a day, many showed sig­nifi­cant gains on a widely used in­tel­li­gence test. Most im­pres­sive, per­haps, is that these gains per­sisted for three months, even though the chil­dren had stopped train­ing…these school­child­ren showed gains in fluid in­tel­li­gence roughly equal to five IQ points after one month of train­ing…There are two im­por­tant caveats to this re­search. The first is that not every kid showed such dra­matic im­prove­ments after train­ing. Ini­tial ev­i­dence sug­gests that chil­dren who failed to in­crease their fluid in­tel­li­gence found the ex­er­cise too diffi­cult or bor­ing and thus did­n’t fully en­gage with the train­ing.

From Dis­cover’s blogs, “Can in­tel­li­gence be boosted by a sim­ple task? For some…”, come ad­di­tional de­tails:

She [Jaeg­gi] re­cruited 62 chil­dren, aged be­tween seven and ten. While half of them sim­ply learned some ba­sic gen­eral knowl­edge ques­tions, the other half trained with a cheer­ful com­put­erised n-back task. They saw a stream of im­ages where a tar­get ob­ject ap­peared in one of six lo­ca­tions - say, a frog in a lily pond. They had to press a but­ton if the frog was in the same place as it was two im­ages ago, forc­ing them to store a con­tin­u­ously up­dated stream of im­ages in their minds. If the chil­dren got bet­ter at the task, this gap in­creased so they had to keep more im­ages in their heads. If they strug­gled, the gap was short­ened.

Be­fore and after the train­ing ses­sions, all the chil­dren did two rea­son­ing tests de­signed to mea­sure their fluid in­tel­li­gence. At first, the re­sults looked dis­ap­point­ing. On av­er­age, the n-back chil­dren did­n’t be­come any bet­ter at these tests than their peers who stud­ied the knowl­edge ques­tions. But ac­cord­ing to Jaeg­gi, that’s be­cause some of them did­n’t take to the train­ing. When she di­vided the chil­dren ac­cord­ing to how much they im­proved at the n-back task, she saw that those who showed the most progress also im­proved in fluid in­tel­li­gence. The oth­ers did not. Best of all, these ben­e­fits lasted for 3 months after the train­ing. That’s a first for this type of study, al­though Jaeggi her­self says that the effect is “not ro­bust.” Over this time pe­ri­od, all the chil­dren showed im­prove­ments in their fluid in­tel­li­gence, “prob­a­bly [as] a re­sult of the nat­ural course of de­vel­op­ment”.

…Philip Ack­er­man, who stud­ies learn­ing and brain train­ing at the Uni­ver­sity of Illi­nois, says, “I am con­cerned about the small sam­ple, es­pe­cially after split­ting the groups on the ba­sis of their per­for­mance im­prove­ments.” He has a point - the group that showed big im­prove­ments in the n-back train­ing only in­cluded 18 chil­dren….Why did some of the chil­dren ben­e­fit from the train­ing while oth­ers did not? Per­haps they were sim­ply un­in­ter­ested in the task, no mat­ter how colour­fully it was dressed up with storks and vam­pires. In Jaeg­gi’s ear­lier study with adults, every vol­un­teer signed up them­selves and were “in­trin­si­cally mo­ti­vated to par­tic­i­pate and train.” By con­trast, the kids in this lat­est study were signed up by their par­ents and teach­ers, and some might only have con­tin­ued be­cause they were told to do so.

It’s also pos­si­ble that the chang­ing diffi­culty of the game was frus­trat­ing for some of the chil­dren. Jaeggi says, “The chil­dren who did not ben­e­fit from the train­ing found the work­ing mem­ory in­ter­ven­tion too effort­ful and diffi­cult, were eas­ily frus­trat­ed, and be­came dis­en­gaged. This makes sense when you think of phys­i­cal train­ing - if you don’t try and re­ally run and just walk in­stead, you won’t im­prove your car­dio­vas­cu­lar fit­ness.” In­deed, a re­cent study on IQ test­ing which found that they re­flect mo­ti­va­tion as well as in­tel­li­gence.

Schweizer et al 2011


This study in­ves­ti­gated whether brain-train­ing (work­ing mem­ory [WM] train­ing) im­proves cog­ni­tive func­tions be­yond the train­ing task (trans­fer effect­s), es­pe­cially re­gard­ing the con­trol of emo­tional ma­te­r­ial since it con­sti­tutes much of the in­for­ma­tion we process dai­ly. Forty-five par­tic­i­pants re­ceived WM train­ing us­ing ei­ther emo­tional or neu­tral ma­te­ri­al, or an un­de­mand­ing con­trol task. WM train­ing, re­gard­less of train­ing ma­te­ri­al, led to trans­fer gains on an­other WM task and in fluid in­tel­li­gence. How­ev­er, only brain-train­ing with emo­tional ma­te­r­ial yielded trans­fer­able gains to im­proved con­trol over affec­tive in­for­ma­tion on an emo­tional Stroop task. The data sup­port the re­al­ity of trans­fer­able ben­e­fits of de­mand­ing WM train­ing and sug­gest that trans­fer­able gains across to affec­tive con­texts re­quire train­ing with ma­te­r­ial con­gru­ent to those con­texts. These find­ings con­sti­tute pre­lim­i­nary ev­i­dence that in­ten­sive cog­ni­tively de­mand­ing brain-train­ing can im­prove not only our ab­stract prob­lem-solv­ing ca­pac­i­ty, but also ame­lio­rate cog­ni­tive con­trol processes (e.g. de­ci­sion-mak­ing) in our daily emo­tive en­vi­ron­ments.


  1. There seems to be an IQ in­crease of around one ques­tion on the RPM (but there’s an odd­ity with the con­trol group which they think they cor­rect for64)
  2. The RPM does not seem to have been ad­min­is­tered speeded65
  3. The emo­tional as­pect seems to be just re­plac­ing the ‘neu­tral’ ex­ist­ing stim­uli like col­ors or let­ters or pi­ano keys with more loaded ones66, nor does this tweak seem to change the DNB/WM/IQ scores of that group67

Their later study “Train­ing the Emo­tional Brain: Im­prov­ing Affec­tive Con­trol through Emo­tional Work­ing Mem­ory Train­ing” did not use any mea­sure of fluid in­tel­li­gence.

Kundu et al 2011

“Re­lat­ing in­di­vid­ual differ­ences in short­-term mem­o­ry-derived EEG to cog­ni­tive train­ing effects” (coded as Kundu1 in the meta-analy­sis); 3 con­trols (Tetris), 3 ex­per­i­men­tals (Brain Work­shop) for 1000 min­utes. RAPM showed a slight in­crease. Ex­tremely small ex­per­i­men­tal size, which may form part of the data for Kundu et al 2012.

Zhong 2011

“The Effect Of Train­ing Work­ing Mem­ory And At­ten­tion On Pupils’ Fluid In­tel­li­gence” (ab­stract), Zhong 2011; orig­i­nal en­crypted file (8M), screen­shots of all pages in the­sis (20M); dis­cus­sion

Ap­pears to have found IQ gains, but no dose-re­sponse effect, us­ing a no-con­tact con­trol group. Diffi­cult to un­der­stand: trans­la­tion as­sis­tance from Chi­nese speak­ers would be ap­pre­ci­at­ed.

Jausovec 2012

“Work­ing mem­ory train­ing: Im­prov­ing in­tel­li­gence - Chang­ing brain ac­tiv­ity”, Jaušovec 2012:

The main ob­jec­tives of the study were: to in­ves­ti­gate whether train­ing on work­ing mem­ory (WM) could im­prove fluid in­tel­li­gence, and to in­ves­ti­gate the effects WM train­ing had on neu­ro­elec­tric (elec­troen­cephalog­ra­phy - EEG) and he­mo­dy­namic (n­ear-in­frared spec­troscopy - NIRS) pat­terns of brain ac­tiv­i­ty. In a par­al­lel group ex­per­i­men­tal de­sign, re­spon­dents of the work­ing mem­ory group after 30 h of train­ing sig­nifi­cantly in­creased per­for­mance on all tests of fluid in­tel­li­gence. By con­trast, re­spon­dents of the ac­tive con­trol group (par­tic­i­pat­ing in a 30-h com­mu­ni­ca­tion train­ing course) showed no im­prove­ments in per­for­mance. The in­flu­ence of WM train­ing on pat­terns of neu­ro­elec­tric brain ac­tiv­ity was most pro­nounced in the theta and al­pha bands. Theta and low­er-1 al­pha band syn­chro­niza­tion was ac­com­pa­nied by in­creased low­er-2 and up­per al­pha de­syn­chro­niza­tion. The he­mo­dy­namic pat­terns of brain ac­tiv­ity after the train­ing changed from higher right hemi­spheric ac­ti­va­tion to a bal­anced ac­tiv­ity of both frontal ar­eas. The neu­ro­elec­tric as well as he­mo­dy­namic pat­terns of brain ac­tiv­ity sug­gest that the train­ing in­flu­enced WM main­te­nance func­tions as well as processes di­rected by the cen­tral ex­ec­u­tive. The changes in up­per al­pha band de­syn­chro­niza­tion could fur­ther in­di­cate that processes re­lated to long term mem­ory were also in­flu­enced.

14 ex­per­i­men­tal & 15 con­trols; the test­ing was a lit­tle un­usu­al:

Re­spon­dents solved four test-bat­ter­ies, for which the pro­ce­dure was the same dur­ing pre- and post-test­ing. The same test-bat­ter­ies were used on pre- and post-test­ing. The digit span sub­test (WAIS-R) was ad­min­is­tered sep­a­rate­ly, ac­cord­ing to the di­rec­tions in the test man­ual (Wech­sler, 1981). The other three tests (RAPM, ver­bal analo­gies and spa­tial ro­ta­tion) were ad­min­is­tered while the re­spon­dents’ EEG and NIRS mea­sures were record­ed.

The RAPM was based on a mod­i­fied ver­sion of Raven’s pro­gres­sive ma­tri­ces (Raven, 1990), a widely used and well es­tab­lished test of fluid in­tel­li­gence (Stern­berg, Fer­rari, Clinken­beard, & Grig­orenko, 1996). The cor­re­la­tion be­tween this mod­i­fied ver­sion of RAPM and WAIS-R was r = .56, (p < .05, n = 97). Sim­i­lar cor­re­la­tions of the or­der of 0.40-0.75, were also re­ported for the stan­dard ver­sion of RAPM (Court & Raven, 1995). There­fore it can be con­cluded that the mod­i­fied ap­pli­ca­tion of the RAPM did not sig­nifi­cantly al­ter its met­ric char­ac­ter­is­tics. Used were 50 test items - 25 easy (Ad­vanced Pro­gres­sive Ma­tri­ces Set I - 12 items and the B Set of the Col­ored Pro­gres­sive Ma­tri­ces), and 25 diffi­cult items (Ad­vanced Pro­gres­sive Ma­tri­ces Set II, items 12-36). Par­tic­i­pants saw a fig­ural ma­trix with the lower right en­try miss­ing. They had to de­ter­mine which of the four op­tions fit­ted into the miss­ing space. The tasks were pre­sented on a com­puter screen (po­si­tioned about 80-100 cm in front of the re­spon­den­t), at fixed 10 or 14 s in­ter­stim­u­lus in­ter­vals. They were ex­posed for 6 s (easy) or 10 s (d­iffi­cult) fol­low­ing a 2-s in­ter­val, when a cross was pre­sent­ed. Dur­ing this time the par­tic­i­pants were in­structed to press a but­ton on a re­sponse pad (1-4) which in­di­cated their an­swer.

At 25 hard ques­tions, and <14s a ques­tion, that im­plies the RAPM was ad­min­is­tered in <5.8 min­utes. They com­ment:

To fur­ther in­ves­ti­gate pos­si­ble in­flu­ences of task diffi­culty on the ob­served per­for­mance gains on the RAPM a GLM for re­peated mea­sures test/retest  easy/d­iffi­cult-items  group (WM, AC) was con­duct­ed. The analy­sis showed only a sig­nifi­cant in­ter­ac­tion effect for the test/retest con­di­tion and type of train­ing used in the two groups (F(1, 27) = 4.47; p < .05; par­tial eta2 = .15). A GLM con­ducted for the WM group showed only a sig­nifi­cant test/retest effect (F(1, 13) = 30.11; p < .05; par­tial eta2 = .70), but no in­ter­ac­tion be­tween the test/retest con­di­tions and the diffi­culty level (F(1, 13) = 1.79; p = .17 not-sig­nifi­cant; par­tial eta2 = .12). As can be seen in Fig. 4 after WM train­ing an about equal in­crease in re­spon­dents’ per­for­mance for the easy and diffi­cult test items was ob­served. On the other hand, no in­creases in per­for­mance, nei­ther for the easy nor for the diffi­cult test items, in re­spon­dents of the ac­tive con­trol group were ob­served (F(1, 14) = .47; p = .50 not- sig­nifi­cant; par­tial eta2 = .03).

(Even on the “easy” ques­tions, no group per­formed bet­ter than 76% ac­cu­ra­cy.)

Clouter 2013

“The Effects of Dual n-back Train­ing on the Com­po­nents of Work­ing Mem­ory and Fluid In­tel­li­gence: An In­di­vid­ual Differ­ences Ap­proach”, Clouter 2013:

A num­ber of re­cent stud­ies have pro­vided ev­i­dence that train­ing work­ing mem­ory can lead to im­prove­ments in fluid in­tel­li­gence, work­ing mem­ory span, and per­for­mance on other un­trained tasks. How­ev­er, in ad­di­tion to a num­ber of mixed re­sults, many of these stud­ies suffer from de­sign lim­i­ta­tions. The aim of the present study was to ex­per­i­men­tally in­ves­ti­gate the effects of a dual n-back work­ing mem­ory train­ing task on a va­ri­ety of mea­sures of fluid in­tel­li­gence, rea­son­ing, work­ing mem­ory span, and at­ten­tional con­trol. The present study com­pared a train­ing group with an ac­tive con­trol group (a placebo group), us­ing ap­pro­pri­ate meth­ods that over­came the lim­i­ta­tions of pre­vi­ous stud­ies. The dual n-back train­ing group im­proved more than the ac­tive con­trol group on some, but not all out­come mea­sures. Differ­en­tial im­prove­ment for the train­ing group was ob­served on fluid in­tel­li­gence, work­ing mem­ory ca­pac­i­ty, and re­sponse times on con­flict tri­als in the Stroop task. In ad­di­tion, in­di­vid­ual differ­ences in pre-train­ing fluid in­tel­li­gence scores and ini­tial per­for­mance on the train­ing task ex­plain some of the vari­ance in out­come mea­sure im­prove­ments. We dis­cuss these re­sults in the con­text of pre­vi­ous stud­ies, and sug­gest that ad­di­tional work is needed in or­der to fur­ther un­der­stand the vari­ables re­spon­si­ble for trans­fer from train­ing.

Jaeggi et al 2013

“The role of in­di­vid­ual differ­ences in cog­ni­tive train­ing and trans­fer”:

Work­ing mem­ory (WM) train­ing has re­cently be­come a topic of in­tense in­ter­est and con­tro­ver­sy. Al­though sev­eral re­cent stud­ies have re­ported near- and far-trans­fer effects as a re­sult of train­ing WM-re­lated skills, oth­ers have failed to show far trans­fer, sug­gest­ing that gen­er­al­iza­tion effects are elu­sive. Al­so, many of the ear­lier in­ter­ven­tion at­tempts have been crit­i­cized on method­olog­i­cal grounds. The present study re­solves some of the method­olog­i­cal lim­i­ta­tions of pre­vi­ous stud­ies and also con­sid­ers in­di­vid­ual differ­ences as po­ten­tial ex­pla­na­tions for the differ­ing trans­fer effects across stud­ies. We re­cruited in­trin­si­cally mo­ti­vated par­tic­i­pants and as­sessed their need for cog­ni­tion (NFC; Ca­cioppo & Petty Jour­nal of Per­son­al­ity and So­cial Psy­chol­ogy 42:116-131, 1982) and their im­plicit the­o­ries of in­tel­li­gence (D­weck, 1999) prior to train­ing. We as­sessed the effi­cacy of two in­ter­ven­tions by com­par­ing par­tic­i­pants’ im­prove­ments on a bat­tery of fluid in­tel­li­gence tests against those of an ac­tive con­trol group. We ob­served that trans­fer to a com­pos­ite mea­sure of fluid rea­son­ing re­sulted from both WM in­ter­ven­tions. In ad­di­tion, we un­cov­ered fac­tors that con­tributed to train­ing suc­cess, in­clud­ing mo­ti­va­tion, need for cog­ni­tion, pre­ex­ist­ing abil­i­ty, and im­plicit the­o­ries about in­tel­li­gence.

This is quite a com­plex study, with a lot of analy­sis I don’t think I en­tirely un­der­stand. The quick sum­mary is ta­ble 2 on pg10: the DNB group fell on APM, rose on BOMAT (nei­ther sta­tis­ti­cal­ly-sig­nifi­can­t); the SNB group in­creased on APM & BOMAT (but only BOMAT was sta­tis­ti­cal­ly-sig­nifi­can­t).

Michael J. Kane has writ­ten some crit­i­cal com­ments on the re­sults.

Savage 2013

“Near and Far Trans­fer of Work­ing Mem­ory Train­ing Re­lated Gains in Healthy Adults”, Sav­age 2013:

En­hanc­ing in­tel­li­gence through work­ing mem­ory train­ing is an at­trac­tive con­cept, par­tic­u­larly for mid­dle-aged adults. How­ev­er, in­ves­ti­ga­tions of work­ing mem­ory train­ing ben­e­fits are lim­ited to younger or older adults, and re­sults are in­con­sis­tent. This study in­ves­ti­gates work­ing mem­ory train­ing in mid­dle age-range adults. Fifty healthy adults, aged 30-60, com­pleted mea­sures of work­ing mem­o­ry, pro­cess­ing speed, and fluid in­tel­li­gence be­fore and after a 5-week we­b-based work­ing mem­ory (ex­per­i­men­tal) or pro­cess­ing speed (ac­tive con­trol) train­ing pro­gram. Base­line in­tel­li­gence and per­son­al­ity were mea­sured as po­ten­tial in­di­vid­ual char­ac­ter­is­tics as­so­ci­ated with change. Im­proved per­for­mance on work­ing mem­ory and pro­cess­ing speed tasks were ex­pe­ri­enced by both groups; how­ev­er, only the work­ing mem­ory train­ing group im­proved in fluid in­tel­li­gence. Agree­able­ness emerged as a per­son­al­ity fac­tor as­so­ci­ated with work­ing mem­ory train­ing re­lated change. Al­beit lim­ited by pow­er, find­ings sug­gest that dual n-back work­ing mem­ory train­ing not only en­hances work­ing mem­ory but also fluid in­tel­li­gence in mid­dle-aged healthy adults.

The per­son­al­ity cor­re­la­tions seem to differ with Stud­er-Luethi.

Stepankova et al 2013

“The Mal­leabil­ity of Work­ing Mem­ory and Vi­su­ospa­tial Skills: A Ran­dom­ized Con­trolled Study in Older Adults”, Stepankova et al 2013:

There is ac­cu­mu­lat­ing ev­i­dence that train­ing on work­ing mem­ory (WM) gen­er­al­izes to other non­trained do­mains, and there are re­ports of trans­fer effects ex­tend­ing as far as to mea­sures of fluid in­tel­li­gence. Al­though there have been sev­eral demon­stra­tions of such trans­fer effects in young adults and chil­dren, they have been diffi­cult to demon­strate in older adults. In this study, we in­ves­ti­gated the gen­er­al­iz­ing effects of an adap­tive WM in­ter­ven­tion on non­trained mea­sures of WM and vi­su­ospa­tial skills. We ran­domly as­signed healthy older adults to train on a ver­bal n-back task over the course of a month for ei­ther 10 or 20 ses­sions. Their per­for­mance change was com­pared with that of a con­trol group. Our re­sults re­vealed re­li­able group effects in non­trained stan­dard clin­i­cal mea­sures of WM and vi­su­ospa­tial skills in that both train­ing groups out­per­formed the con­trol group. We also ob­served a dose-re­sponse effect, that is, a pos­i­tive re­la­tion­ship be­tween train­ing fre­quency and the gain in vi­su­ospa­tial skills; this find­ing was fur­ther con­firmed by a pos­i­tive cor­re­la­tion be­tween train­ing im­prove­ment and trans­fer. The im­prove­ments in vi­su­ospa­tial skills emerged even though the in­ter­ven­tion was re­stricted to the ver­bal do­main. Our work has im­por­tant im­pli­ca­tions in that our data pro­vide fur­ther ev­i­dence for plas­tic­ity of cog­ni­tive func­tions in old age.

Horvat 2014

“The effect of work­ing mem­ory train­ing on cog­ni­tive abil­i­ties”, Hor­vat 2014; Slove­ni­an, Eng­lish ab­stract:

In the last few years, there is a grow­ing ev­i­dence in psy­cho­log­i­cal lit­er­a­ture in­di­cat­ing that work­ing mem­ory train­ing could serve as a use­ful tool to im­prove per­for­mance on non-trained tasks that mea­sure higher cog­ni­tive abil­i­ties; how­ev­er, re­sults of differ­ent stud­ies re­main in­con­sis­tent. The aim of the present mas­ter the­sis was to dis­cover whether work­ing mem­ory train­ing could im­prove short­-term mem­ory ca­pac­ity and in­crease test scores on test of fluid in­tel­li­gence in nor­mal de­vel­op­ing chil­dren.

Fi­nal sam­ple con­sisted of 29 par­tic­i­pants, be­tween 13 to 15 years old; 14 of them were in ex­per­i­men­tal group, 15 was con­trols. Ex­per­i­men­tal group com­pleted se­ries of ten work­ing mem­ory train­ings, based on adap­tive dual n-back task. Con­trol group was pas­sive and did not do any train­ing in the mean­time.

Re­sults of our study showed that all par­tic­i­pants in ex­per­i­men­tal group im­proved their per­for­mance on trained task. There was no sta­tis­ti­cally sig­nifi­cant effect of ex­per­i­men­tal group on mea­sures of digit span and vi­su­ospa­tial mem­ory span be­fore and after train­ing, when com­par­ing with per­for­mance of con­trol group. How­ev­er, ex­per­i­men­tal group im­proved more on mea­sure of fluid in­tel­li­gence com­pared with con­trol group.

Find­ing of our study sug­gest the im­por­tance of in­ves­ti­gat­ing fac­tors as­so­ci­ated with effec­tive­ness of work­ing mem­ory train­ing in fu­ture re­search.

Heinzel et al 2016

“Neural cor­re­lates of train­ing and trans­fer effects in work­ing mem­ory in older adults”, Heinzel et al 2016:

As in­di­cated by pre­vi­ous re­search, ag­ing is as­so­ci­ated with a de­cline in work­ing mem­ory (WM) func­tion­ing, re­lated to al­ter­ations in fron­to-pari­etal neural ac­ti­va­tions. At the same time, pre­vi­ous stud­ies showed that WM train­ing in older adults may im­prove the per­for­mance in the trained task (train­ing effec­t), and more im­por­tant­ly, also in un­trained WM tasks (trans­fer effect­s). How­ev­er, neural cor­re­lates of these trans­fer effects that would im­prove un­der­stand­ing of its un­der­ly­ing mech­a­nisms, have not been shown in older par­tic­i­pants as yet. In this study, we in­ves­ti­gated blood­-oxy­gen-level-de­pen­dent (BOLD) sig­nal changes dur­ing n-back per­for­mance and an un­trained de­layed recog­ni­tion (Stern­berg) task fol­low­ing 12 ses­sions (45 min­utes each) of adap­tive n-back train­ing in older adults. The Stern­berg task used in this study al­lowed to test for neural train­ing effects in­de­pen­dent of spe­cific task affor­dances of the trained task and to sep­a­rate main­te­nance from up­dat­ing process­es. Thir­ty-two healthy older par­tic­i­pants (60-75 years) were as­signed ei­ther to an n-back train­ing or a no-con­tact con­trol group. Be­fore (t1) and after (t2) train­ing/wait­ing pe­ri­od, both the n-back task and the Stern­berg task were con­ducted while BOLD sig­nal was mea­sured us­ing func­tional Mag­netic Res­o­nance Imag­ing (fMRI) in all par­tic­i­pants. In ad­di­tion, neu­ropsy­cho­log­i­cal tests were per­formed out­side the scan­ner. WM per­for­mance im­proved with train­ing and be­hav­ioral trans­fer to tests mea­sur­ing ex­ec­u­tive func­tions, pro­cess­ing speed, and fluid in­tel­li­gence was found. In the train­ing group, BOLD sig­nal in right lat­eral mid­dle frontal gyrus/ cau­dal su­pe­rior frontal sul­cus (Brod­mann area, BA 6/8) de­creased in both the trained n-back and the up­dat­ing con­di­tion of the un­trained Stern­berg task at t2, com­pared to the con­trol group. FMRI find­ings in­di­cate a train­ing-re­lated in­crease in pro­cess­ing effi­ciency of WM net­works, po­ten­tially re­lated to the process of WM up­dat­ing. Per­for­mance gains in un­trained tasks sug­gest that trans­fer to other cog­ni­tive tasks re­mains pos­si­ble in ag­ing.


Moody 2009 (re: Jaeggi 2008)

Jaeggi 2008, you may re­mem­ber, showed that train­ing on N-back im­proved work­ing mem­o­ry, but it also boosted scores on tests of Gf. The lat­ter would be a ma­jor re­sult - in­deed, unique - and is one of the main re­search re­sults en­cour­ag­ing peo­ple to do N-back in a non-re­search set­ting. Peo­ple want to be­lieve that N-back is effi­ca­cious and par­tic­u­larly that it will do more than boost work­ing mem­o­ry. So we need to be wary of (for those of you who read too much fan­ta­sy, you’ll know this as ).

For­tu­nate­ly, we can dis­cuss at length the work of one David E. Moody who has pub­lished a crit­i­cism of how the odd method­ol­ogy of Jaeggi 2008 un­der­mines this re­sult. He’s worth quot­ing at length, since be­sides be­ing im­por­tant to un­der­stand­ing Jaeg­gi’s study, it’s an in­ter­est­ing ex­am­ple of how sub­tle is­sues can be im­por­tant in psy­chol­o­gy:

"The sub­jects were di­vided into four groups, differ­ing in the num­ber of days of train­ing they re­ceived on the task of work­ing mem­o­ry. The group that re­ceived the least train­ing (8 days) was tested on Raven’s Ad­vanced Pro­gres­sive Ma­tri­ces (Raven, 1990), a widely used and well-estab­lished test of fluid in­tel­li­gence. This group, how­ev­er, demon­strated neg­li­gi­ble im­prove­ment be­tween pre- and post-test per­for­mance.

The other three groups were not tested us­ing Raven’s Ma­tri­ces, but rather on an al­ter­na­tive test of much more re­cent ori­gin. The Bochumer Ma­tri­ces Test (BOMAT) (Hossiep, Tur­ck, & Hasel­la, 1999) is sim­i­lar to Raven’s in that it con­sists of vi­sual analo­gies. In both tests, a se­ries of geo­met­ric and other fig­ures is pre­sented in a ma­trix for­mat and the sub­ject is re­quired to in­fer a pat­tern in or­der to pre­dict the next fig­ure in the se­ries. The au­thors pro­vide no rea­son for switch­ing from Raven’s to the BOMAT.

The BOMAT differs from Raven’s in some im­por­tant re­spects, but is sim­i­lar in one cru­cial at­trib­ute: both tests are pro­gres­sive in na­ture, which means that test items are se­quen­tially arranged in or­der of in­creas­ing diffi­cul­ty. A high score on the test, there­fore, is pred­i­cated on sub­jects’ abil­ity to solve the more diffi­cult items.

How­ev­er, this pro­gres­sive fea­ture of the test was effec­tively elim­i­nated by the man­ner in which Jaeggi et al. ad­min­is­tered it. The BOMAT is a 29-item test which sub­jects are sup­posed to be al­lowed 45 min to com­plete. Re­mark­ably, how­ev­er, Jaeggi et al. re­duced the al­lot­ted time from 45 min to 10. The effect of this re­stric­tion was to make it im­pos­si­ble for sub­jects to pro­ceed to the more diffi­cult items on the test. The large ma­jor­ity of the sub­ject­s-re­gard­less of the num­ber of days of train­ing they re­ceived-an­swered less than 14 test items cor­rect­ly.
By virtue of the man­ner in which they ad­min­is­tered the BOMAT, Jaeggi et al. trans­formed it from a test of fluid in­tel­li­gence into a speed test of abil­ity to solve the eas­ier vi­sual analo­gies. The time re­stric­tion not only made it im­pos­si­ble for sub­jects to pro­ceed to the more diffi­cult items, it also lim­ited the op­por­tu­nity to learn about the test-and so im­prove per­for­mance-in the process of tak­ing it. This fac­tor can­not be ne­glected be­cause test per­for­mance does im­prove with prac­tice, as demon­strated by the con­trol groups in the Jaeggi study, whose im­prove­ment from pre- to post-test was about half that of the ex­per­i­men­tal groups. The same learn­ing process that oc­curs from one ad­min­is­tra­tion of the test to the next may also op­er­ate within a given ad­min­is­tra­tion of the test-pro­vided sub­jects are al­lowed suffi­cient time to com­plete it.

Since the whole weight of their con­clu­sion rests upon the va­lid­ity of their mea­sure of fluid in­tel­li­gence, one might as­sume the au­thors would present a care­ful de­fense of the man­ner in which they ad­min­is­tered the BOMAT. In­stead they do not even men­tion that sub­jects are nor­mally al­lowed 45 min to com­plete the test. Nor do they men­tion that the test has 29 items, of which most of their sub­jects com­pleted less than half.

The au­thors’ en­tire ra­tio­nale for re­duc­ing the al­lot­ted time to 10 min is con­fined to a foot­note. That foot­note reads as fol­lows:

Al­though this pro­ce­dure differs from the stan­dard­ized pro­ce­dure, there is ev­i­dence that this timed pro­ce­dure has lit­tle in­flu­ence on rel­a­tive stand­ing in these tests, in that the cor­re­la­tion of speeded and non-speeded ver­sions is very high (r = 0.95; ref. 37).

The ref­er­ence given in the foot­note is to a 1988 study (F­rear­son & Eysenck, 1986) that is not in fact de­signed to sup­port the con­clu­sion stated by Jaeggi et al. The 1988 study merely con­tains a foot­note of its own, which refers in turn to un­pub­lished re­search con­ducted forty years ear­li­er. That re­search in­volved Raven’s ma­tri­ces, not the BOMAT, and en­tailed a re­duc­tion in time of at most 50%, not more than 75%, as in the Jaeggi study.

So in­stead of offer­ing a rea­soned de­fense of their pro­ce­dure, Jaeggi et al. pro­vide merely a foot­note which refers in turn to a foot­note in an­other study. The sec­ond foot­note de­scribes un­pub­lished re­sults, ev­i­dently re­called by mem­ory over a span of 40 years, in­volv­ing a differ­ent test and a much less se­vere re­duc­tion in time.

In this con­text it bears re­peat­ing that the group that was tested on Raven’s ma­tri­ces (with pre­sum­ably the same time re­stric­tion) showed vir­tu­ally no im­prove­ment in test per­for­mance, in spite of eight days’ train­ing on work­ing mem­o­ry. Per­for­mance gains only ap­peared for the groups ad­min­is­tered the BOMAT. But the BOMAT differs in one im­por­tant re­spect from Raven’s. Raven’s ma­tri­ces are pre­sented in a 3 × 3 for­mat, whereas the BOMAT con­sists of a 5 × 3 ma­trix con­fig­u­ra­tion.

With 15 vi­sual fig­ures to keep track of in each test item in­stead of 9, the BOMAT puts added em­pha­sis on sub­jects’ abil­ity to hold de­tails of the fig­ures in work­ing mem­o­ry, es­pe­cially un­der the con­di­tion of a se­vere time con­straint. There­fore it is not sur­pris­ing that ex­ten­sive train­ing on a task of work­ing mem­ory would fa­cil­i­tate per­for­mance on the early and eas­i­est BOMAT test item­s-those that present less of a chal­lenge to fluid in­tel­li­gence.

This in­ter­pre­ta­tion ac­quires added plau­si­bil­ity from the na­ture of one of the two work­ing-mem­ory tasks ad­min­is­tered to the ex­per­i­men­tal groups. The au­thors main­tain that those tasks were “en­tirely differ­ent” from the test of fluid in­tel­li­gence. One of the tasks mer­its that de­scrip­tion: it was a se­quence of let­ters pre­sented au­di­to­rily through head­phones.

But the other work­ing-mem­ory task in­volved re­call of the lo­ca­tion of a small square in one of sev­eral po­si­tions in a vi­sual ma­trix pat­tern. It rep­re­sents in sim­pli­fied form pre­cisely the kind of de­tail re­quired to solve vi­sual analo­gies. Rather than be­ing “en­tirely differ­ent” from the test items on the BOMAT, this task seems well-de­signed to fa­cil­i­tate per­for­mance on that test."

Stern­berg re­viewed Jaeggi 2008

Email from Jaeggi to Pon­tus about the time­lim­it; vi­sual prob­lems bias RPM against women who have a slightly lower av­er­age vi­su­ospa­tial per­for­mance.

Nut­ley 2011 dis­cusses why one test may be in­suffi­cient when an ex­per­i­men­tal in­ter­ven­tion is done:

Since the de­fi­n­i­tion of Gf it­self stems from , us­ing the shared vari­ance of sev­eral tests to de­fine the Gf fac­tor, a sim­i­lar method should be used to mea­sure gains in Gf. An­other is­sue raised by Stern­berg (2008) is that the use of only one sin­gle train­ing task makes it diffi­cult to in­fer if the train­ing effect was due to some spe­cific as­pect of the task rather than the gen­eral effect of train­ing a con­struct.

Ship­stead, Redick, & En­gle 2012 elab­o­rate on how, while ma­trix-style IQ tests are con­sid­ered gold stan­dards, they are not per­fect mea­sures of IQ such that an in­crease in per­for­mance must re­flect an in­crease in un­der­ly­ing in­tel­li­gence:

…far trans­fer tasks are not per­fect mea­sures of abil­i­ty. In many train­ing stud­ies, Raven’s Pro­gres­sive Ma­tri­ces (Ravens; Raven, 1990, 1995, 1998) serves as the sole in­di­ca­tor of Gf. This “ma­trix rea­son­ing” task presents test tak­ers with a se­ries of ab­stract pic­tures that are arranged in a grid. One piece of the grid is miss­ing, and the test taker must choose an op­tion (from among sev­er­al) that com­pletes the se­quence. Jensen (1998) es­ti­mates that 64% of the vari­ance in Ravens per­for­mance can be ex­plained by Gf. Sim­i­lar­ly, Fig­ure 3 in­di­cates that in the study of Kane et al. (2004), 58% of the Ravens vari­ance was ex­plained by Gf. It is clear that Ravens is strongly re­lated to Gf. How­ev­er, 30%-40% of the vari­ance in Ravens is at­trib­ut­able to other in­flu­ences. Thus, when Ravens (or any other task) serves as the sole in­di­ca­tor of far trans­fer, per­for­mance im­prove­ments can be ex­plained with­out as­sum­ing that a gen­eral abil­ity has changed. In­stead, it can be par­si­mo­niously con­cluded that train­ing has in­flu­enced some­thing that is spe­cific to per­form­ing Ravens, but not nec­es­sar­ily ap­plic­a­ble to other rea­son­ing con­texts (Car­roll, 1993; Jensen, 1998; Moody, 2009; Schmiedek et al., 2010; te Ni­jen­huis, van Via­nen, & van der Flier, 2007).

…Pre­emp­tion of crit­i­cisms such as Moody’s (2009) is, how­ev­er, read­ily ac­com­plished through demon­stra­tion of trans­fer to sev­eral mea­sures of an abil­i­ty. Un­for­tu­nate­ly, the prac­tice of equat­ing posttest im­prove­ment on one task with change to cog­ni­tive abil­i­ties is preva­lent within the WM train­ing lit­er­a­ture (cf. Jaeggi et al., 2008; Kling­berg, 2010). This is par­tially dri­ven by the time and mon­e­tary costs as­so­ci­ated with con­duct­ing mul­ti­ses­sion, mul­ti­week stud­ies. Re­gard­less, train­ing stud­ies can greatly im­prove the per­sua­sive­ness of their re­sults by mea­sur­ing trans­fer via sev­eral tasks that differ in pe­riph­eral as­pects but con­verge on an abil­ity of in­ter­est (e.g., a ver­bal, Gf, and spa­tial task from Fig­ure 3). If a train­ing effect is ro­bust, it should be ap­par­ent in all tasks.

Ex­plicit at­tempts at mea­sur­ing speed­ing:

Seidler 2010

“Cog­ni­tive Train­ing As An In­ter­ven­tion To Im­prove Dri­ving Abil­ity In The Older Adult”, a by a group which in­cludes Su­sanne Jaeg­gi, stud­ied the effect of DNB on the dri­ving abil­ity of younger/older adults. As part of the be­fore/after test bat­tery, a Raven’s was ad­min­is­tered:

Type 2 tests in­cluded Raven’s ma­tri­ces (Raven et al., 1990), which is a stan­dard­ized test of fluid in­tel­li­gence, and the BOMAT and ver­bal analo­gies tests of in­tel­li­gence (Hossiep et al., 1995). We have pre­vi­ously shown that work­ing mem­ory train­ing trans­fers to per­for­mance on this task (Jaeggi et al., 2008), and we in­cluded it here for the sake of repli­ca­tion.

They found the :

There were no sig­nifi­cant group by test ses­sion in­ter­ac­tions for the in­tel­li­gence mea­sures or com­plex mo­tor tasks for the young adults, al­though one of the in­tel­li­gence mea­sures ex­hib­ited a trend for trans­fer effects that scaled with train­ing task gains.

…Un­like in our pre­vi­ous work (Jaeggi et al., 2008) we did not ob­serve trans­fer to mea­sures of in­tel­li­gence. This may have been a by-prod­uct of the rather ex­ten­sive pre and post test bat­tery of as­sess­ments that we per­formed, par­tic­u­larly given that one of the in­tel­li­gence mea­sures was al­ways per­formed last in the se­quence of tests. Given this, par­tic­i­pants may have been too fa­tigued and / or un­mo­ti­vated to per­form these tests well.

Jonasson 2011

“In­ves­ti­gat­ing train­ing and trans­fer in com­plex tasks with dual n-back”, bach­e­lor de­gree the­sis:

No clear con­sen­sus ex­ists in the sci­en­tific com­mu­nity of what con­sti­tutes effi­cient du­al-task­ing abil­i­ties. More­over, the train­ing of ex­ec­u­tive com­po­nents has been given in­creased at­ten­tion in the lit­er­a­ture in re­cent years. In­ves­ti­gat­ing trans­fer­abil­ity of cog­ni­tive train­ing in a com­plex task set­ting, thirty sub­jects prac­ticed for five days on a Name-Tag task (con­trols) or a Dual N-Back task (ex­per­i­men­tal), sub­se­quently be­ing tested on two trans­fer tasks; the Au­to­mated Op­er­a­tion Span and a dual task (Trail Mak­ing task + Math­e­mat­i­cal Ad­di­tion task). Dual N-Back train­ing pre­vi­ously trans­ferred to un­re­lated in­tel­li­gence tests and in this study is as­sumed to rely pri­mar­ily on ex­ec­u­tive at­ten­tion. Ex­ec­u­tive at­ten­tion, func­tion­ing to re­solve in­ter­fer­ence and main­tain­ing task-rel­e­vant in­for­ma­tion in work­ing mem­o­ry, has pre­vi­ously been linked to fluid in­tel­li­gence and to du­al-task­ing. How­ev­er, no trans­fer effects were re­vealed. The length of train­ing may have been too short to re­veal any such effects. How­ev­er, the three com­plex tasks cor­re­lated sig­nifi­cant­ly, sug­gest­ing com­mon re­sources, and there­fore hav­ing po­ten­tials as trans­fer tasks. No­tably, sub­jects with the high­est task-spe­cific im­prove­ments per­formed worse on the trans­fer tasks than sub­jects im­prov­ing less, sug­gest­ing that task-spe­cific gains do not di­rectly cor­re­late with any trans­fer effect. At pre­sent, if trans­fer ex­ists in these set­tings, data im­plies that five days of train­ing is in­suffi­cient for a trans­fer to oc­cur. Im­por­tant ques­tions for fu­ture re­search re­lates to the nec­es­sary con­di­tions for trans­fer to oc­cur, such as the amount of train­ing, neural cor­re­lates, at­ten­tion, and mo­ti­va­tion.

Caveats for this study:

  1. It did not at­tempt to mea­sure any form of Gf

  2. It used 30 to­tal sub­jects, or 15 in each group

  3. Train­ing was over 5-6 days, 16-20 min­utes each day (although the DNB sub­jects did in­crease their scores), which may not be enough; al­though Jonas­son com­ments (pg 44-45):

    Nev­er­the­less, train­ing for five days or less has also led to sig­nifi­cant im­prove­ments in per­for­mance on trans­fer tasks (Damos & Wick­ens, 1980; Kramer et al., 1995; Rueda et al., 2005). How­ev­er, the study by Kramer et al. (1995) may have trans­ferred a strat­egy rather than train­ing a spe­cific com­po­nent, and the study by Rueda et al. (2005) found trans­fer in chil­dren be­tween ages four and six, the chil­dren pos­si­bly be­ing more sus­cep­ti­ble to train­ing than adults.

  4. Jonas­son sug­gests that sub­jects were un­mo­ti­vat­ed, per­haps by the train­ing be­ing done at home on Lumosity.com; only one did the full 6 days of train­ing, and in­cen­tives often in­crease per­for­mance on IQ and other tests.

Chooi 2011

“Im­prov­ing In­tel­li­gence by In­creas­ing Work­ing Mem­ory Ca­pac­ity”, PhD the­sis:

…The cur­rent study aimed to repli­cate and ex­tend the orig­i­nal study con­ducted by Jaeggi et al. (2008) in a well-con­trolled ex­per­i­ment that could ex­plain the cause or causes of such trans­fer if in­deed the case. There were a to­tal of 93 par­tic­i­pants who com­pleted the study, and they were ran­domly as­signed to one of three groups - a pas­sive con­trol group, ac­tive con­trol group and ex­per­i­men­tal group. Half of the par­tic­i­pants were ran­domly as­signed to the 8-day con­di­tion and the other half to the 20-day con­di­tion. All par­tic­i­pants com­pleted a bat­tery of tests at pre- and post-tests that con­sisted of short timed tests, a com­plex work­ing mem­ory span and a [un­timed] ma­trix rea­son­ing task. Par­tic­i­pants in the ac­tive con­trol group prac­ticed for ei­ther 8 days or 20 days on the same task as the one used in the ex­per­i­men­tal group, the dual n-back, but at the eas­i­est level to con­trol for Hawthorne effect. Re­sults from the cur­rent study did not sug­gest any sig­nifi­cant im­prove­ment in the men­tal abil­i­ties test­ed, es­pe­cially fluid in­tel­li­gence and work­ing mem­ory ca­pac­i­ty, after train­ing for 8 days or 20 days. This leads to the con­clu­sion that in­creas­ing one’s work­ing mem­ory ca­pac­ity by train­ing and prac­tice did not trans­fer to im­prove­ment on fluid in­tel­li­gence as as­serted by Jaeggi and her col­leagues (2008, 2010).

Jonathan Toomim points out a con­cern about : the mul­ti­ple con­trol groups means that the num­ber of sub­jects do­ing ac­tual n-back­ing is small and the null re­sult is only trust­wor­thy if one ex­pects a dra­matic effect from n-back­ing, a huge effect size taken from Jaeggi 2010 (but not Jaeggi 2008’s smaller effect size). He com­ments: “the for DNB train­ing is prob­a­bly less than 0.98. (Of course, that’s what I be­lieved any­way be­fore I saw this.) The effect size could quite rea­son­ably still be as high as 0.75.” Chooi 2011 seems to have been sum­ma­rized as Chooi & Thomp­son 2012, which dis­cusses the power is­sue fur­ther:

A ma­jor lim­i­ta­tion of the study was the small sam­ple size and pos­si­bly sam­ple char­ac­ter­is­tic, which may have low­ered the power of analy­ses con­duct­ed. When Jaeggi et al. (2010) re­peated the study with 25 stu­dents who trained on the Raven’s Ad­vanced Pro­gres­sive Ma­tri­ces (RAPM) for 20 days, they ob­tained an effect size (Co­hen’s d) of 0.98. Ad­di­tion­al­ly, par­tic­i­pants in the Jaeggi et al. (2010) study were cul­tur­ally differ­ent from the par­tic­i­pants in the cur­rent study. Par­tic­i­pants from the for­mer study were un­der­grad­u­ates from a uni­ver­sity in Tai­wan (mean age=19.4), while those from the cur­rent study were mostly Amer­i­can stu­dents at­tend­ing a Mid­west­ern uni­ver­si­ty. The cur­rent study was de­signed ac­cord­ing to the claims put forth by Jaeggi et al. (2008) as a study of repli­ca­tion and ex­ten­sion. In that study, par­tic­i­pants were healthy, young adults who were slightly older (mean age=25.6 years) than the cur­rent sam­ple (mean age= 20.0), and they were re­cruited from a uni­ver­sity in Bern, Switzer­land. Effect sizes ob­tained from our study for RAPM were not as high as re­ported by Jaeggi et al. (2008, 2010) - d = 0.65 and d = 0.98 re­spec­tive­ly. With such large effect sizes, the analy­sis of paired t-test could achieve a power of 0.80 with 10- 12 par­tic­i­pants. Re­fer­ring to Ta­ble 4, the high­est RAPM effect size (d = 0.50) was from the 8-day pas­sive con­trol group that had 22 par­tic­i­pants and this achieved a power of 0.83. The 20-day train­ing group (n = 13) had an effect size of 0.06 in RAPM, and to achieve a power of 0.80 this group would need more than 1700 par­tic­i­pants. On the other hand, the effect size from the 20-day ac­tive con­trol group with 11 par­tic­i­pants was 0.40, and power could be im­proved by in­creas­ing the num­ber of par­tic­i­pants to 34. These ob­ser­va­tions led us to be­lieve that the lack of im­prove­ments in the test vari­ables was prob­a­bly due to a com­bi­na­tion of low sam­ple size and differ­ences in sam­ple char­ac­ter­is­tics, of which par­tic­i­pants in our study had re­stric­tion of range in in­tel­lec­tual abil­i­ty.

Preece 2011 / Palmer 2011

“The Effect of Work­ing Mem­ory (n-back) Train­ing on Fluid In­tel­li­gence”, David Preece 2011:

The present study repli­cated and ex­tended these re­sults by test­ing the fluid in­tel­li­gence con­struct us­ing a differ­ent type of fluid in­tel­li­gence test, and em­ploy­ing an ‘ac­tive’ rather than ‘no-con­tact’ con­trol group to ac­count for mo­ti­va­tional effects on in­tel­li­gence test per­for­mance. 58 par­tic­i­pants were in­volved and their fluid in­tel­li­gence was as­sessed pre-train­ing us­ing the Fig­ure Weights sub­test from the Wech­sler Adult In­tel­li­gence Scale - Fourth Edi­tion (WAIS-IV). Par­tic­i­pants were ran­domly as­signed to two groups (ex­per­i­men­tal or ac­tive con­trol), and both groups did a train­ing task on their home com­puter for 20 days, for 20 min­utes a day. The ex­per­i­men­tal group trained us­ing a sin­gle n-back task whilst the con­trol group com­pleted gen­eral knowl­edge and vo­cab­u­lary ques­tions. After train­ing, par­tic­i­pants were retested us­ing the Fig­ure Weights sub­test. Par­tic­i­pants’ Fig­ure Weights scores were analysed us­ing an analy­sis of co­vari­ance (ANCOVA). The re­sults of this analy­sis re­vealed no sig­nifi­cant differ­ence be­tween the train­ing groups in terms of per­for­mance on the Fig­ure Weights sub­test, sug­gest­ing that the n-back task was not effec­tive in in­creas­ing fluid rea­son­ing abil­i­ty. These find­ings were in con­trast to those of Jaeggi et al. (2008) and Jaeggi et al. (2010) and sug­gested that differ­ences be­tween the work­ing mem­ory group and con­trol group found in these stud­ies were likely the re­sult of place­bo/­mo­ti­va­tional effects rather than the prop­er­ties of the n-back task it­self.

Sub­jects were also tested on the RAPM pre/­post, but that was re­ported in a sep­a­rate the­sis, Vaughan Palmer’s “Im­prov­ing fluid in­tel­li­gence (Gf) though train­ing”, which is not avail­able. I have emailed the su­per­vis­ing pro­fes­sor for more in­for­ma­tion.

2 closely re­lated the­ses are “Im­prov­ing Mem­ory Us­ing N-Back Train­ing”, Beavon 2012 (short­-term & work­ing mem­o­ry); and “Vi­sual Mem­ory Im­prove­ment in Recog­ni­tion”, Prandl 2012.

Kundu et al 2012

“Be­hav­ioral and EEG Effects of Work­ing Mem­ory Train­ing” (RAPM sup­ple­ment); 13 con­trols and 13 ex­per­i­men­tals trained for 1000 min­utes on dual n-back (Brain Work­shop) or Tetris. “Train­ing does not ap­pear to trans­fer to gf [RAPM] or com­plex span [OSPAN].” This is not a pub­lished study but a con­fer­ence poster, so de­tails such as RAPM scores are not in­clud­ed. It may be re­lated to Kundu et al 2011.

Kundu et al 2013

The in­terim posters Kundu 2011 & 2012 were pub­lished as , Kundu et al 2013:

Al­though long con­sid­ered a na­tively en­dowed and fixed trait, work­ing mem­ory (WM) abil­ity has re­cently been shown to im­prove with in­ten­sive train­ing. What re­mains con­tro­ver­sial and poorly un­der­stood, how­ev­er, are the neural bases of these train­ing effects, and the ex­tent to which WM train­ing gains trans­fer to other cog­ni­tive tasks. Here we present ev­i­dence from hu­man elec­tro­phys­i­ol­ogy (EEG) and si­mul­ta­ne­ous tran­scra­nial mag­netic stim­u­la­tion (TMS) and EEG that the trans­fer of WM train­ing to other cog­ni­tive tasks is sup­ported by changes in task-re­lated effec­tive con­nec­tiv­ity in fron­topari­etal and pari­eto-oc­cip­i­tal net­works that are en­gaged by both the trained and trans­fer tasks. One con­se­quence of this effect is greater effi­ciency of stim­u­lus pro­cess­ing, as ev­i­denced by changes in EEG in­dices of in­di­vid­ual differ­ences in short­-term mem­ory ca­pac­ity and in vi­sual search per­for­mance. Trans­fer to search-re­lated ac­tiv­ity pro­vides ev­i­dence that some­thing more fun­da­men­tal than task-spe­cific strat­egy or stim­u­lus-spe­cific rep­re­sen­ta­tions have been learned. Fur­ther­more, these pat­terns of train­ing and trans­fer high­light the role of com­mon neural sys­tems in de­ter­min­ing in­di­vid­ual differ­ences in as­pects of vi­su­ospa­tial cog­ni­tion.

Salminen 2012

, Salmi­nen & Strobach & Schu­bert, Fron­tiers in Hu­man Neu­ro­science:

Re­cent stud­ies have re­ported im­prove­ments in a va­ri­ety of cog­ni­tive func­tions fol­low­ing sole work­ing mem­ory (WM) train­ing. In spite of the emer­gence of sev­eral suc­cess­ful train­ing par­a­digms, the scope of trans­fer effects has re­mained mixed. This is most likely due to the het­ero­gene­ity of cog­ni­tive func­tions that have been mea­sured and tasks that have been ap­plied. In the present study, we ap­proached this is­sue sys­tem­at­i­cally by in­ves­ti­gat­ing trans­fer effects from WM train­ing to differ­ent as­pects of ex­ec­u­tive func­tion­ing. Our train­ing task was a de­mand­ing WM task that re­quires si­mul­ta­ne­ous per­for­mance of a vi­sual and an au­di­tory n-back task, while the trans­fer tasks tapped WM up­dat­ing, co­or­di­na­tion of the per­for­mance of mul­ti­ple si­mul­ta­ne­ous tasks (i.e., du­al-tasks) and se­quen­tial tasks (i.e., task switch­ing), and the tem­po­ral dis­tri­b­u­tion of at­ten­tional pro­cess­ing. Ad­di­tion­al­ly, we ex­am­ined whether WM train­ing im­proves rea­son­ing abil­i­ties; a hy­poth­e­sis that has so far gained mixed sup­port. Fol­low­ing train­ing, par­tic­i­pants showed im­prove­ments in the trained task as well as in the trans­fer WM up­dat­ing task. As for the other ex­ec­u­tive func­tions, trained par­tic­i­pants im­proved in a task switch­ing sit­u­a­tion and in at­ten­tional pro­cess­ing. There was no trans­fer to the du­al-task sit­u­a­tion or to rea­son­ing skills. These re­sults, there­fore, con­firm pre­vi­ous find­ings that WM can be trained, and ad­di­tion­al­ly, they show that the train­ing effects can gen­er­al­ize to var­i­ous other tasks tap­ping on ex­ec­u­tive func­tions.

Pas­sive con­trol group; un­speeded RAPM test.

Redick et al 2012

“No ev­i­dence of trans­fer after work­ing mem­ory train­ing: A con­trolled, ran­dom­ized study” (sup­ple­ment), Redick et al 2012; ab­stract:

Nu­mer­ous re­cent stud­ies seem to pro­vide ev­i­dence for the gen­eral in­tel­lec­tual ben­e­fits of work­ing mem­ory train­ing. In re­views of the train­ing lit­er­a­ture, Ship­stead, Redick, and En­gle (2010, in press) ar­gued that the field should treat re­cent re­sults with a crit­i­cal eye. Many pub­lished work­ing mem­ory train­ing stud­ies suffer from de­sign lim­i­ta­tions (no-con­tact con­trol groups, sin­gle mea­sures of cog­ni­tive con­struct­s), mixed re­sults (trans­fer of train­ing gains to some tasks but not oth­ers, in­con­sis­tent trans­fer to the same tasks across stud­ies), and lack of the­o­ret­i­cal ground­ing (i­den­ti­fy­ing the mech­a­nisms re­spon­si­ble for ob­served trans­fer). The cur­rent study com­pared young adults who re­ceived 20 ses­sions of prac­tice on an adap­tive dual n-back pro­gram (work­ing mem­ory train­ing group) or an adap­tive vi­sual search pro­gram (ac­tive place­bo-con­trol group) with a no-con­tact con­trol group that re­ceived no prac­tice. In ad­di­tion, all sub­jects com­pleted pre-test, mid-test, and post-test ses­sions, com­pris­ing mul­ti­ple mea­sures of fluid in­tel­li­gence, mul­ti­task­ing, work­ing mem­ory ca­pac­i­ty, crys­tal­lized in­tel­li­gence, and per­cep­tual speed. De­spite im­prove­ments on both the dual n-back and vi­sual search tasks with prac­tice, and de­spite a high level of sta­tis­ti­cal pow­er, there was no pos­i­tive trans­fer to any of the cog­ni­tive abil­ity tests. We dis­cuss these re­sults in the con­text of pre­vi­ous work­ing mem­ory train­ing re­search, and ad­dress is­sues for fu­ture work­ing mem­ory train­ing stud­ies.

75 sub­jects; RAPM was speed­ed.

Rudebeck 2012

, Rude­beck et al 2012:

One cur­rent chal­lenge in cog­ni­tive train­ing is to cre­ate a train­ing regime that ben­e­fits mul­ti­ple cog­ni­tive do­mains, in­clud­ing episodic mem­o­ry, with­out re­ly­ing on a large bat­tery of tasks, which can be time-con­sum­ing and diffi­cult to learn. By giv­ing care­ful con­sid­er­a­tion to the neural cor­re­lates un­der­ly­ing episodic and work­ing mem­o­ry, we de­vised a com­put­er­ized work­ing mem­ory train­ing task in which neu­ro­log­i­cally healthy par­tic­i­pants were re­quired to mon­i­tor and de­tect rep­e­ti­tions in two streams of spa­tial in­for­ma­tion (s­pa­tial lo­ca­tion and scene iden­ti­ty) pre­sented si­mul­ta­ne­ously (i.e. a dual n-back par­a­dig­m). Par­tic­i­pants’ episodic mem­ory abil­i­ties were as­sessed be­fore and after train­ing us­ing two ob­ject and scene recog­ni­tion mem­ory tasks in­cor­po­rat­ing mem­ory con­fi­dence judg­ments. Fur­ther­more, to de­ter­mine the gen­er­al­iz­abil­ity of the effects of train­ing, we also as­sessed fluid in­tel­li­gence us­ing a ma­trix rea­son­ing task. By ex­am­in­ing the differ­ence be­tween pre- and post-train­ing per­for­mance (i.e. gain scores), we found that the train­ers, com­pared to non-train­ers, ex­hib­ited a sig­nifi­cant im­prove­ment in fluid in­tel­li­gence after 20 days. In­ter­est­ing­ly, pre-train­ing fluid in­tel­li­gence per­for­mance, but not train­ing task im­prove­ment, was a sig­nifi­cant pre­dic­tor of post-train­ing fluid in­tel­li­gence im­prove­ment, with lower pre-train­ing fluid in­tel­li­gence as­so­ci­ated with greater post-train­ing gain. Cru­cial­ly, train­ers who im­proved the most on the train­ing task also showed an im­prove­ment in recog­ni­tion mem­ory as cap­tured by d-prime scores and es­ti­mates of rec­ol­lec­tion and fa­mil­iar­ity mem­o­ry. Train­ing task im­prove­ment was a sig­nifi­cant pre­dic­tor of gains in recog­ni­tion and fa­mil­iar­ity mem­ory per­for­mance, with greater train­ing im­prove­ment lead­ing to more marked gains. In con­trast, lower pre-train­ing rec­ol­lec­tion mem­ory scores, and not train­ing task im­prove­ment, led to greater rec­ol­lec­tion mem­ory per­for­mance after train­ing. Our find­ings demon­strate that prac­tice on a sin­gle work­ing mem­ory task can po­ten­tially im­prove as­pects of both episodic mem­ory and fluid in­tel­li­gence, and that an ex­ten­sive train­ing regime with mul­ti­ple tasks may not be nec­es­sary.

Speeded BOMAT (“Due to time re­stric­tions and the pos­si­bil­ity of ceil­ing effects as­so­ci­ated with some Gf tests, par­tic­i­pants were given 10 min­utes to com­plete as many pat­terns as they could in each as­sess­ment ses­sion (for a sim­i­lar pro­ce­dure see Jaeggi et al 2008).”); 55 sub­jects to­tal, ex­per­i­men­tals trained for 400 min­utes, pas­sive con­trol group. The im­prove­ment pre­dic­tor sounds like a post hoc analy­sis and may be some­thing like re­gres­sion to the mean.

Heinzel et al 2013

“Work­ing mem­ory train­ing im­prove­ments and gains in non-trained cog­ni­tive tasks in young and older adults”, Heinzel et al 2013:

Pre­vi­ous stud­ies on work­ing mem­ory train­ing have in­di­cated that trans­fer to non-trained tasks of other cog­ni­tive do­mains may be pos­si­ble. The aim of this study is to com­pare work­ing mem­ory train­ing and trans­fer effects be­tween younger and older adults (n = 60). A novel ap­proach to adap­tive n-back train­ing (12 ses­sions) was im­ple­mented by vary­ing the work­ing mem­ory load and the pre­sen­ta­tion speed. All par­tic­i­pants com­pleted a neu­ropsy­cho­log­i­cal bat­tery of tests be­fore and after the train­ing. On av­er­age, younger train­ing par­tic­i­pants achieved diffi­culty level 12 after train­ing, while older train­ing par­tic��i­pants only reached diffi­culty level 5. In younger par­tic­i­pants, trans­fer to Ver­bal Flu­ency and Digit Sym­bol Sub­sti­tu­tion test was found. In older par­tic­i­pants, we ob­served a trans­fer to Digit Span For­ward, CERAD De­layed Re­call, and Digit Sym­bol Sub­sti­tu­tion test. Re­sults sug­gest that work­ing mem­ory train­ing may be a ben­e­fi­cial in­ter­ven­tion for main­tain­ing and im­prov­ing cog­ni­tive func­tion­ing in old age.

Sin­gle n-back; pas­sive con­trol group; no trans­fer in young or old train­ing group to “Raven’s Stan­dard Pro­gres­sive Ma­tri­ces (Raven’s SPM) and the Fig­ural Re­la­tions sub­test of a Ger­man in­tel­li­gence test (Leis­tung­spruef­sys­tem, LPS, Horn, 1983)” (in­creased but sam­ple size is too small to reach sta­tis­ti­cal-sig­nifi­cance in the young group); RPM speeded (7.5 min­utes). See pg19 for graphs of the IQ test per­for­mance.

Onken 2013

“Trans­fer von Ar­beits­gedächt­nis­train­ing auf die flu­ide In­tel­li­genz”, Jo­hanna Onken 2013; some sort of re-re­port­ing or ver­sion of the Heinzel da­ta.

Fluid in­tel­li­gence de­scribes the abil­ity to think ab­stract, to adapt to new sit­u­a­tions and to solve un­known prob­lems. It is im­por­tant for learn­ing as well as for aca­d­e­mic and pro­fes­sional suc­cess. Work­ing mem­ory is char­ac­ter­ized as a cog­ni­tive sys­tem, that saves in­for­ma­tion over a short pe­riod of time in spite of pos­si­ble dis­trac­tions. More­over, work­ing mem­ory is able to as­sess the rel­e­vance of in­for­ma­tion while re­quire­ments change. Effec­tive im­plicit train­ing is able to in­crease the work­ing mem­ory ca­pac­i­ty. Fur­ther­more it was shown that work­ing mem­ory train­ing may also cause trans­fer effects to higher cog­ni­tive abil­i­ties such as fluid in­tel­li­gence. To clar­ify the un­der­ly­ing processes of this trans­fer, var­i­ous trans­fer mod­els were pre­sent­ed, which ei­ther ac­cen­tu­ate the rel­e­vance of pro­cess­ing speed, ex­ec­u­tive func­tions or short time mem­o­ry. The pur­pose of this sur­vey was to con­firm trans­fer effects of work­ing mem­ory train­ing to differ­ent cog­ni­tive abil­i­ties and, on the other hand, to in­ves­ti­gate the mech­a­nism of the trans­fer ac­cord­ing to the pro­posed trans­fer mod­els. 30 healthy sub­jects [age 22-30 years] par­tic­i­pated in the study and were ran­domly as­signed to ei­ther train­ing or con­trol group. The train­ing group prac­ticed an adap­tive N-back work­ing mem­ory task for four weeks. Be­fore, after one week and after four weeks of the train­ing, a range of neu­ropsy­cho­log­i­cal tasks was per­formed by the par­tic­i­pants, test­ing for differ­ent cog­ni­tive abil­i­ties. Rel­a­tive to the con­trol group that did not par­tic­i­pate in the train­ing, trans­fer effects to pro­cess­ing speed, ex­ec­u­tive func­tions and fluid in­tel­li­gence tasks have been found. Ad­di­tion­al­ly, the train­ing re­sulted in a sig­nifi­cant short­en­ing of re­ac­tion time. In sum­ma­ry, the present study demon­strates that com­plex cog­ni­tive abil­i­ties can be im­proved through effec­tive work­ing mem­ory train­ing. The ques­tion on which cog­ni­tive mech­a­nisms the trans­fer is based could not be an­swered de­fin­i­tively by this study. The re­sults sug­gest that the adap­tive work­ing mem­ory train­ing has led mainly to faster basal cog­ni­tive process­es, which in turn re­sulted in a faster pro­cess­ing of in­tel­li­gence tests.

30 sub­jects, pas­sive con­trol group, 4 weeks; con­trols paid 50 eu­ros, ex­per­i­men­tals 150 eu­ros, 480 min­utes of train­ing, sin­gle n-back. IQ tests ad­min­is­tered: RPM, LPS, MWT-B. On pg40 are all the post-test re­sults: “Tabelle 3.7: Deskrip­tive Daten der Neu­ropsy­cholo­gie im Posttest (t3)”; dis­cus­sion of RPM re­sults on pg46.

Thompson et al 2013


…The cur­rent study at­tempted to repli­cate and ex­pand those re­sults by ad­min­is­ter­ing a broad as­sess­ment of cog­ni­tive abil­i­ties and per­son­al­ity traits to young adults who un­der­went 20 ses­sions of an adap­tive dual n-back work­ing mem­ory train­ing pro­gram and com­par­ing their post-train­ing per­for­mance on those tests to a matched set of young adults who un­der­went 20 ses­sions of an adap­tive at­ten­tional track­ing pro­gram. Pre- and post-train­ing mea­sure­ments of fluid in­tel­li­gence, stan­dard­ized in­tel­li­gence tests, speed of pro­cess­ing, read­ing skills, and other tests of work­ing mem­ory were as­sessed. Both train­ing groups ex­hib­ited sub­stan­tial and spe­cific im­prove­ments on the trained tasks that per­sisted for at least 6 months post-train­ing, but no trans­fer of im­prove­ment was ob­served to any of the non-trained mea­sure­ments when com­pared to a third un­trained group serv­ing as a pas­sive con­trol. These find­ings fail to sup­port the idea that adap­tive work­ing mem­ory train­ing in healthy young adults en­hances work­ing mem­ory ca­pac­ity in non-trained tasks, fluid in­tel­li­gence, or other mea­sures of cog­ni­tive abil­i­ties.

Co­vari­ate de­tails:

…Two groups of young adults, strat­i­fied so as to be equated on ini­tial fluid IQ scores, were ran­domly as­signed to two con­di­tions (a ran­dom­ized con­trolled trial or RCT). The ex­per­i­men­tal group per­formed the dual n-back task (as in the orig­i­nal Jaeggi et al., 2008 study [6]) for ap­prox­i­mately 40 min­utes per day, 5 days per week for 4 weeks (20 ses­sions of 30 blocks per ses­sion, ex­ceed­ing the max­i­mum of 19 ses­sions of 20 blocks per day in the orig­i­nal Jaeggi et al., 2008 study). An ac­tive con­trol group per­formed a vi­su­ospa­tial skill learn­ing task, mul­ti­ple ob­ject track­ing (or MOT), on an iden­ti­cal train­ing sched­ule. We also tested a no-con­tact group equated for ini­tial fluid IQ in case both kinds of train­ing en­hanced cog­ni­tive abil­i­ties…­Par­tic­i­pants were given 25 min­utes to com­plete each half of the RAPM…Participants in the train­ing groups were paid $20 per train­ing ses­sion, with a $20 bonus per week for com­plet­ing all five train­ing ses­sions in that week. All par­tic­i­pants were paid $20 per hour for be­hav­ioral test­ing, and $30 per hour for imag­ing ses­sions (data from imag­ing ses­sions are re­ported sep­a­rate­ly)….After re­cruit­ment, par­tic­i­pants un­der­went ap­prox­i­mately six hours of be­hav­ioral test­ing spread across three days and two hours of struc­tural and func­tional mag­netic res­o­nance imag­ing. [Thomp­son says there were ad­di­tional pay­ments for imag­ing not men­tioned, so the true ex­pect­ed-value of par­tic­i­pa­tion was $740.]

In line with my meta-analy­sis’s null on a dose-re­sponse effect:

One method of as­sess­ing whether the amount of train­ing im­prove­ment affects the de­gree of trans­fer is to mea­sure the cor­re­la­tion be­tween train­ing and trans­fer gains. For both the n-back and MOT groups, a pos­i­tive cor­re­la­tion was ob­served be­tween the amount of im­prove­ment dur­ing train­ing and the amount of im­prove­ment on the trained task be­tween the pre- and post-assess­ment (n-back r = .85, p<.0001; MOT r = .77, p<.0001). How­ev­er, the amount of train­ing gain did not sig­nifi­cantly pre­dict im­prove­ment on any trans­fer task; par­tic­i­pants who im­proved to a greater ex­tent on the train­ing tasks did not im­prove more or less on po­ten­tial trans­fer tasks than did par­tic­i­pants who im­proved to a lesser ex­tent (all n-back r val­ues <.33, all p’s >.15; all MOT r val­ues <.38, all p’s >.11). Fig­ure S2 de­picts the ab­sence of a re­la­tion be­tween im­prove­ment on trained tasks and the post-train­ing changes in the RAPM and the com­bined span tasks.

Note also that the post hoc split of chil­dren into ‘im­provers’ and not in that Jaeggi pa­per does not repli­cate here ei­ther:

An­other analy­sis that has pre­vi­ously re­vealed a differ­ence in trans­fer be­tween par­tic­i­pants who ex­hib­ited larger or smaller train­ing gains has been a di­vi­sion of par­tic­i­pants into groups based on train­ing gains above or be­low the group me­dian (me­dian split) [15]. Such a me­dian split of par­tic­i­pants in the present study who per­formed the n-back train­ing yielded no sig­nifi­cant differ­ences in trans­fer be­tween groups (all n-back t-ra­tios <1.78, all p’s >.09). The only trans­fer mea­sure that ap­proached sig­nifi­cance (at p = .09) was on the RAPM test, in which the par­tic­i­pants who im­proved less on the trained n-back task had higher scores on the post-train­ing be­hav­ioral test­ing. Sim­i­lar­ly, when sep­a­rat­ing the MOT par­tic­i­pants into two groups based on me­dian MOT im­prove­ment, the two groups showed no sig­nifi­cant differ­ences in trans­fer per­for­mance (all MOT t-ra­tios <1.74, all p’s >.10).

The per­son­al­ity cor­re­lates from Stud­er-Luethi 2012 also don’t work:

We also ex­am­ined whether per­son­al­ity as­sess­ments were as­so­ci­ated with differ­ent train­ing or trans­fer out­comes. Nei­ther the Dweck mea­sure of at­ti­tude to­ward in­tel­li­gence (a “growth mind­set”) nor mea­sures of con­sci­en­tious­ness or grit cor­re­lated sig­nifi­cantly with train­ing gains on ei­ther train­ing task, al­though there was a trend to­ward a sig­nifi­cant neg­a­tive cor­re­la­tion be­tween the growth mind­set and im­prove­ment on the n-back train­ing task (r = −.44, p = .051), such that par­tic­i­pants who viewed in­tel­li­gence as more mal­leable had less im­prove­ment across their n-back train­ing. A greater growth mind­set score was pos­i­tively cor­re­lat­ed, how­ev­er, with im­prove­ment on the Ravens Ad­vanced Pro­gres­sive Ma­tri­ces in the n-back group (r = .53, p = .017) and in the pas­sive con­trol group (r = .51, p = .027), but not in the MOT con­trol group (r = .031, p>.9). No other trans­fer mea­sures were sig­nifi­cantly pre­dicted by growth mind­set scores.

Al­though the con­sci­en­tious­ness scores and “grit” scores were highly cor­re­lated in each of the three treat­ment groups (n-back r = .75, p<.001; MOT r = .70, p<.001; pas­sive r = .76, p<.001), the two mea­sures differed in their cor­re­la­tions with the be­hav­ioral out­come mea­sures. A higher “grit” score pre­dicted less im­prove­ment on the RAPM for the n-back group (r = −.45, p = .049) and the MOT group (r = −.58, p = .009), such that par­tic­i­pants who viewed them­selves as hav­ing more “grit” im­proved less on the RAPM after train­ing, al­though this re­la­tion­ship did not hold for the No-Con­tact group (r = .17, p = .5). Sim­i­lar­ly, a higher score on the con­sci­en­tious­ness mea­sure pre­dicted less im­prove­ment on the RAPM for the MOT group (r = −.57, p = .01), such that par­tic­i­pants who saw them­selves as more con­sci­en­tious im­proved less on the RAPM after train­ing, al­though this was not ob­served in ei­ther of the other two groups (n-back r = −.21, p = .37; no-con­tact r = −.04, p = .85). Fi­nal­ly, a high con­sci­en­tious­ness score pre­dicted a lower Pair Can­cel­la­tion im­prove­ment within the MOT group (r = −.47, p = .04), but not in the n-back or no-con­tact con­trol groups (n-back r = −.07, p = .77; no-con­tact r = −.13, p = .58). No other trans­fer mea­sures were sig­nifi­cantly pre­dicted by ei­ther con­sci­en­tious­ness or grit scores.

Smith et al 2013

“Ex­plor­ing the effec­tive­ness of com­mer­cial and cus­tom-built games for cog­ni­tive train­ing”, Smith et al 2013

There is in­creas­ing in­ter­est in quan­ti­fy­ing the effec­tive­ness of com­puter games in non-en­ter­tain­ment do­mains. We have ex­plored gen­eral in­tel­li­gence im­prove­ments for par­tic­i­pants us­ing ei­ther a com­mer­cial-off-the-shelf (COTS) game [Brain Age], a cus­tom do-it-y­our­self (DIY) train­ing sys­tem for a work­ing mem­ory task [DNB] or an on­line strat­egy game to a con­trol group (with­out train­ing). Forty uni­ver­sity level par­tic­i­pants were di­vided into four groups (COTS, DIY, Gam­ing, [Pas­sive] Con­trol) and were eval­u­ated three times (pre-in­ter­ven­tion, post-in­ter­ven­tion, 1-week fol­low-up) with three weeks of train­ing. In gen­eral in­tel­li­gence tests both cog­ni­tive train­ing sys­tems (COTS and DIY groups) failed to pro­duce [s­ta­tis­ti­cal­ly-]sig­nifi­cant im­prove­ments in com­par­i­son to a con­trol group or a gam­ing group. Also nei­ther cog­ni­tive train­ing sys­tem pro­duced [s­ta­tis­ti­cal­ly-]sig­nifi­cant im­prove­ments over the in­ter­ven­tion or fol­low-up pe­ri­ods.

Dual n-back; RAPM (10 min­utes each test); 1 pas­sive con­trol group, 2 ac­tives; 340 min­utes train­ing; min­i­mal com­pen­sa­tion (course credit & en­try into “a prize draw”, which was worth 100£). Very small sam­ple sizes (~10 in each of the 4 group­s).

Nussbaumer et al 2013

“Lim­i­ta­tions and chances of work­ing mem­ory train­ing”, Nuss­baumer et al 2013:

Re­cent stud­ies show con­tro­ver­sial re­sults on the train­abil­ity of work­ing mem­ory (WM) ca­pac­ity be­ing a lim­it­ing fac­tor of hu­man cog­ni­tion. In or­der to con­tribute to this open ques­tion we in­ves­ti­gated if par­tic­i­pants im­prove in trained tasks and whether gains gen­er­al­ize to un­trained WM tasks, math­e­mat­i­cal prob­lem solv­ing and in­tel­li­gence tests.

83 adults trained over a three week pe­riod (7.5 hours to­tal) in one of the fol­low­ing con­di­tions: A high, a medium or a low WM load group. The present find­ings show that task spe­cific char­ac­ter­is­tics could be learned but that there was no trans­fer be­tween trained and un­trained tasks which had no com­mon el­e­ments. Pos­i­tive trans­fer oc­curred be­tween two tasks fo­cus­ing on in­hibitory process­es. It might be pos­si­ble to en­hance this spe­cific com­po­nent of WM but not WM ca­pac­ity as such. A pos­si­ble en­hance­ment in a learn­ing test is of high ed­u­ca­tional in­ter­est and worth­while to be in­ves­ti­gated fur­ther.

One of the two IQ tests was the RAPM; this was dual n-back, but it was adap­tive only in the “high” ex­per­i­men­tal group (so the “medium” and “low” groups are largely ir­rel­e­van­t). Pa­per does not pro­vide RAPM score de­tails, so I emailed the lead au­thor.

Oelhafen et al 2013

“In­creased pari­etal ac­tiv­ity after train­ing of in­ter­fer­ence con­trol”,

…In the cur­rent study, we ex­am­ined whether train­ing on two vari­ants of the adap­tive dual n-back task would affect un­trained task per­for­mance and the cor­re­spond­ing elec­tro­phys­i­o­log­i­cal even­t-re­lated po­ten­tials (ERPs). 43 healthy young adults trained for three weeks with a high or low in­ter­fer­ence train­ing vari­ant of the dual n-back task, or they were as­signed to a pas­sive con­trol group. While n-back train­ing with high in­ter­fer­ence led to par­tial im­prove­ments in the At­ten­tion Net­work Test (ANT), we did not find trans­fer to mea­sures of work­ing mem­ory and fluid in­tel­li­gence. ERP analy­sis in the n-back task and the ANT in­di­cated over­lap­ping processes in the P3 time range. More­over, in the ANT, we de­tected in­creased pari­etal ac­tiv­ity for the in­ter­fer­ence train­ing group alone. In con­trast, we did not find elec­tro­phys­i­o­log­i­cal differ­ences be­tween the low in­ter­fer­ence train­ing and the con­trol group. These find­ings sug­gest that train­ing on an in­ter­fer­ence con­trol task leads to higher elec­tro­phys­i­o­log­i­cal ac­tiv­ity in the pari­etal cor­tex, which may be re­lated to im­prove­ments in pro­cess­ing speed, at­ten­tional con­trol, or both.

Sprenger et al 2013

“Train­ing work­ing mem­o­ry: Lim­its of trans­fer”, Sprenger et al 2013; ab­stract:

In two ex­per­i­ments (to­tal­ing 253 adult par­tic­i­pants), we ex­am­ined the ex­tent to which in­ten­sive work­ing mem­ory train­ing led to im­prove­ments on un­trained mea­sures of cog­ni­tive abil­i­ty. Al­though par­tic­i­pants showed im­prove­ment on the trained task and on tasks that ei­ther shared task char­ac­ter­is­tics or stim­uli, we found no ev­i­dence that train­ing led to gen­eral im­prove­ments in work­ing mem­o­ry. Us­ing Bayes Fac­tor analy­sis, we show that the data gen­er­ally sup­port the hy­poth­e­sis that work­ing mem­ory train­ing was in­effec­tive at im­prov­ing gen­eral cog­ni­tive abil­i­ty. This con­clu­sion held even after con­trol­ling for a num­ber of in­di­vid­ual differ­ences, in­clud­ing need for cog­ni­tion, be­liefs in the mal­leabil­ity of in­tel­li­gence, and age.

Colom et al 2013

“Adap­tive n-back train­ing does not im­prove fluid in­tel­li­gence at the con­struct lev­el; gains on in­di­vid­ual tests sug­gest train­ing may en­hance vi­su­ospa­tial pro­cess­ing”, Colom et al 2013:

Short­-term adap­tive cog­ni­tive train­ing based on the n-back task is re­ported to in­crease scores on in­di­vid­ual abil­ity tests, but the key ques­tion of whether such in­creases gen­er­al­ize to the in­tel­li­gence con­struct is not clear. Here we eval­u­ate flu­id/ab­stract in­tel­li­gence (Gf), crys­tal­lized/ver­bal in­tel­li­gence (Gc), work­ing mem­ory ca­pac­ity (WMC), and at­ten­tion con­trol (ATT) us­ing di­verse mea­sures, with equiv­a­lent ver­sions, for es­ti­mat­ing any changes at the con­struct level after train­ing. Be­gin­ning with a sam­ple of 169 par­tic­i­pants, two groups of twen­ty-eight women each were se­lected and matched for their gen­eral cog­ni­tive abil­ity scores and de­mo­graphic vari­ables. Un­der strict su­per­vi­sion in the lab­o­ra­to­ry, the train­ing group com­pleted an in­ten­sive adap­tive train­ing pro­gram based on the n-back task (vi­su­al, au­di­to­ry, and dual ver­sions) across twen­ty-four ses­sions dis­trib­uted over twelve weeks. Re­sults showed this group had the ex­pected sys­tem­atic im­prove­ments in n-back per­for­mance over time; this per­for­mance sys­tem­at­i­cally cor­re­lated across ses­sions with Gf, Gc, and WMC, but not with ATT. How­ev­er, the main find­ing showed no sig­nifi­cant changes in the as­sessed psy­cho­log­i­cal con­structs for the train­ing group as com­pared with the con­trol group. Nev­er­the­less, post-hoc analy­ses sug­gested that spe­cific tests and tasks tap­ping vi­su­ospa­tial pro­cess­ing might be sen­si­tive to train­ing.

<– One hun­dred and sixty nine psy­chol­ogy un­der­grad­u­ates com­pleted a bat­tery of twelve in­tel­li­gence tests and cog­ni­tive tasks mea­sur­ing flu­id-ab­stract and crys­tal­lized-ver­bal in­tel­li­gence, work­ing mem­ory ca­pac­i­ty, and at­ten­tion con­trol. After com­put­ing a gen­eral in­dex from the six in­tel­li­gence tests, two groups of twen­ty-eight fe­males were re­cruited for the study. They were paid for their par­tic­i­pa­tion€. Mem­bers of each group were care­fully matched for their gen­eral in­tel­li­gence in­dex, so they were per­fectly over­lapped and rep­re­sented a wide range of scores. All par­tic­i­pants were right hand­ed, as as­sessed by the Ed­in­burgh Test (Old­field, 1971). They also com­pleted a set of ques­tions ask­ing for med­ical or psy­chi­atric dis­or­ders, as well as sub­stance in­take. The re­cruit­ment process fol­lowed the Helsinki guide­lines (World Med­ical As­so­ci­a­tion, 2008) and the lo­cal ethics com­mit­tee ap­proved the study. De­scrip­tive sta­tis­tics for the de­mo­graphic vari­ables and per­for­mance on the cog­ni­tive mea­sures for the two groups of par­tic­i­pants (train­ing and con­trol) can be seen in the Ap­pen­dix (Table A.1.). [200€ if as­signed to the train­ing group and 100€ if as­signed to the con­trol group.] [150 eu­ros in dol­lars is $204]

The col­lec­tive psy­cho­log­i­cal as­sess­ment for the pretest stage was done from Sep­tem­ber 19 to Oc­to­ber 14 2011. Par­tic­i­pants were as­sessed in groups not greater than twen­ty-five. The data ob­tained for the com­plete group (N = 169) were an­a­lyzed for re­cruit­ing the train­ing (N = 28) and con­trol (N = 28) groups based on the gen­eral in­dex com­puted from the mea­sures of fluid and crys­tal­lized in­tel­li­gence (Table A.1.). The adap­tive cog­ni­tive train­ing pro­gram be­gan in No­vem­ber 14 2011, re­mained ac­tive un­til 2012-02-17, and lasted for twelve weeks (with a break from De­cem­ber 24 2011 to Jan­u­ary 9 2012). The psy­cho­log­i­cal as­sess­ment for the posttest was done in­di­vid­u­ally from Feb­ru­ary 20 to March 09 (in­tel­li­gence tests) and from March 12 to March 30 (cog­ni­tive tasks) 2012.

In­tel­li­gence and cog­ni­tive con­structs were as­sessed by three mea­sures each. As noted above, fluid in­tel­li­gence (Gf) re­quires ab­stract prob­lem solv­ing abil­i­ties, whereas crys­tal­lized in­tel­li­gence (Gc) in­volves the men­tal ma­nip­u­la­tion of cul­tural knowl­edge. Gf was mea­sured by screen­ing ver­sions (odd num­bered items and even num­bered items for the pretest and posttest eval­u­a­tions, re­spec­tive­ly) of the Raven Ad­vanced Pro­gres­sive Ma­tri­ces Test (RAPM), the ab­stract rea­son­ing sub­test from the Differ­en­tial Ap­ti­tude Test (DAT-AR), and the in­duc­tive rea­son­ing sub­test from the Pri­mary Men­tal Abil­i­ties Bat­tery (PMA-R). Gc was mea­sured by screen­ing ver­sions (odd num­bered items and even num­bered items for the pretest and posttest eval­u­a­tions, re­spec­tive­ly) of the ver­bal rea­son­ing sub­test from the DAT (DAT-VR), the nu­mer­i­cal rea­son­ing sub­test from the DAT (DAT-NR), and the vo­cab­u­lary sub­test from the PMA (PMA-V). Gf and Gc were mea­sured by tests with (PMA sub­tests) and with­out (RAPM and DAT sub­tests) highly speeded con­straints.

The frame­work for the cog­ni­tive train­ing pro­gram fol­lowed the guide­lines re­ported by Jaeggi et al. (2008) but it was re-pro­grammed for Vi­sual Ba­sic (2008 Ver­sion). Nev­er­the­less, there were some differ­ences: (a) the train­ing be­gan with four ses­sions (weeks 1 and 2) with a vi­sual adap­tive n-back ver­sion and four ses­sions (weeks 3 and 4) with an au­di­tory adap­tive n-back ver­sion be­fore fac­ing the six­teen ses­sions of the adap­tive n-back dual pro­gram (weeks 5 to 12), and (b) while the train­ing pro­gram is usu­ally com­pleted in one mon­th, here we ex­tended the train­ing pe­riod to three months (12 week­s). There were two train­ing ses­sions per week last­ing around 30 min each and they took place un­der strict su­per­vi­sion in the lab­o­ra­to­ry. Par­tic­i­pants worked within in­di­vid­ual cab­ins and the ex­per­i­menter was al­ways avail­able for at­tend­ing any re­quest they might have. Data were an­a­lyzed every week for check­ing their progress at both the in­di­vid­ual and the group lev­el. Par­tic­i­pants re­ceived sys­tem­atic feed­back re­gard­ing their per­for­mance. Fur­ther­more, every two weeks par­tic­i­pants com­pleted a mo­ti­va­tion ques­tion­naire ask­ing for their (a) in­volve­ment with the task, (b) per­ceived diffi­culty lev­el, (c) per­ceived chal­leng­ing of the task lev­els, and (d) ex­pec­ta­tions for fu­ture achieve­ment. At the end of the train­ing pe­riod par­tic­i­pants were asked with re­spect to their gen­eral eval­u­a­tion of the pro­gram. Us­ing a rat­ing scale from 0 to 10, av­er­age val­ues were (a) 8.1 (range 8.0 to 8.2 across ses­sion­s), (b) 7.9 (range 7.4 to 8.5 across ses­sion­s), (c) 8.0 (range 7.8 to 8.2 across ses­sion­s), and (d) 7 (range 6.5 to 7.7 across ses­sion­s). [12 weeks * 2 ses­sions * 30 min­utes = 720 min­utes]

The con­trol group was pas­sive. After the re­cruit­ment process, mem­bers of this no-con­tact con­trol group were in­vited to fol­low their nor­mal life as uni­ver­sity stu­dents. As rea­soned in some of our pre­vi­ous re­search re­ports ad­dress­ing the po­ten­tial effect of cog­ni­tive train­ing, and ac­cord­ing to the main the­o­ret­i­cal frame­work, we were not in­ter­ested in com­par­ing differ­ent types of train­ing, but in the com­par­i­son be­tween a spe­cific cog­ni­tive train­ing and do­ing noth­ing be­yond reg­u­lar life.

Four out of six in­tel­li­gence tests were ap­plied with­out se­vere time con­straints. For the RAPM there was more than one minute per item (20 min­utes for 18 item­s). For DAT-AR DAT-NR and DAT-VR, there were ap­prox­i­mately 30 sec­onds per item (10 min­utes for 20 item­s). For the speeded tests (PMA-R and PMA-V) there were be­tween 5 and 12 sec­onds per item (PMA-R: 3 min­utes for 15 items and PMA-V: 2 min for 25 item­s.)

posttest: train­ing, n=28 RAPM: 37.25 (6.23) con­trol n=28 RAPM: 35.46 (8.26)


Mean differ­ences be­tween the odd and even items were sig­nifi­cant (p < 0.001 for all the tests, ex­clud­ing the DAT-VR) which im­plies that pretest (odd) and posttest (even) scores must not be di­rectly com­pared. –>

Burki et al 2014

“In­di­vid­ual differ­ences in cog­ni­tive plas­tic­i­ty: an in­ves­ti­ga­tion of train­ing curves in younger and older adults”, Burki et al 2014

To date, cog­ni­tive in­ter­ven­tion re­search has pro­vided mixed but nev­er­the­less promis­ing ev­i­dence with re­spect to the effects of cog­ni­tive train­ing on un­trained tasks (trans­fer). How­ev­er, the mech­a­nisms be­hind learn­ing, train­ing effects and their pre­dic­tors are not fully un­der­stood. More­over, in­di­vid­ual differ­ences, which may con­sti­tute an im­por­tant fac­tor im­pact­ing train­ing out­come, are usu­ally ne­glect­ed. We sug­gest in­ves­ti­gat­ing in­di­vid­ual train­ing per­for­mance across train­ing ses­sions in or­der to gain fin­er-grained knowl­edge of train­ing gains, on the one hand, and as­sess­ing the po­ten­tial im­pact of pre­dic­tors such as age and fluid in­tel­li­gence on learn­ing rate, on the other hand. To this aim, we pro­pose to model in­di­vid­ual learn­ing curves to ex­am­ine the in­tra-in­di­vid­ual change in train­ing as well as in­ter-in­di­vid­ual differ­ences in in­tra-in­di­vid­ual change. We rec­om­mend in­tro­duc­ing a la­tent growth curve model (LGCM) analy­sis, a method fre­quently ap­plied to learn­ing data but rarely used in cog­ni­tive train­ing re­search. Such ad­vanced analy­ses of the train­ing phase al­low iden­ti­fy­ing fac­tors to be re­spected when de­sign­ing effec­tive tai­lor-made train­ing in­ter­ven­tions. To il­lus­trate the pro­posed ap­proach, a LGCM analy­sis us­ing data of a 10-day work­ing mem­ory train­ing study in younger and older adults is re­port­ed.

Re­pub­li­ca­tion of a the­sis.

Pugin et al 2014

“Work­ing mem­ory train­ing shows im­me­di­ate and long-term effects on cog­ni­tive per­for­mance in chil­dren and ado­les­cents”, Pu­gin et al 2014:

Work­ing mem­ory is im­por­tant for men­tal rea­son­ing and learn­ing process­es. Sev­eral stud­ies in adults and school-age chil­dren have shown per­for­mance im­prove­ment in cog­ni­tive tests after work­ing mem­ory train­ing. Our aim was to ex­am­ine not only im­me­di­ate but also long-term effects of in­ten­sive work­ing mem­ory train­ing on cog­ni­tive per­for­mance tests in chil­dren and ado­les­cents. Four­teen healthy male sub­jects be­tween 10 and 16 years trained a vi­su­ospa­tial n-back task over 3 weeks (30 min dai­ly), while 15 in­di­vid­u­als of the same age range served as a pas­sive con­trol group. Sig­nifi­cant differ­ences in im­me­di­ate (after 3 weeks of train­ing) and long-term effects (after 2-6 months) in an au­di­tory n-back task were ob­served com­pared to con­trols (2.5 fold im­me­di­ate and 4.7 fold long-term in­crease in the train­ing group com­pared to the con­trol­s). The im­prove­ment was more pro­nounced in sub­jects who im­proved their per­for­mance dur­ing the train­ing. Other cog­ni­tive func­tions (ma­tri­ces test and Stroop task) did not change when com­par­ing the train­ing group to the con­trol group. We con­clude that spa­tial work­ing mem­ory train­ing in chil­dren and ado­les­cents boosts per­for­mance in sim­i­lar mem­ory tasks such as the au­di­tory n-back task. The sus­tained per­for­mance im­prove­ment sev­eral months after the train­ing sup­ports the effec­tive­ness of the train­ing.

Heffernan 2014

“The Gen­er­al­iz­abil­ity of Dual N-Back Train­ing in Younger Adults”, Heffer­nan 2014 (Hal­i­fax, Nova Sco­tia; Canada):

In­tro­duc­tion: The pop­u­lar­ity of cog­ni­tive train­ing has in­creased in re­cent years. Ac­cu­mu­lat­ing ev­i­dence shows that train­ing can some­times im­prove trained and non-trained cog­ni­tive func­tions, and these im­prove­ments may be re­lated to in­di­vid­ual differ­ences in ini­tial ca­pac­ity and per­for­mance on the train­ing task. The cur­rent study as­sessed the effec­tive­ness of a cus­tom-de­signed n-back task (the N-IGMA) ver­sus an ac­tive con­trol task (Block­mas­ter) at im­prov­ing var­i­ous forms of work­ing mem­ory ca­pac­i­ty, at­ten­tion, and fluid in­tel­li­gence. Three mea­sures of work­ing mem­ory ca­pac­ity were con­sid­ered: ver­bal, vi­su­ospa­tial and ob­served ac­tion. Meth­ods: Out­come mea­sures were as­sessed pre- and post-train­ing. Nine­teen healthy young adults (19-30 years of age) trained at-home for 30 min­utes per day, five days a week for three weeks with ei­ther the N-IGMA (n=9) or Block­mas­ter (n=10) at-home games. Re­sults: Pre-post changes were ob­served for some out­come mea­sures and these were equal for the N-IGMA and ac­tive con­trol group. Out­come im­prove­ments could be due to sim­ple test/re-test ben­e­fits or al­ter­na­tively the N-IGMA and Block­mas­ter tasks may pro­duce equiv­a­lent train­ing effects. Im­prove­ments in the train­ing tasks did not cor­re­late with the changes in the out­come mea­sures, sug­gest­ing im­prove­ments in the out­come mea­sures might not be at­trib­ut­able to trans­fer of learn­ing. For ver­bal work­ing mem­ory on­ly, par­tic­i­pants with higher (ver­sus low­er) ini­tial fluid in­tel­li­gence demon­strated larger im­prove­ments on the out­come mea­sures sug­gest­ing that in fu­ture re­search train­ing tasks might need to be tai­lored to the in­di­vid­ual par­tic­i­pant. Pre-assess­ment but not change scores were re­lated for ob­served ac­tion and vi­su­ospa­tial work­ing mem­o­ry, con­sis­tent with some over­lap be­tween con­tent do­mains. Con­clu­sion: De­spite specifi­cally tar­get­ing work­ing mem­o­ry, the N-IGMA was not bet­ter than a vi­su­ospa­tial con­trol game at im­prov­ing a va­ri­ety of cog­ni­tive out­come mea­sures in this small sam­ple. Re­sults sug­gest that the in­di­vid­u­al’s ini­tial cog­ni­tive ca­pac­ity might need to be con­sid­ered in fu­ture train­ing stud­ies. Cau­tion should be used in ex­trap­o­lat­ing the re­sults of this study to other pop­u­la­tions of in­ter­est (e.g., older adults or in­di­vid­u­als with cog­ni­tive deficits) since the present in­ves­ti­ga­tion in­cluded rel­a­tively high func­tion­ing in­di­vid­u­als.

Hancock 2013

“Pro­cess­ing speed and work­ing mem­ory train­ing in Mul­ti­ple Scle­ro­sis: a blinded ran­dom­ized con­trolled trial”, Han­cock 2013

Be­tween 40-65% of pa­tients with mul­ti­ple scle­ro­sis (MS) ex­pe­ri­ence cog­ni­tive deficits as­so­ci­ated with the dis­ease. The two most com­mon ar­eas affected are in­for­ma­tion pro­cess­ing speed and work­ing mem­o­ry. In­for­ma­tion pro­cess­ing speed has been posited as a core cog­ni­tive deficit in MS, and work­ing mem­ory has been shown to im­pact per­for­mance on a wide va­ri­ety of do­mains for MS pa­tients. Cur­rent­ly, clin­i­cians have few re­li­able op­tions for ad­dress­ing cog­ni­tive deficits in MS. The cur­rent study aimed to in­ves­ti­gate the effect of com­put­er­ized, home­-based cog­ni­tive train­ing fo­cused specifi­cally on im­prov­ing in­for­ma­tion pro­cess­ing speed and work­ing mem­ory for MS pa­tients. Par­tic­i­pants were re­cruited and ran­dom­ized into ei­ther the Ac­tive Train­ing or Sham Train­ing group, tested with a neu­rocog­ni­tive bat­tery at base­line, com­pleted six weeks of train­ing, and then were again tested with a neu­rocog­ni­tive bat­tery at fol­low-up. After cor­rect­ing for mul­ti­ple com­par­isons, re­sults in­di­cated that the Ac­tive Train­ing group scored higher on the Paced Au­di­tory Se­r­ial Ad­di­tion Test (a test of in­for­ma­tion pro­cess­ing speed and at­ten­tion) fol­low­ing cog­ni­tive train­ing, and data trended to­ward sig­nifi­cance on the Con­trolled Oral Word As­so­ci­a­tions Task (a test of ex­ec­u­tive func­tion­ing), Let­ter Num­ber Se­quenc­ing (a test of work­ing mem­o­ry), Brief Vi­su­ospa­tial Mem­ory Test (a test of vi­sual mem­o­ry), and the Con­ners’ Con­tin­u­ous Per­for­mance Test (a test of at­ten­tion). Re­sults pro­vide pre­lim­i­nary ev­i­dence that cog­ni­tive train­ing with MS pa­tients may pro­duce mod­er­ate im­prove­ment in se­lect ar­eas of cog­ni­tive func­tion­ing. Fol­low-up stud­ies with larger sam­ples should be con­ducted to de­ter­mine whether these re­sults can be repli­cat­ed, and also to de­ter­mine the func­tional out­come of im­prove­ments on neu­rocog­ni­tive tests.

Waris et al 2015

, Waris et al 2015

Dur­ing the past decade, work­ing mem­ory train­ing has at­tracted much in­ter­est. How­ev­er, the train­ing out­comes have var­ied be­tween stud­ies and method­olog­i­cal prob­lems have ham­pered the in­ter­pre­ta­tion of re­sults. The cur­rent study ex­am­ined trans­fer after work­ing mem­ory up­dat­ing train­ing by em­ploy­ing an ex­ten­sive bat­tery of pre-post cog­ni­tive mea­sures with a fo­cus on near trans­fer. Thir­ty-one healthy Finnish young adults were ran­dom­ized into ei­ther a work­ing mem­ory train­ing group or an ac­tive con­trol group. The work­ing mem­ory train­ing group prac­ticed with three work­ing mem­ory tasks, while the con­trol group trained with three com­mer­cial com­puter games with a low work­ing mem­ory load. The par­tic­i­pants trained thrice a week for five weeks, with one train­ing ses­sion last­ing about 45 min­utes. Com­pared to the con­trol group, the work­ing mem­ory train­ing group showed strongest trans­fer to an n-back task, fol­lowed by work­ing mem­ory up­dat­ing, which in turn was fol­lowed by ac­tive work­ing mem­ory ca­pac­i­ty. Our re­sults sup­port the view that work­ing mem­ory train­ing pro­duces near trans­fer effects, and that the de­gree of trans­fer de­pends on the cog­ni­tive over­lap be­tween the train­ing and trans­fer mea­sures.

Baniqued et al 2015

, Ban­iqued et al 2015:

Al­though some stud­ies have shown that cog­ni­tive train­ing can pro­duce im­prove­ments to un­trained cog­ni­tive do­mains (far trans­fer), many oth­ers fail to show these effects, es­pe­cially when it comes to im­prov­ing fluid in­tel­li­gence. The cur­rent study was de­signed to over­come sev­eral lim­i­ta­tions of pre­vi­ous train­ing stud­ies by in­cor­po­rat­ing train­ing ex­pectancy as­sess­ments, an ac­tive con­trol group, and “Mind Fron­tiers,” a video game-based mo­bile pro­gram com­prised of six adap­tive, cog­ni­tively de­mand­ing train­ing tasks that have been found to lead to in­creased scores in fluid in­tel­li­gence (Gf) tests. We hy­poth­e­size that such in­te­grated train­ing may lead to broad im­prove­ments in cog­ni­tive abil­i­ties by tar­get­ing as­pects of work­ing mem­o­ry, ex­ec­u­tive func­tion, rea­son­ing, and prob­lem solv­ing. Ninety par­tic­i­pants com­pleted 20 hour-and-a-half long train­ing ses­sions over four to five weeks, 45 of whom played Mind Fron­tiers and 45 of whom com­pleted vi­sual search and change de­tec­tion tasks (ac­tive con­trol). After train­ing, the Mind Fron­tiers group im­proved in work­ing mem­ory n-back tests, a com­pos­ite mea­sure of per­cep­tual speed, and a com­pos­ite mea­sure of re­ac­tion time in rea­son­ing tests. No train­ing-re­lated im­prove­ments were found in rea­son­ing ac­cu­racy or other work­ing mem­ory tests, nor in com­pos­ite mea­sures of episodic mem­o­ry, se­lec­tive at­ten­tion, di­vided at­ten­tion, and mul­ti­-task­ing. Per­ceived self­-im­prove­ment in the tested abil­i­ties did not differ be­tween groups. A gen­eral ex­pectancy differ­ence in prob­lem-solv­ing was ob­served be­tween groups, but this per­ceived ben­e­fit did not cor­re­late with train­ing-re­lated im­prove­ment. In sum­ma­ry, al­though these find­ings pro­vide mod­est ev­i­dence re­gard­ing the effi­cacy of an in­te­grated cog­ni­tive train­ing pro­gram, more re­search is needed to de­ter­mine the util­ity of Mind Fron­tiers as a cog­ni­tive train­ing tool.

Kuper & Karbach 2015

“In­creased train­ing com­plex­ity re­duces the effec­tive­ness of brief work­ing mem­ory train­ing: ev­i­dence from short­-term sin­gle and dual n-back train­ing in­ter­ven­tions”, Ku­per & Kar­bach 2015:

N-back train­ing has re­cently come un­der in­tense sci­en­tific scrutiny due to re­ports of train­ing-re­lated im­prove­ments in gen­eral fluid in­tel­li­gence. As of yet, rel­a­tively lit­tle is known about the effects of short­-term n-back train­ing in­ter­ven­tions, how­ev­er. In a pretest-train­ing-posttest de­sign, we com­pared brief dual and sin­gle n-back train­ing reg­i­men in terms of train­ing gains and trans­fer effects rel­a­tive to a pas­sive con­trol group. Trans­fer effects in­di­cated that, in the short­-term, sin­gle n-back train­ing may be the more effec­tive train­ing task: At the short train­ing du­ra­tion we em­ployed, train­ing group showed far trans­fer to spe­cific task switch costs, Stroop in­hi­bi­tion or ma­trix rea­son­ing in­dex­ing fluid in­tel­li­gence. Yet, both types of train­ing re­sulted in a re­duc­tion of gen­eral task switch costs in­di­cat­ing im­proved cog­ni­tive con­trol dur­ing the sus­tained main­te­nance of com­pet­ing task sets. Sin­gle but not dual n-back train­ing ad­di­tion­ally yielded near trans­fer to an un­trained work­ing mem­ory up­dat­ing task.

Lindeløv et al 2016

“Train­ing and trans­fer effects of N-back train­ing for brain-in­jured and healthy sub­jects”, Lin­deløv et al 2016:

Work­ing mem­ory im­pair­ments are preva­lent among pa­tients with ac­quired brain in­jury (ABI). Com­put­erised train­ing tar­get­ing work­ing mem­ory has been re­searched ex­ten­sively us­ing sam­ples from healthy pop­u­la­tions but this field re­mains iso­lated from sim­i­lar re­search in ABI pa­tients. We re­port the re­sults of an ac­tively con­trolled ran­domised con­trolled trial in which 17 pa­tients and 18 healthy sub­jects com­pleted train­ing on an N-back task. The healthy group had su­pe­rior im­prove­ments on both train­ing tasks (SMD = 6.1 and 3.3) whereas the ABI group im­proved much less (SMD = 0.5 and 1.1). Nei­ther group demon­strated trans­fer to un­trained tasks. We con­clude that com­put­erised train­ing fa­cil­i­tates im­prove­ment of spe­cific skills rather than high­-level cog­ni­tion in healthy and ABI sub­jects alike. The ac­qui­si­tion of these spe­cific skills seems to be im­paired by brain in­jury. The most effec­tive use of com­put­er-based cog­ni­tive train­ing may be to make the task re­sem­ble the tar­geted be­hav­iour(s) closely in or­der to ex­ploit the stim­u­lus-speci­ficity of learn­ing.

Schwarb et al 2015

“Work­ing mem­ory train­ing im­proves vi­sual short­-term mem­ory ca­pac­ity”, Schwarb et al 2015

Lawlor-Savage & Goghari 2016

, Lawlor-Sav­age & Goghari 2016:

En­hanc­ing cog­ni­tive abil­ity is an at­trac­tive con­cept, par­tic­u­larly for mid­dle-aged adults in­ter­ested in main­tain­ing cog­ni­tive func­tion­ing and pre­vent­ing age-re­lated de­clines. Com­put­er­ized work­ing mem­ory train­ing has been in­ves­ti­gated as a safe method of cog­ni­tive en­hance­ment in younger and older adults, al­though few stud­ies have con­sid­ered the po­ten­tial im­pact of work­ing mem­ory train­ing on mid­dle-aged adults. This study in­ves­ti­gated dual n-back work­ing mem­ory train­ing in healthy adults aged 30-60. Fifty-seven adults com­pleted mea­sures of work­ing mem­o­ry, pro­cess­ing speed, and fluid in­tel­li­gence be­fore and after a 5-week we­b-based dual n-back or ac­tive con­trol (pro­cess­ing speed) train­ing pro­gram. Re­sults: Re­peated mea­sures mul­ti­vari­ate analy­sis of vari­ance failed to iden­tify im­prove­ments across the three cog­ni­tive com­pos­ites, work­ing mem­o­ry, pro­cess­ing speed, and fluid in­tel­li­gence, after train­ing. Fol­low-up Bayesian analy­ses sup­ported null find­ings for train­ing effects for each in­di­vid­ual com­pos­ite. Find­ings sug­gest that dual n-back work­ing mem­ory train­ing may not ben­e­fit work­ing mem­ory or fluid in­tel­li­gence in healthy adults. Fur­ther in­ves­ti­ga­tion is nec­es­sary to clar­ify if other forms of work­ing mem­ory train­ing may be ben­e­fi­cial, and what fac­tors im­pact train­ing-re­lated ben­e­fits, should they oc­cur, in this pop­u­la­tion.

Studer-Luethi et al 2015

“Work­ing mem­ory train­ing in chil­dren: Effec­tive­ness de­pends on tem­pera­ment”, Stud­er-Luethi et al 2015:

Stud­ies re­veal­ing trans­fer effects of work­ing mem­ory (WM) train­ing on non-trained cog­ni­tive per­for­mance of chil­dren hold promis­ing im­pli­ca­tions for scholas­tic learn­ing. How­ev­er, the re­sults of ex­ist­ing train­ing stud­ies are not con­sis­tent and pro­voke de­bates about the po­ten­tial and lim­i­ta­tions of cog­ni­tive en­hance­ment. To ex­am­ine the in­flu­ence of in­di­vid­ual differ­ences on train­ing out­comes is a promis­ing ap­proach for find­ing causes for such in­con­sis­ten­cies. In this study, we im­ple­mented WM train­ing in an el­e­men­tary school set­ting. The aim was to in­ves­ti­gate near and far trans­fer effects on cog­ni­tive abil­i­ties and aca­d­e­mic achieve­ment and to ex­am­ine the mod­er­at­ing effects of a dis­po­si­tional and a reg­u­la­tive tem­pera­ment fac­tor, neu­roti­cism and effort­ful con­trol. Nine­ty-nine sec­ond-graders were ran­domly as­signed to 20 ses­sions of com­put­er-based adap­tive WM train­ing, com­put­er-based read­ing train­ing, or a no-con­tact con­trol group. For the WM train­ing group, our analy­ses re­veal near trans­fer on a vi­sual WM task, far trans­fer on a vo­cab­u­lary task as a proxy for crys­tal­lized in­tel­li­gence, and in­creased aca­d­e­mic achieve­ment in read­ing and math by trend. Con­sid­er­ing in­di­vid­ual differ­ences in tem­pera­ment, we found that effort­ful con­trol pre­dicts larger train­ing mean and gain scores and that there is a mod­er­a­tion effect of both tem­pera­ment fac­tors on post-train­ing im­prove­ment: WM train­ing con­di­tion pre­dicted higher post-train­ing gains com­pared to both con­trol con­di­tions only in chil­dren with high effort­ful con­trol or low neu­roti­cism. Our re­sults sug­gest that a short but in­ten­sive WM train­ing pro­gram can en­hance cog­ni­tive abil­i­ties in chil­dren, but that suffi­cient self­-reg­u­la­tive abil­i­ties and emo­tional sta­bil­ity are nec­es­sary for WM train­ing to be effec­tive….We found no sig­nifi­cant train­ing group in­ter­ac­tion on the per­for­mance in the Raven’s Pro­gres­sive Ma­tri­ces (F(2,92) = 1.57, p = .22, ηp2 = .004)…We found no sig­nifi­cant long-term effects in the vari­ables mem­ory span, cog­ni­tive con­trol, Gf, Gc, and scholas­tic tests (all T < 1.4).

Minear et al 2016

“A si­mul­ta­ne­ous ex­am­i­na­tion of two forms of work­ing mem­ory train­ing: Ev­i­dence for near trans­fer only”, Min­ear et al 2016

The effi­cacy of work­ing-mem­ory train­ing is a topic of con­sid­er­able de­bate, with some stud­ies show­ing trans­fer to mea­sures such as fluid in­tel­li­gence while oth­ers have not. We re­port the re­sults of a study de­signed to ex­am­ine two forms of work­ing-mem­ory train­ing, one us­ing a spa­tial n-back and the other a ver­bal com­plex span. Thir­ty-one un­der­grad­u­ates com­pleted 4 weeks of n-back train­ing and 32 com­pleted 4 weeks of ver­bal com­plex span train­ing. We also in­cluded two ac­tive con­trol groups. One group trained on a non-adap­tive ver­sion of n-back and the other trained on a re­al-time strat­egy video game. All par­tic­i­pants com­pleted pre- and post-train­ing mea­sures of a large bat­tery of trans­fer tasks used to cre­ate com­pos­ite mea­sures of short­-term and work­ing mem­ory in both ver­bal and vi­su­ospa­tial do­mains as well as ver­bal rea­son­ing and fluid in­tel­li­gence. We only found clear ev­i­dence for near trans­fer from the spa­tial n-back train­ing to new forms of n-back, and this was the case for both adap­tive and non-adap­tive n-back.

Studer-Luethi et al 2016

“Work­ing mem­ory train­ing in chil­dren: Effec­tive­ness de­pends on tem­pera­ment”, Stud­er-Luethi et al 2016:

Stud­ies re­veal­ing trans­fer effects of work­ing mem­ory (WM) train­ing on non-trained cog­ni­tive per­for­mance of chil­dren hold promis­ing im­pli­ca­tions for scholas­tic learn­ing. How­ev­er, the re­sults of ex­ist­ing train­ing stud­ies are not con­sis­tent and pro­voke de­bates about the po­ten­tial and lim­i­ta­tions of cog­ni­tive en­hance­ment. To ex­am­ine the in­flu­ence of in­di­vid­ual differ­ences on train­ing out­comes is a promis­ing ap­proach for find­ing causes for such in­con­sis­ten­cies. In this study, we im­ple­mented WM train­ing in an el­e­men­tary school set­ting. The aim was to in­ves­ti­gate near and far trans­fer effects on cog­ni­tive abil­i­ties and aca­d­e­mic achieve­ment and to ex­am­ine the mod­er­at­ing effects of a dis­po­si­tional and a reg­u­la­tive tem­pera­ment fac­tor, neu­roti­cism and effort­ful con­trol. Nine­ty-nine sec­ond-graders were ran­domly as­signed to 20 ses­sions of com­put­er-based adap­tive WM train­ing, com­put­er-based read­ing train­ing, or a no-con­tact con­trol group. For the WM train­ing group, our analy­ses re­veal near trans­fer on a vi­sual WM task, far trans­fer on a vo­cab­u­lary task as a proxy for crys­tal­lized in­tel­li­gence, and in­creased aca­d­e­mic achieve­ment in read­ing and math by trend. Con­sid­er­ing in­di­vid­ual differ­ences in tem­pera­ment, we found that effort­ful con­trol pre­dicts larger train­ing mean and gain scores and that there is a mod­er­a­tion effect of both tem­pera­ment fac­tors on post-train­ing im­prove­ment: WM train­ing con­di­tion pre­dicted higher post-train­ing gains com­pared to both con­trol con­di­tions only in chil­dren with high effort­ful con­trol or low neu­roti­cism. Our re­sults sug­gest that a short but in­ten­sive WM train­ing pro­gram can en­hance cog­ni­tive abil­i­ties in chil­dren, but that suffi­cient self­-reg­u­la­tive abil­i­ties and emo­tional sta­bil­ity are nec­es­sary for WM train­ing to be effec­tive.

…Fluid in­tel­li­gence was as­sessed us­ing ei­ther the even or odd items of Raven’s Pro­gres­sive Ma­tri­ces in coun­ter­bal­anced or­der (RPM, 30 items; Raven, 1998). After two prac­tice tri­als, chil­dren were al­lowed to work for 10 min and cross the right so­lu­tion for each task. The num­ber of cor­rect so­lu­tions pro­vided in this time limit was used as the de­pen­dent vari­able.

…We found no sig­nifi­cant train­ing group in­ter­ac­tion on the per­for­mance in the Raven’s Pro­gres­sive Ma­tri­ces (F(2,92) = 1.57, p = .22, ηp2 = .004).


I con­struct a meta-analy­sis of the >19 stud­ies which mea­sure IQ after an n-back in­ter­ven­tion, con­firm­ing that there is a gain of smal­l­-to-medium effect size. I also in­ves­ti­gate sev­eral n-back claims, crit­i­cisms, and in­di­ca­tors of bi­as, find­ing:

Due to its length and tech­ni­cal de­tail, my meta-analy­sis has been moved to .

Does it really work?

N-back improves working memory

There are quite a few stud­ies show­ing sig­nifi­cant in­creases in work­ing mem­o­ry: WM is some­thing that can be trained. See for ex­am­ple “Changes in cor­ti­cal ac­tiv­ity after train­ing of work­ing mem­ory - a sin­gle-sub­ject analy­sis.” or “In­creased pre­frontal and pari­etal ac­tiv­ity after train­ing of work­ing mem­ory”.

There are a few stud­ies show­ing that DNB train­ing en­hances Gf; see the sup­port sec­tion. There is also a study show­ing that WM train­ing (not DNB) en­hances Gc68.

IQ Tests


Be­cause N-back is sup­posed to im­prove your pure ‘’ (Gf), and not, say, your Eng­lish vo­cab­u­lary, the most ac­cu­rate tests for see­ing whether N-back has done any­thing are go­ing to be ones that avoid vo­cab­u­lary or lit­er­a­ture or tests of sub­jec­t-area knowl­edge. That is, ‘cul­ture-neu­tral’ IQ tests. (A non-neu­tral test fo­cuses more on your ‘crys­tal­lized in­tel­li­gence’, while N-back is sup­posed to affect ‘fluid in­tel­li­gence’; they do affect each other a lit­tle but it’s bet­ter to test fluid in­tel­li­gence with a fluid in­tel­li­gence test.)

As one ML mem­ber writes:

The WAIS test in­volves crys­tal­lized in­tel­li­gence and is un­suit­able for judg­ing fluid in­tel­li­gence. High work­ing mem­ory will not spawn the abil­ity to solve com­plex math­e­mat­i­cal and ver­bal prob­lems on its own, you have to put your ex­tended ca­pac­ity to learn­ing. All very-high­-level IQ tests are largely crys­tal­lized IQ tests, there­fore work­ing mem­ory gains will not be im­me­di­ately ap­par­ent by their mea­sure.ref

Available tests

The gold-s­tan­dard of cul­ture-neu­tral IQ tests is . Un­for­tu­nate­ly, Raven’s is not avail­able for free on­line, but there are a num­ber of clones one can use - bear­ing in mind their likely in­ac­cu­racy and the fact that many of them do not ran­dom­ize their ques­tions. It’s a very good idea, if you plan to n-back for a long time, to take an IQ test be­fore and an IQ test after, both to find out whether you im­proved and so you can tell the rest of us. But the in­ter­val has to be a long one: if you are test­ing at the be­gin­ning and end of your train­ing there is prob­a­bly go­ing to be a which will dis­tort your sec­ond score up­wards69; it’s strongly rec­om­mended you take a par­tic­u­lar test only on­ce, or with in­ter­vals on the or­der of months (and prefer­ably years).

The tests are:

Raven-style ma­trix tests can be me­chan­i­cally gen­er­ated by the San­dia Gen­er­ated Ma­trix Tool; the gen­er­ated ma­trix test scores sta­tis­ti­cally look very sim­i­lar to SPM test scores ac­cord­ing to the pa­per, “Recre­at­ing Raven’s: Soft­ware for sys­tem­at­i­cally gen­er­at­ing large num­bers of Raven-like ma­trix prob­lems with normed prop­er­ties”.

If Raven-style tests bore you or you’ve gone through the pre­vi­ous ones, there are a wealth of diffi­cult tests at Miyaguchi’s “Un­com­monly Diffi­cult IQ Tests”, and a per­son­al­ity test, the IPIP-NEO, is free (although the con­nec­tion to IQ is min­i­mal).

Other tests that might be use­ful in­clude tests: they pro­vide a non-d­u­al-N-back method of mea­sur­ing WM be­fore one be­gins train­ing and then after. There is also Cogtest suite of spans and at­ten­tion tasks or the http://cog­ni­tive­fun.net/ site (which im­ple­ments many tasks). The Au­to­mated Op­er­a­tion Span (OSPAN) Task could be used as well.

IQ test results

Re­ports of IQ tests have been mixed. Some re­sults have been stun­ning, oth­ers have shown noth­ing.


LSaul posted about his ap­par­ent rise in IQ back in Oc­to­ber. From what I re­mem­ber, he had re­cently failed to qual­ify for MENSA, which re­quires a score of about 131 (98th per­centile). He then got a 151 (99.97th per­centile) on a pro­fes­sion­ally ad­min­is­tered IQ test (WAIS) three months lat­er, after 2 months of reg­u­lar du­al-n-back use. –MR

(A >20 point gain sounds very im­pres­sive. But pos­si­ble con­found­ing fac­tors here are that LSaul ap­par­ently took 2 differ­ent IQ tests; be­sides the gen­eral in­com­pa­ra­bil­ity of differ­ent IQ tests, it sounds as if the first test was a cul­ture-neu­tral one, while the WAIS has com­po­nents such as ver­bal tests - the sec­ond might well be ‘eas­ier’ for LSaul than the first.)

Mike L. writes:

Em­pir­i­cally speak­ing, how­ev­er: I took a WAIS-IV IQ test (ad­min­is­tered pro­fes­sion­al­ly) around a year ago and got a 110. I took a de­riv­a­tive of the same test re­cently (mind you, after about 20 days of DNB train­ing) and got a score of 121.

The blog­ger of “In­hu­man Ex­per­i­ment”, who played for ~22 days and went from ~2.6-back to ~4-back, re­ports:

The other test proved to be quite good (you can find it here). In this one, the ques­tions vary, the diffi­culty is ad­justed on the go de­pend­ing on whether you an­swer them cor­rect­ly, and there’s a time limit of 45 sec­onds per ques­tion, which makes this test bet­ter suited for re-tak­ing. My first test, taken be­fore play­ing the game, gave me a score of 126; my sec­ond test, taken yes­ter­day, gave me a score of 132 (an in­crease of about 5%)….As you can see, it’s kind of diffi­cult to draw any mean­ing­ful con­clu­sions from this. Yes, there was a slight in­crease in my score, but I would say a sim­i­lar in­crease could’ve been pos­si­ble even with­out play­ing the game. I think the vari­a­tion in the IQ test ques­tions re­duces the “learn­ing by heart” effect, but that’s im­pos­si­ble to say with­out a con­trol group.

Pon­tus Granström writes that

I scored 133 on www.­men­sa.d­k/iqtest.swf to­day. I have never scored that high be­fore I re­ally feel the “DNB think­ing” kick­ing in.

(He ap­par­ently took that test about a year ago, and avers that his orig­i­nal score on it ‘was 122. Well be­low 130.’)

Pheonexia writes:

Ap­prox­i­mately three years ago I took the “Eu­ro­pean IQ Test.” It was posted on some mes­sage board and the au­thor of the thread said the test was cred­i­ble. At that time, I scored 126.

I’ve been n-back­ing since early Feb­ru­ary, so I fig­ured I’d try it again to­day. I googled “Eu­ro­pean IQ Test” and clicked the first re­sult, a test from Nanyang Tech­no­log­i­cal Uni­ver­sity in Sin­ga­pore.. I don’t re­call any of the ex­act ques­tions for the first one I took three years ago, but the for­mat of this test seemed al­most iden­ti­cal. To­day I scored 144, 18 points higher than be­fore. http://www3.n­tu.e­du.s­g/home­/cz­zhao/iq/test.htm

To me, this is anec­do­tal ev­i­dence that n-back­ing does in­crease in­tel­li­gence. I’ll try again for an­other three months and take a com­pletely differ­ent test. I will ad­mit, how­ev­er, that I rec­og­nized one of the first ques­tions as the Fi­bonacci se­quence, so I at­tribute that to crys­tal­lized, not fluid in­tel­li­gence. The high­est score this test al­lows for is 171, mean­ing you got ZERO ques­tions wrong. I got 6 wrong and 3 half ques­tions wrong where it re­quires two an­swers (that was my worst sec­tion), so ei­ther 7.5 or 9 out of 33 ques­tions wrong.


I took one of the IQ tests I did pre­vi­ously [pre­vi­ously linked as “High IQ So­ci­ety On­line Test”] and scored 109 on, I just took it again and scored 116…I don’t know about retest effect, but all the ques­tions were differ­ent.

Toto writes in TNB(PIA) may im­prove in­tel­li­gence”:

While DNB proved in­effec­tive for me (at least it did­n’t in­crease my IQ, though it im­proved mem­o­ry) TNB may have made a differ­ence. I took 2 high­-range tests dur­ing the last 2 months and the re­sults were higher than I ex­pected - my IQ was some­where be­tween 130 and 135 on good on­line tests, I scored 132 on a su­per­vised test (Raven’s SM). My re­sults on CFNSE http://www.e­ti­enne.se/cfnse/ and GEThttp://www.epiq­so­ci­ety.net/get/ were ap­prox­i­mately 10 points higher - 6 on CFNSE (8 on my sec­ond at­tempt) and 21 on G.E.T . It could be be­cause of a flaw of these tests, or they may not test the same abil­ity as timed tests (though the cor­re­la­tion be­tween them and fa­mous su­per­vised timed tests is said to be very high), it may be for some other rea­son as well, but it could be be­cause of TNB. I had tried CFNSE long ago and scored 0 (but I prob­a­bly did­n’t try hard enough then).

christo­pher lines re­ports:

I did a cou­ple of the on­line IQ tests after about 10 days (s­cored 126 in one of them [iqtest.dk] and 106 [iqout.­com] in an­oth­er); I re­peated the same tests about a month later (about 1 month ago) and scored (133 and 109). I have no idea why the tests gave such big differ­ences in scores but I de­fi­nately [sic] think its eas­ier the sec­ond time you do the tests be­cause I re­mem­bered the strate­gies for solv­ing the prob­lems which took some time to fig­ure out when I first did the tests. I am kind of against keep re-do­ing the tests be­cause of learn­ing effects and a bit truobled [sic] that differ­ent test pro­duce such differ­ent re­sults.

Tofu writes:

I’ve pur­posely not been do­ing any­thing to prac­tice for the tests or any­thing else I thought could in­crease my score so I would­n’t have to fac­tor other things into an im­prove­ment in iq, which makes im­prove­ments more likely at­trib­ut­able to dual n-back. Be­fore I took the test I scored at 117, a score about 1 in about 8 peo­ple can get (7.78 to be ex­ac­t), and yes­ter­day I scored at 127 (a score that 1 in 28 peo­ple would get). Its a pretty big differ­ence I would say.

After a year of N-back­ing, Tofu has 3 sets of IQ test re­sults us­ing Self­-S­cor­ing IQ Tests (). To sum­ma­rize:

  1. 0 months (D3B; ~71%): 27,25 = 117 IQ
  2. 3 (T4B; ~76%): 37,4070 = 128
  3. 12 (Q6B; ~47%): 34,42 = 128

Other rel­e­vant tests for To­fu:

As a side­note- after 6 months I took a prac­tice with­out any study­ing and got a 146, roughly 30th per­centile, and I took an IQ test from http://iqtest.d­k/­main.swf after 1 year which I scored a 115 on. Al­so, in high school I took a pro­fes­sion­ally ad­min­is­tered IQ test and got a 137 which may have been high be­cause they took my age into ac­count in the scor­ing like the old school IQ tests used to do, but I’m not sure if they ac­tu­ally did that.


Last year I scored 123 in www.iqtest.dk and to­day I made 140. If you elim­i­nate sta­tis­tic de­vi­a­tions, even if it’s just 5-10 points it’s very good IMO.


I do ac­tu­ally have gains to re­port on the “Ad­vanced Cul­ture Fair Test” found on iq­com­par­ison­site.­com that I just took to­day. Facts: I scored 29 raw (out of 36) IQ 146 or 99.9%ile, com­pared to my 130 or 98%ile raw 21 that I scored when I took the test over a year ago.

…For com­par­i­son to other fluid mea­sures, this re­sult is 3 points higher than my Get-gamma score and 2 points higher than my GIGI cer­ti­fied and 13 points higher than my iqtestdk re­sult which lands in the same place every time I take it (last time I took it was less than a month ago). My cur­rent DNB level av­er­ages 8+ over mul­ti­ple (10-20) ses­sions.

mile­stones later re­ported:

Last night I re­took the iqtest.org.uk and scored higher on a sec­ond try than I did a few months ago—145 up from 133. This could be due to 1. con­sis­tent quad back prac­tice 2. be­ing back on cre­a­tine as I have been for the last month 3. Omega 3/epa/­fish oil 4. just a nor­mal swing in scores due to other fac­tors, in­clud­ing fa­mil­iar­ity with the items. Or, of course, maybe some com­bi­na­tion of the above

Lach­lan Jones:

Hey guys I’ve been us­ing brain work­shop (D­ual N back) for about 2 months now and would like to re­port an in­crease in IQ from 124 to 132 (on pro­fes­sion­ally ad­min­is­tered IQ tests that were su­per­vised) The IQ tests were sep­a­rated by a pe­riod of about a year as well.

(It’s a lit­tle un­clear whether this was an im­prove­ment or not; the sec­ond score was on the test, but Lach­lan has­n’t said what the first one was. Since differ­ent tests have differ­ent norms and what not, Lach­lan’s scores could ac­tu­ally be the same or de­clin­ing.)

Jan Pouchlý:

2008/06 DNB for cca 1 mon­th, 1 - 2 hours a day, 5 times a week; after 2 weeks prob­a­bly gain +8 points IQ (I think it was Wech­sler IQ 140 ad­min­is­tered by school psy­chol­o­gist and after 5 months Raven IQ 148 ad­min­is­tered by some Mensa guy). No prob­lems at all. Bet­ter dream re­call.

Ar­gumzio com­ments on Jan:

The differ­ence be­tween (what I as­sume both are) WAIS-III and RAPM is fairly sig­nifi­cant; the for­mer is about 2.667 sigma (FSIQ), and the lat­ter is just over 3 sig­ma. For those who wish to know, both are set with a stan­dard de­vi­a­tion of 15.

Keep in mind, how­ev­er, that with WAIS-III you get the full treat­ment while with RAPM your fluid abil­ity is as­sessed as in the orig­i­nal Jaeggi study, so Jan’s per­for­mance on other fac­tors may have de­pressed and con­cealed his (al­ready) high Gf, or per­for­mance, ca­pa­bil­i­ties. That’s why it is para­mount to use the same test, or a test that is es­sen­tially of the same de­sign.

Mug­gin­Buns (gains may’ve been from prac­tice, or from a ‘fea­ture se­lec­tion’-based game Mug­gin­Buns is de­vel­op­ing):

"http://www.iqtest.d­k/­main.swf - 126 3 months ago

http­s://www.gi­giassess­men­t.­com/shop/in­dex.php - 126 3 months ago

http://www.iqtest.d­k/­main.swf - 140 2 weeks ago

http://www.cere­bral­s.org/w­p/?­page_id=44 - 137 yes­ter­day"

Min Mae:

Pre­vi­ously I had a RAPM IQ test re­sult was 112 by cer­ti­fied psy­chol­o­gist. In 2009 Au­gust I prac­ticed DNB 2.5 hours a day for 20 days with time off two days, Sat­ur­day and Sun­day. After 20 days I took RAPM test in my uni­ver­sity by cer­ti­fied psy­chol­o­gist. I got IQ gain was 12.1 points. In the test i was only able to an­swer more ques­tions that re­lated to changes in po­si­tion of ob­jects in the test (RAPM).

At 2010/04/12 I started SNB(single N-Back, vi­sual modal­i­ty) with train­ing time was same as on DNB train­ing time for 20 days. IQ gain by RAPM test was no change.That time i also was able to an­swer more ques­tions that re­lated to changes in po­si­tion of ob­jects in the RAPM test.

Colin Dick­er­man in the thread “IQ test in one mon­th!”:

I took a free an­droid IQ test (I’m com­put­er­less) and scored 123 about 4 months ago. I’ve started n-back­ing again and after 4 weeks of con­sis­tent effort, I av­er­age around 70% at dual 5 back­…Okay, I jumped the gun a lit­tle bit and re­took the test a day ear­ly. I scored 126…[My] N-back level stag­nated in the mid 60% at 6-back.


Ok, so a few months ago the most i could get on the http://­men­sa.d­k/iqtest/ was about 95,115-126. Well now its 136….with the stan­dard de­vi­a­tion 24 of course…I got to ad­mit that score was on one of my bad days, and I was­n’t re­ally fo­cused, plus I did­n’t spend much time on the ques­tions. Prob­a­bly 2 months ago the high­est was 126.


Be­fore n-back­ing, my IQ lay in the re­gion be­tween 109 and 120 (most on­line tests al­ways put me in the 113-120 range, but the MENSA test only gave me a re­sult of 109). I’ve prob­a­bly com­pleted 10 IQ tests over the last 3 years and my scores seem to be rel­a­tively con­sis­ten­t….­So, I’ve spent about one and a half months on du­al-n-back. I did the http://www.iqtest.d­k/­main.swf test and got an IQ of 123.


iqtest.dk (Eng­lish) I first at­tempted this test more than 2 years ago where I ob­tained a score of be­tween 110-115. I at­tempted this test again to­day where I achieved a score of 138.

N-level: Well, as most of you may know, over­time I’ve very much just been rolling around in the mud of what BW has to offer, so be­cause I haven’t stayed in one coun­try long enough to call it home, it’s pretty much im­pos­si­ble for me to at­tribute my new ‘world view’ to one par­tic­u­lar mode or another…DNB: 4-back - 8-back = Time taken to reach lev­el, 10+ months Quad-n-back: 2-back - 6-back = Time taken to reach lev­el, be­tween 6-8 months Vari­able-arith­metic n-back: 3-back - 7-back = Time taken to reach lev­el, be­tween 3-4 months… [de­scrip­tion of daily rou­tine]


I’ve done about 40 half-hour ses­sions of dual n back and have made gains within the task-ie higher n-back score. Per­son­ally I don’t feel much smarter but I’ve no­ticed I read faster and can com­pre­hend what I am read­ing at a faster speed as well. Pre­vi­ously I scored a 109, then after 40 ses­sions I scored at 122 on the Den­mark Mensa IQ test. http://www.iqtest.d­k/­main.swf. My con­cern is that this sup­posed gain has not made a not­i­ca­ble im­prove­ment in my re­al-world in­tel­li­gence, and that the Den­mark IQ test is un­re­li­able…I re­cently took a Mensa puz­zle brain-teaser and scored a 18/30, which seems fairly medioc­re? I don’t know..I was pretty stumped by some of the ques­tions. Did­n’t make me feel too smart….Up­date: I took the same IQ test(­Den­mark Men­sa) again and I scored a 126, 4 points higher than my pre­vi­ous score of 122. Be­tween tak­ing the tests I had prac­ticed dual n back for about 14 half hour ses­sions.


I liked IQ tests, es­pe­cially the iqtest.dk. I did it for the last time be­tween 1-2 years ago. My score was 110, I’m pretty sure. I scored never higher than 110, but also not much lower than 110. Guess i had just av­er­age in­tel­li­gence and i was feel­ing that way too. On the Wech­sler test i did 3-4 years ago i scored 107. so if i for­mu­late it cor­rectly my fluid in­tel­li­gence was in line with my gen­eral in­tel­li­gence. Now i do it for 11 con­se­quent days, just 25 min­utes a day and mostly at this point 2-3-4 back. But after the third day i felt much more clar­ity and bet­ter abil­ity to for­mu­late things, be­cause my mem­ory seemed so much bet­ter. Take note that I’m re­ally an ex­tremely sen­si­tive per­son, so that is prob­a­bly the rea­son i felt it so quick­ly. To­day i de­cided to do the iqtest.dk test again, be­cause i was ex­cited to do it and not wait till the 19th day. My ex­pec­ta­tion was a iq in the 110-115 range, but guess what i scored 126…A few min­utes ago i ended the iqtest.dk again and scored 122. This means for me I’m ap­prox­i­mately as smart as 12 days be­fore, fur­ther­more i think i should­n’t be any differ­ences in raw in­tel­li­gence.

He later re­ported ad­di­tional re­sults:

i did the full wais-iii 12-13 months ago. i scored 111 on the POI, which i think is the best mea­sure for gf (although not a pure mea­sure, but more com­pre­hen­sive than just ma­tri­ces) This where the scores within the POI:

  • Pic­ture com­ple­tion 11
  • Block De­sign 10
  • Ma­trix Rea­son­ing 15

I trained 2.5 months from Feb­ru­ary to April 2012. Note that i am 21 years old (in­tel­li­gence is in some de­gree mal­leable till 22/23 years old, right?) Well, i did the wais-iii again and have the re­sults since a week. My POI is now 125 and this is how it looks:

  • Pic­ture com­ple­tion 8 (-3)
  • Block De­sign 18 (+8)
  • Ma­trix rea­son­ing 16 (+1)

karthik bm:

I started dual n back­ing about 5months ago.After train­ing for first 2 months, I took an IQ test and Iqtest.dk. My score was in low 120s.(­took the test mul­ti­ple times and got al­most the same score each and every time)

For the next 3 months apart from n back­ing, I in­cluded med­i­ta­tion, im­age stream­ing and jug­gling into my sched­ule. Yes­ter­day I took the same test at iqtest.dk and got the score as 133.


I first started play­ing DNB some 3 years ago, try­ing to play three rounds every­day (skipped at most 20% of the to­tal days), re­gard­less of the n-back lev­el.

  • 2 years ago (high­est con­sis­tent DNB: 4), I took my first MENSA test - and scored 130 (SD 24), top 10%.
  • 2 weeks ago (high­est con­sis­tent DNB: 7), I took my sec­ond MENSA test - and scored 156 (SD 24), top 1%, so I am join­ing.


I’ve been do­ing n-back since June of 2011 and I’m av­er­ag­ing now be­tween 8.0 and 8.25. I lay off for as long as six months at a time but get back to where I ended just two days after re­sum­ing. Be­fore I started I toke iqtest.dk and scored 105. After six months and reg­u­larly scor­ing above 6 I re­took the same test and scored 115. After one year and scor­ing above 7 I scored 127 IQ points. I re­took the test just now and I scored 115.

No improvement

Some have not:

I took the On­line Den­mark IQ test again [after N-back train­ing] and I got 140 (the same re­sult) I took a stan­dard­ized (and charged) on­line IQ test from www.iqtest.­com and I got 134 (though it may be a bit higher be­cause Eng­lish is not my mother tongue) –Crypto

jt­toto re­ports a null re­sult:

6 months ago I posted my IQ on this site after tak­ing the Mensa Nor­way test… [see IQ tests sec­tion] I scored a 135. After 6 months of dual n-back, triple n-back, and quad n-back train­ing, I took the same ex­act test. I scored ex­actly the same, 135. Grant­ed, I took 7 less min­utes to com­plete the test, but this was due to fa­mil­iar­ity of some of the ques­tions. That be­ing said, I have been see­ing sig­nifi­cant in­creases in my digit span and other WM gains, so while my ap­ti­tude on ques­tions like the Raven’s may not have in­creased, my mem­ory has.

(It’s worth not­ing that Jt­to­to’s ex­pe­ri­ence does­n’t rule out an IQ in­crease of some sort, as the orig­i­nal 135 score was from an IQ test he took after at least 10 hours of n-back­ing over 5 days, ac­cord­ing to an ear­lier email; what it shows is that Jt­toto did­n’t ben­e­fit or the ben­e­fits hap­pened early on, or there’s some other con­found­ing fac­tor. Test re­sults can be very diffi­cult to in­ter­pret.)

moe writes:

“After 6 months of train­ing I de­cided to take the tri 52 again and there has been no im­prove­ment in in­tel­li­gence (or should I say ab­stract rea­son­ing abil­i­ty), I’m still at 144 sd15 on that test. My digit span has gone up a bit from 9 for­ward 8 re­verse to be­tween 10-12 for­ward and re­verse de­pend­ing on how I’m feel­ing. I’m still not sure if the im­prove­ments in digit span are gen­uine mem­ory im­prove­ments or in­creased skill at chunk­ing.”

Jt­toto fur­ther wrote in re­sponse to moe:

“Yes, I’ve con­tin­ued to train QnB my­self (about 3-4 times a week). Based on the iqout.­com test, if any­thing, I’ve gone down a lit­tle!. This is not sur­pris­ing and prob­a­bly not at­trib­uted to n-back­ing. I’m at the age where cog­ni­tive de­cline be­gins and I was de­pressed that day. At the same time, one would think I would see mea­sur­able gains by now.”

"I’ve had pretty unim­pres­sive find­ings. I’ve used Brain Work­shop 4.4 for about four months, with about a half-hours use 4-6 days a week. I used Den­mark IQ test and scored a 112 and after DNB I scored a 110.

My max DNB level was 11. Hours and hours and no gain in iq."


“…here they are (in or­der of test tak­ing): 119, 125, 125, 107, 153, 131, (and I would say be­tween 125 and 131 was my real iq) from differ­ent on­line tests al­most 2 years ago be­fore start­ing n-back­ing. after two years (I took the same bunch of on­line iq tests 3 weeks ago be­fore try­ing faster tri­als) I got: 126, 135, 124, 125. so there was­n’t much of a change. but I had been play­ing n-back softly for a long time. I ex­pected my iq to jump at least by 5 to 10 points, from what I felt in my life. then after my week of faster tri­als, I did this iq test: and I got 149. if any­one who al­ready knows his iq wants to try it, I’d be cu­ri­ous to know if they also score higher than ex­pect­ed, at first try of course. I thought the 153 I once got was pure chance, but maybe it was­n’t com­plete­ly, and that would be cool.”


“Been train­ing for about 5 weeks now, 30 mins a day and made very quick progress ini­tial­ly, and now shut­tling be­tween n=7 and n=8 and oc­ca­sion­ally reach­ing n=9 (when I set out, I be­gin with n=2 and the value of N for the next round de­pends on my per­for­mance in the round I just fin­ished)…I took a few in­tel­li­gence tests (mostly cul­ture in­sen­si­tive), and the scores have ac­tu­ally”DROPPED" some 3-4 per­cent. Al­though I guess that does­n’t mean much be­cause I took those tests to­wards the end of the day at work and was some­what ex­haust­ed, but it sure as hell means that there is no in­crease in my in­tel­li­gence ei­ther!!"


“I have used du­al, sin­gle and com­bi­na­tion n back reg­u­larly for al­most 2 years and no pos­i­tive re­sults come from it. I have the ex­act same IQ as I have ac­cord­ing to Den­mark IQ test. Not even a cou­ple of points high­er….Just to clar­i­fy, I have used n back, seen no im­prove­ment based on IQ tests or re­al-life ben­e­fits.”

Keep in mind, that if IQ is im­proved, that does­n’t nec­es­sar­ily mean any­thing un­less one em­ploys it to some end. It would be a shame to boost one’s IQ through N-back, but never use it be­cause one was too busy play­ing!

Other effects

Be­tween 2008 and 2011, I col­lected a num­ber of anec­do­tal re­ports about the effects of n-back­ing; there are many other anec­dotes out there, but the fol­low­ing are a good rep­re­sen­ta­tion - for what they’re worth71.

Be­sides these col­lected re­ports, there is an on­go­ing group sur­vey (spread­sheet re­sults); n-back­ers are strongly en­cour­aged to sub­mit their dat­a­points & opin­ions there as well.


  • Ashir­go: “To be hon­est, I do not feel any ob­vi­ous differ­ence. There are mo­ments in which I per­ceive a sig­nifi­cant im­prove­ment, though, as well as par­tic­u­lars task which are much eas­ier now.”

    “I have also ex­pe­ri­enced bet­ter dream re­call­ing, with all these rever­ies and other hal­lu­ci­na­tions in­clud­ed. I am more hap­pier now than ever. I did doubt it would be ever pos­si­ble! I am also more prone to get ex­cit­ed…Now peo­ple in my moth­er­land are just bor­ing to lis­ten to. They speak too slow and seem as though it took them pains to ex­press any­thing. I did not no­tice that after I had done my first ninety days of n-back, but now (after 2.5 months) it is just con­spic­u­ous.”ref

    “My change of opin­ion72 can be eas­ily at­trib­uted to the im­prove­ment of mood, in co­in­ci­dence with the mere fact that the win­ter days have passed and now there is a bright and sunny Spring in my coun­try”; when asked if the pre­vi­ous means Ashirgo at­trib­utes all the im­prove­ment to the weath­er, Ashirgo replied: "For­tu­nate­ly, I can at­tribute many changes to n-back, I can now han­dle var­i­ous tasks with lit­tle effort and it takes me much less time in com­par­i­son with oth­ers (e­spe­cially when I know what to do). Nev­er­the­less, the main prob­lem for me is that I am also oc­cu­pied with few things that I sup­pose to be able to test my newly ac­quired po­ten­tial, there­fore I can­not say that ‘changes’ are ex­plicit every­where.

    On the other hand, I am start­ing to be­lieve that any im­prove­ments (that one can ex­pect) so smoothly and swiftly be­come a nat­ural part of one’s ca­pa­bil­i­ties that it makes them hardly no­tice­able un­til some test­s/mea­sures are tak­en."

  • chin­mi04: “For me, it defi­nitely has taught me how to fo­cus. But I’m still not sure whether that has some­thing to do with merely com­ing to re­al­ize the im­por­tance of fo­cus­ing, or whether the pro­gram has re­ally phys­i­cally rewired my brain to fo­cus bet­ter. In any case, it ap­pears that I’m now faster at men­tal rea­son­ing, cre­ative think­ing and speak­ing flu­en­cy. But again, the effects are not so clear as to com­pletely elim­i­nate any doubt re­gard­ing the con­nec­tion with the n-back pro­gram.”

    “I have been main­tain­ing a per­sonal blog on Word­Press since 3 years ago. Av­er­age post per month : a lit­tle over 1. Then I started with du­al-n-back at the end of No­vem­ber… num­ber of posts in Jan­u­ary : 7! (none are about n-back)”

  • Ar­seneLupin: “Not much, yet, but I feel that I can eas­ier get a hold of a dis­cus­sion. The feel­ing is the same as when I am mas­ter­ing a cer­tain n-back in the game (a bit hard to ex­plain).”

  • John: “I feel much sharper since I started in the mid­dle of last No­vem­ber…My pro­duc­tiv­ity is much higher these days. I’m a non-fic­tion writer, so hav­ing a higher work­ing mem­ory and fluid in­tel­li­gence di­rectly leads to bet­ter (and faster) per­for­mance. It’s amaz­ing to see the stuff I pro­duce to­day and com­pare it to be­fore I be­gan the Dual N-Back train­ing. Al­so, I am si­mul­ta­ne­ously learn­ing Ger­man, French and Span­ish, and I’m cer­tain this is help­ing me learn those lan­guages faster.”

  • Ginkgo: “DN-Back has prob­a­bly helped me with one of my hob­bies.”

  • Ba­maDoc: “I note a sub­jec­tive differ­ence in re­call. There might be some in­crease in at­ten­tion, but I cer­tainly do no­tice a differ­ence in re­call. It might be place­bo, but I am con­vinced enough that I con­tinue to find time to use the pro­gram.”ref

  • kar­nau­trahl: “Since No­vem­ber how­ev­er, I be­gan to read the Neu­ro­science book in more de­tail. I men­tioned late De­cem­ber I think that I was find­ing I could un­der­stand more stuff. I’ve spent about £1000 on books since No­vem­ber. The large ma­jor­ity are books on the brain, source from Ama­zon re­views, read­ing lists and out of my own pi­rate list when I liked a book. I stopped Dual n Back in De­cem­ber, ear­ly. The ben­e­fits have stayed how­ev­er. I tested this the other day, very eas­ily go­ing to 3 n back, which was mostly where I was be­fore. I guess in a way I’m try­ing to say that for me, whilst the fo­cus may have been on G in­crease and IQ etc, now the fo­cus is on—what’s re­ally hap­pened and what can I do with it. What I can do with it is choose to con­cen­trate long enough to gen­uinely un­der­stand fairly tech­ni­cal in depth chap­ters on sub­jects often new to me.”ref Kar­nau­trahl writes more on his self­-im­prove­ments in his thread “Sec­ond lot of train­ing start­ed-and long term ex­pe­ri­ence over­all.”, and de­scribes an in­ci­dent in which though he stopped us­ing DNB 3 months pre­vi­ous­ly, he still dealt with a tech­ni­cal is­sue much faster and more effec­tively than he feels he would’ve be­fore.

  • nega­tron: “One per­haps co­in­ci­den­tal thing I no­ticed is that dream rec­ol­lec­tion went up sub­stan­tial­ly. A good while after I stopped I de­vel­oped an odd cu­rios­ity for what I pre­vi­ously con­sid­ered un­pleas­ant ma­te­ri­al, such as ad­vanced math­e­mat­ics. Never imag­ined I’d con­sider the thought of ad­vanced cal­cu­lus ex­cit­ing. I be­gan read­ing up on such sub­jects far more fre­quently than I used to. This was well after I’ve long for­got­ten about dual n-back so I find it hard to at­tribute it to a placebo effect, be­liev­ing that I’m more adapted to this ma­te­r­i­al. On the other hand I don’t re­call read­ing any­thing about mo­ti­va­tional ben­e­fits to dual n-back train­ing so I still con­sider this con­jec­ture and per­haps an event­ful co­in­ci­dence just the same.”ref

  • su­tur: “i did­n’t re­ally no­tice any con­crete changes in my think­ing process, which prob­a­bly, if ex­is­tent, are rather hard to de­tect re­li­ably any­way. one thing i did no­tice how­ever is an in­creased sense of calm­ness. i used to move my legs around an aw­ful lot while sit­ting which i now don’t feel the urge to any­more. but of course this could be placebo or some­thing else en­tire­ly. i also seem to be able to read text (in books or on screen) more flu­ently now with less dan­ger of dis­trac­tion. how­ev­er, per­son­ally i am quite skep­tic when peo­ple de­scribe the changes they no­tice. changes in cog­ni­tive ca­pac­ity are prob­a­bly quite sub­tle, build up slowly and are hard to no­tice through in­tro­spec­tion.”ref

  • as­triaos: “By ‘ro­bust’, I mean prac­ti­cally every­thing I do is qual­i­ta­tively differ­ent from how I did things 30 days pre­vi­ous to the dual n-back train­ing. For in­stance, in physics class I went from vaguely un­der­stand­ing most of the con­cepts cov­ered in class to a mas­tery thor­ough enough that now my ques­tions usu­ally tran­scend the scope of the in­-class and text­book ma­te­ri­al, rou­tinely stu­pe­fy­ing my physics teacher into longer-than-av­er­age paus­es. It’s the same ex­pe­ri­ence for all of my class­es. Some­how, I’ve learned more-than-I usu­ally learn of physic­s/­gov­ern­ment/ etc. (all of my class­es, and any topic in gen­er­al) in­for­ma­tion from sources out­side of class, and with­out what I con­sider sig­nifi­cant effort. I feel like my learn­ing speed has gone up by some fac­tor greater than 1; I can fol­low longer ar­gu­ments with greater pre­ci­sion; my vo­cab­u­lary has im­proved; I can pay at­ten­tion longer; my prob­lem solv­ing skills are sig­nifi­cantly bet­ter… Re­al­ly, it’s amaz­ing how much cog­ni­tion de­pends on at­ten­tion!”ref

  • flashquar­ter­mas­ter re­ports N-back cured his chronic fa­tigue syn­drome?

  • UOChris1: “Harry Kahne was said to have de­vel­oped the abil­ity to per­form sev­eral tasks at one time in­volv­ing no less the 16 differ­ent ar­eas of the brain….­Sur­pris­ing­ly, I am slowly de­vel­op­ing the abil­ity si­mul­ta­ne­ously per­form quad com­bi­na­tion 3-back while recit­ing the al­pha­bet back­wards. The prac­tice is very diffi­cult and re­quires loads of con­cen­tra­tion but I am ex­pe­ri­enc­ing per­ceiv­able gains in clar­ity of thought from one day of prac­tice to the next whereas my gains from Brain Work­shop alone were not per­ceiv­able on a daily ba­sis.” UOChris1 wrote of an­other mode: “Triple-N-Back at .5sec in­ter­vals and pi­ano notes in­stead of let­ters has greatly im­proved my sub­jec­tively per­ceived flu­id­ity of thought. I am much more en­gaged in class, can read much quick­er, and am com­ing up with many more cre­ative so­lu­tions now than ever be­fore. I did­n’t no­tice the im­prove­ments as much when I was us­ing slower in­ter­val­s—I feel I make more de­ci­sion cy­cles in a given amount of time be­fore com­ing to a so­lu­tion.”

  • Pon­tus Granström “I cer­tainly feel calmer hap­pier and more mo­ti­vated after do­ing DNB, it has to do with the in­crease of dopamine re­cep­tors no doubt!”

  • Chris War­ren sum­ma­rizes the re­sults of his in­ten­sive prac­tice (cov­ered above): “For those that are cu­ri­ous, I no­ticed the largest change in my thought processes on Wednes­day. My abil­i­ties were no­tice­ably differ­ent, to the ex­tent that, at some points, it was, well, star­tling. I’ve started get­ting used to the feel­ing, so I can’t re­ally com­pare my in­tel­li­gence now vs. Wednes­day. How­ev­er, I’m com­pletely con­fi­dent that I’ve be­come smarter. Un­der the kind of stress I’ve put my brain through, I can’t imag­ine a sce­nario where that would­n’t hap­pen.”

    “After the first cou­ple days of train­ing, I ex­pe­ri­enced a very rapid in­crease in in­tel­li­gence. It sud­denly be­came eas­ier to think. I can’t give you any hard ev­i­dence, since I did­n’t bother to take any tests be­fore I start­ed. How­ev­er, I can give you this: when I woke up Wednes­day morn­ing, I felt the same as I did after the first time I tried n-back. Ex­cept the feel­ing was 10 times stronger, and my think­ing was no­tice­ably faster and more com­pre­hen­sive.”

  • Ra­man re­ports an ini­tial null re­sult: “19 days with n-back are over… no sub­jec­tive ben­e­fits as such. But I am aware at what point I am com­fort­able or not. e.g. y’­day play­ing the game was effort­less, and to­day my brain felt sort of sticky, the se­quence was just not stick­ing in my brain. very strange what a few hours can do.”

  • iwan tuli­jef says that “Long time ago I was di­ag­nosed Adhd [sic] and for long time I took meds and this train­ing helped me to re­duce my meds nearly to ze­ro, com­pared with the doses I took be­fore. Un­for­tu­nately this haven’t fixed the whole thing. But what I no­ticed was, hmm… those things are very diffi­cult to de­scribe…. that time by time I got more con­trol about my men­tal life. Ob­vi­ous effects in so­cial mat­ters were e.g. that I could fol­low con­ver­sa­tions bet­ter and be­have more nat­u­ral­ly. In my ed­u­ca­tion mat­ters, e.g. that I un­der­stood maths proofs bet­ter. There are a lot of de­tails. In­ter­est­ing was, as these is­sues are, to un­der­state it a bit, not unim­por­tant for me, that in the be­gin­ning when I re­marked changes, I got a bit eu­phoric, so the first effects of n-back feeled like the strongest.” and warns us that “It’s very diffi­cult and very ques­tion­able to take ob­jec­tive in­for­ma­tions out of sub­jec­tive self eval­u­a­tion.” (I­wan trained for 3-4 months, 20 rounds a day in the morn­ing & evening.)

  • jt­toto saw no gain on an IQ test, but thinks he’s ben­e­fited any­way: “My friends have al­ways called me inat­ten­tive and ab­sen­t-mind­ed, but since play­ing n-back no one has called me that for a while. I now never for­get where I park my car, when I used to do that nearly every other day. I feel more at­ten­tive. Even if my abil­ity to solve prob­lems has­n’t im­proved, the gains in my mem­ory are real and mea­sur­able.”

  • re­ece: “Not that I’ve no­ticed [an im­prove­ment in ]. I have no­ticed an im­prove­ment in my work­ing mem­ory how­ev­er—seems eas­ier to jug­gle a few ideas in my head at the same time which pre­sum­ably the quad-n-back has helped with.” “I re­cently no­ticed that it ap­pears to have made me bet­ter at play­ing ping pong and tetris. Oddly enough how­ev­er, it does­n’t ap­pear to have im­proved my re­ac­tion time…­Work­ing mem­ory has im­proved, how­ever other things I’ve al­ways strug­gled with such as un­cued long term mem­ory re­call have not… I’m still very ab­sen­t-minded and be­lieve n-back has made me more eas­ily dis­tractable (low­ered la­tent in­hi­bi­tion?), al­though to be fair, I may have brought this on my­self by play­ing quad n-back and this was not some­thing I no­ticed when only play­ing dual n-back. I seem to be able to get by on about one hour less sleep per night and per­form bet­ter cog­ni­tively when sleep de­prived. Dream re­call has in­creased sig­nifi­cantly as has lu­cid dream­ing. I do take a few nootrop­ics, how­ever I’ve been tak­ing the same ones for years…Ver­bal flu­ency ap­pears to have im­proved, proper spelling and punc­tu­a­tion are things I’ve al­ways strug­gled with and do not ap­pear to have ame­lio­rated re­sul­tant from n-back train­ing.” (Poll) “In my ex­pe­ri­ence with dual and mul­ti­modal n-back, the ben­e­fits I’ve most ob­served have been in­creased mul­ti­task­ing abil­ity and in­creased con­cen­tra­tion in the pres­ence of dis­trac­tions. For me, the ben­e­fits of n-back train­ing are most ap­par­ent on days I don’t take my ADHD med­ica­tion. I have been train­ing DNB with po­si­tion-sound and col­or-im­age modes late­ly. I used QNB for sev­eral months in the past, how­ever I (sub­jec­tive­ly) be­lieve DNB is giv­ing me the most ben­e­fit.” (ADHD thread)

  • Michael Camp­bell: “Some­thing very mi­nor to some, but was good for me; I’m able to con­cen­trate while read­ing a lot more than I have been able to in the past.”

  • ex­i­gentsky: "I’ve seen im­prove­ments in ex­ec­u­tive func­tion and mo­ti­va­tion. After DNB, I am more in­clined to study and com­plete long pend­ing items. How­ev­er, there is a con­found­ing vari­able. I don’t usu­ally do DNB when in an un­healthy state of mind (for ex­am­ple, with lit­tle sleep and ex­tremely high stress). Still, I be­lieve that I can at­tribute some of the effects only to DNB.

    In terms of work­ing mem­ory and other cog­ni­tive mea­sures, I’m not sure. I don’t no­tice any­thing dra­matic but also haven’t stuck to a DNB regime for more than a few week­s."

  • cev: "I think I’ve put my fin­ger on a par­tic­u­lar ben­e­fit of dnb train­ing: it seems to help my brain’s ‘in­ter­nal clock’ - I am bet­ter able to or­der my thoughts in time.

    DNB has also helped my foos­ball (!) play­ing: at a high level the game in­volves com­plex strings of mo­tor move­ments and since I’ve been train­ing, I’ve found that my co­or­di­na­tion of these move­ments has greatly im­proved de­spite no longer prac­tis­ing."

  • erm: “I can rely on this to dras­ti­cally re­duce anx­i­ety, flight­i­ness, im­prove con­cen­tra­tion. It also seems to whet my ap­petite for in­tel­lec­tual work and in­crease pur­pose­ful­ness across the board.”

  • Tofu, after a year of n-back­ing: "N-back train­ing may have some­how im­proved my ver­bal in­tel­li­gence, but since ver­bal in­tel­li­gence is a form of crys­tal­lized in­tel­li­gence and train­ing work­ing mem­ory is sup­posed to pri­mar­ily im­prove fluid in­tel­li­gence, it prob­a­bly did­n’t. My score on the ver­bal sub­test went up and then down which would make no sense if it did have any in­flu­ence…S­ince my IQ score in­creased from the first test to the sec­ond test, and stayed the same from the sec­ond test to the third test it could be pos­si­bly that work­ing mem­ory only con­tributes to IQ up to a cer­tain point. All in all, I feel more in­clined to say that n-back train­ing has only a lit­tle if any effect on IQ though which is why rea­son I’m prob­a­bly go­ing to stop do­ing the n-back train­ing.

    On a more pos­i­tive note, since I started n-back train­ing I have no­ticed bet­ter con­cen­tra­tion which I had a se­ri­ous prob­lem with be­fore. In gen­er­al, I feel like I think more clearly and I at least feel like I’ve be­come smarter too. I’ve reached a pretty high level in n-back and any gains I’ve made in the last month or two have been small, so I think I’ve reached a long-term plateau which is an­other rea­son for me to stop the train­ing. From my ex­pe­ri­ence when I stop the n-back train­ing for a month or two and re­turn to n-back train­ing I still per­form at the same level any­way. It seems like the effects from train­ing are go­ing to last a while which is also good news. Over­all, I feel like the n-back train­ing was worth it but if I had it to do over I would have prob­a­bly stopped after a cou­ple of month­s."

  • kriegerlie: “i’ve defnitely had some ben­e­fit, like pon­tus said, dunno about be­ing smarter, but my fo­cus is in­cred­i­ble now. I can do what i thought I could never do, purely be­cause I can fo­cus more. Placebo or not. It’s a defi­nite effect.”

  • Rotem: “DNB works, It’s one of the best in­vest­ments I made in my life. I have much less anx­i­ety ( I suffered from GAD my life was a night­mare), more con­fi­dence and I guar­an­tee more in­tel­li­gence - I can feel it…”

  • chortly: “For a while I imag­ined that my work­ing mem­ory mus­cles were in­deed strength­en­ing, the main sen­sa­tion be­ing that I could re­tain the var­i­ous threads of a com­pli­cated con­ver­sa­tion bet­ter as they dan­gled and were for­got­ten by the other con­ver­sa­tion­al­ists. But that was prob­a­bly just wish­ful think­ing. Be­cause it’s bor­ing and diffi­cult, I haven’t stuck with it, though I keep in­tend­ing to.”

  • JHar­ris: “I’ve been work­ing with the dual n-back pro­gram for a bit of time now. Im­prove­ment is slow, but seems to be hap­pen­ing; I just had a 68% run at dual 3-back. Ob­ser­va­tions like this are not re­ally sci­en­tific and hell­ishly sub­ject to bi­as, but I think I may be notic­ing it slightly eas­ier to think effec­tive­ly.”

  • Neu­ro­hacker (in a thread on ): “I’m defi­nitely find­ing it help­ful, even if it’s just giv­ing me some prac­tice at fo­cus­ing…as a com­ple­men­tary strat­egy [to med­ica­tion], it’s cer­tainly work­ing won­ders.”

  • iwan tuli­jef: “n-Back helped me a lot. Es­pe­cially in the be­gin­ning when I started with DNB, the effect was as­tound­ing. I got much faster in un­der­stand­ing writ­ten and spo­ken word­s.In the be­gin­ning I think the func­tion of my work­ing mem­ory was re­ally bad. What then hap­pened is that I got ha­bit­u­ated to the effect and the in­creases were small­er, so notic­ing im­prove­ments got more diffi­cult.”

  • Arkan­j3l: “On a side, I re­ally en­joy the lu­cid feel­ing I get after an hour of n-back. I start to look at things and ideas seem to flow into my head very vividly (I’ve made some of my best Lego cre­ations after an n-back ses­sion :p).”

  • Michael Lo­gan: “…and learned very, very quickly that I had a short term mem­ory and at­ten­tion is­sue. The dual n back task laughed at me, but I vowed to over­come my inat­ten­tion and short term mem­ory is­sues, and within a few prac­tices, I no­ticed an im­prove­ment not only in my scores on the com­put­er­ized game, but in ses­sion with my clients….So Mind Sparke does pro­vide that kind of novel learn­ing chal­lenge. I have not taken an IQ test, but I do be­lieve the use of the tool is help­ing me build cog­ni­tive re­serve for the later stages of my life.”

  • mile­stones: “I’m grate­ful for the gains I seemed to have re­ceived from train­ing dual n back. I used to be ex­tremely for­get­ful with re­mem­ber­ing where I put things and now it’s very easy to re­trace steps and re­call where I placed xyz item. As far as IQ tests go, I did see a gain on a well de­signed (un­timed) cul­ture fair test of about 1 stan­dard de­vi­a­tion after train­ing one DNB on and off for close 2 years. (Other tests with lower ceil­ings, how­ev­er, showed no or mar­ginal gain­s).” A later post: “The gains I’m see­ing are: faster en­cod­ing speed; faster and more ac­cu­rate re­trieval of data from long term mem­o­ry; as well as an in­crease in data-se­quenc­ing speed (the lat­ter is a rel­a­tive weak­ness of mine that now seems to have been helped by con­sis­tent quad-back train­ing—-though I’ve not tested any trans­fer so this is sub­jec­tive). Al­so, though my fluid in­tel­li­gence has prob­a­bly ceased gain­ing, it seems I’m func­tion­ing at higher bands of abil­ity far more reg­u­lar­ly—even when I’m tired or slug­gish.”

  • Lach­lan Jones wrote, after a be­fore/after IQ re­port, “The most sig­nifi­cant real word ap­pli­ca­tion for me has been im­prove­ments in my pi­ano play­ing. I am a pi­anist and can re­port sig­nifi­cant im­prove­ments in my sight read­ing and the rate at which I learn new pieces.”

  • un­funf: “While I haven’t taken an IQ test to see if it has gar­nered any IQ im­prove­ment, I can say I started off at dual 4-back only 2 weeks ago and I am now near­ing dual 6-back. I can also at­test to a pretty large work­ing mem­ory im­prove­ment, be­yond what I would call placebo (the effects of which I am very well aware). Even if it is not very effec­tive, I still say this game is fun.”

  • Neu­roGuy: “Du­al-N-Back has sub­jec­tively done more for me in less then two weeks than any sin­gle nootrop­ic, can hardly imag­ine it com­bined with spaced-repepi­tion.”

  • TeC­NoY­oTTa: “I also want to re­port that after train­ing on DNB I found that I am dream­ing al­most every day…by the way I re­mem­ber that this effect was not di­rectly after train­ing…un­for­tu­nately I stopped us­ing DNB from about 2 months or some­thing like that and now I dream less”

  • dime­coin: “I make no claims, other than anec­do­tal - in that it seems to re­lax me and able to han­dle stress bet­ter when I do it reg­u­lar­ly.”

  • Arbo Arba: “I did find a lot of changes come to my brain and per­son­al­i­ty, but I’m not sure if it’s from im­prov­ing WM or if it’s just from spend­ing a lot of time in an al­pha-wave dom­i­nant state. I think it’s be­ing in a pro­longed al­pha-brain wave dom­i­nant state, tbh, be­cause I found that when I was younger and took up heavy read­ing projects I felt the same im­prove­ments—that is, hav­ing more fo­cus, be­ing able to ‘hear’ my­self think very dis­tinctly to the point where I could com­pose po­em­s/e­mails in my head with­out effort. I don’t know why this is, but it hap­pens so much with me that I can’t doubt that there is a real effect on my per­son­al­ity and de­fault men­tal state when I’m do­ing ‘in­tel­lec­tual’ things.”

  • Akiyama Shinichi: “I train 3 times a day and every se­sion last about 20 min­utes. After a month a went to my chess club and com­pletely crash play­ers who was at com­pletely differ­ent lev­el. I chose one of the strongest player (at my level of course), be­cause he was able to tell me if I’ve re­aly im­proved. Then I had to re­veal my se­cret, and after month I tell how it works for them. I no­tice that I im­prove not only in chess. I’m a pi­ano player and it’s re­ally chal­leng­ing. I was learn­ing very slow­ly, but yes­ter­day my teacher told me that in two weeks I learnt much more than in the last 2 months. He was even sus­pect­ing me that I take lec­tures from other teacher, not only from him. And that’s not all. I’m a stu­dent and one a month every of us have to pre­pare pre­sen­ta­tion on some top­ic. A few days ago was my turn. I did­n’t no­tice it by my­self, but one of my friend told me that I was very well-pre­pared, be­cause I stopped mak­ing that an­noy­ing sound like”umm“,”yyyy" when I was think­ing what to say. When I was per­form­ing my pre­sen­ta­tion I don’t have to think what to say next be­cause I’d al­ready know and did­n’t have to think about it much."

  • whois­bam­bam: “My mind feels faster. I also seem to have less men­tal fa­tigue dur­ing study­ing, n-back­ing, etc. I am more con­fi­dent. I am con­fi­dent that my mem­ory has im­proved in­de­pen­dent of n-back­ing (what they call ‘far trans­fer’ effec­t). I am not say­ing it is a HUGE differ­ence. I can not say the same is true for any sup­ple­ment i have taken other than pos­si­bly some small effect with mag­ne­sium l-thre­onate which also seems to make me ‘less men­tally tired’ in par­tic­u­lar, in­ter­est­ing­ly.”

  • Christo­pher Dzi­ało: “I’ve trained with n-back for sev­eral months and have no­ticed a pro­found abil­ity to sight read mu­sic and lo­cate the notes, my speed and over­all dex­ter­ity has dras­ti­cally in­creased and I shall con­tinue to n-back and grow my mu­si­cal tal­ent.”

No benefits

  • Con­fuzedd: “[asked if felt ‘sharper’]: Noth­ing.”
  • Chris: “One thing I have no­ticed is the rec­ol­lec­tion of a num­ber of very un­pleas­ant im­ages in dreams. Specifi­cal­ly, im­ages of bod­ily dis­ease, mu­ti­la­tion, in­jury and post-mortem de­com­po­si­tion. I find it diffi­cult to be­lieve it’s just a co­in­ci­dence, be­cause I can’t re­mem­ber when I last had such a dream, and I’ve had maybe half a dozen since I started dual n-back. But per­haps it’s sim­ply ow­ing to bet­ter re­call.”ref
  • Pheonexia: “now I’m at 6-back and am con­sis­tently be­tween 50 and 80% ac­cu­rate….All that said, I have NOT no­ticed any differ­ences in my men­tal ca­pac­i­ty, in­tel­li­gence, daily life, or even abil­ity to re­mem­ber things that just hap­pened. I still some­times for­get peo­ple’s names right after they tell me them. I’m go­ing to keep train­ing though, be­cause just be­cause I haven’t con­sciously no­ticed these things, I have faith in sci­en­tific stud­ies, so with enough train­ing hope­fully I’ll yield some pos­i­tive ben­e­fits.”
  • The­Q17 re­ports lit­tle to no ben­e­fit: “At any rate, I don’t feel study­ing is any eas­ier al­though it was­n’t re­ally diffi­cult to be­gin with for me. Per­haps I’ll give it an­other go over break and re­port back. My goal orig­i­nally was to get to P5B be­fore adding a sec­ond sound stim­u­lus mak­ing a Sex­tu­ple Nback but I don’t know if Shamanu made an up­dated ver­sion to make that any eas­i­er. I’m also kind of on the fence about the effect on the depth of train­ing. It may have been more ben­e­fi­cial to do higher N lev­els in­stead of more stim­uli.”
  • Jonathan Graehl: “I can do dual 4-back with 95%+ ac­cu­racy and 5-back with 60%, and I’ve likely plateaued (nat­u­ral­ly, my skill rapidly im­proved at first). I en­joy it as”prac­tice fo­cus­ing on some­thing“, but haven’t no­ticed any ev­i­dence of any gen­eral im­prove­ment in mem­ory or other men­tal abil­i­ties.”
  • Will New­some: “After do­ing 100 tri­als of dual N back stretched over a week (mostly 4 back) I no­ticed that I felt slightly more con­scious: my emo­tions were more salient, I en­joyed sim­ple things more, and I just felt gen­er­ally more alive. There were tons of free vari­ables for me, though, so I doubt cau­sa­tion.”
  • steven0461: “I did maybe 10-15 half-hour ses­sions of mostly D5B-D6B last year over the course of a few weeks and did­n’t no­tice any effects.”
  • Egg­plantWiz­ard (D3B->D10B): “I would say that there has been some form of im­prove­men­t—though it’s not clear if the im­prove­ment is task-spe­cific. I haven’t no­ticed any sig­nifi­cant differ­ence in my day to day life, but (to be im­mod­est in the name of effi­ciency for a mo­ment) I had a very good mem­ory to be­gin with, and I would say strong fluid in­tel­li­gence. It’s pos­si­ble that peo­ple start­ing from po­si­tions of lower fluid in­tel­li­gence would see a more pro­nounced ben­e­fit.”
  • Matt: “…I’ve cer­tainly im­proved at n-back type tasks, I can’t say that I’ve no­ticed any im­prove­ment while han­dling real life prob­lems. I think the effects do gen­er­al­ize - I’m quite good at highly g-loaded tasks like the now, even with­out much prac­tice - but the range of tasks which are sub­ject to im­prove­ment from n-back­ing seems lim­it­ed. I’m bet­ter at tasks in­volv­ing men­tal up­dat­ing, but my short term mem­ory has only slightly im­proved, if at all. I don’t have an ac­cu­rate way of mea­sur­ing my change in Gf (or g), as most of the fluid rea­son­ing tasks avail­able on­line use the same/sim­i­lar rule pat­terns or aren’t ac­cu­rately normed, but as I said be­fore, my real life prob­lem solv­ing abil­i­ties have not sub­jec­tively im­proved…”
  • Je­lani Sims: “I’ve been do­ing DNB since the group start­ed, I haven’t no­ticed any­thing out of the or­di­nary in terms of cog­ni­tion. But I never took a be­fore and after IQ test and I haven’t re­ally done any­thing that I found men­tally diffi­cult be­fore. So it’s very hard for me to gauge men­tal im­prove­ments with noth­ing for me to base it on. I also changed my di­et, started mind­ful­ness med­i­ta­tion and ex­er­cis­ing around the same time I started DNB, in an over­all at­tempt to de­lay brain de­cline. Mak­ing it even more diffi­cult to at­tribute any­thing di­rectly to DNB. What I can say is I have been stuck on 12 for 4 months now, each level was in­creas­ingly more diffi­cult to pass and 12 seems to be some sort of tem­po­rary plateau.”
  • ar­gumzio: “I’ve seen no net ben­e­fit. Com­pared to im­proved nu­tri­tion, ex­er­cise, sleep­ing, and the oc­ca­sional nootropic (e.g., Pirac­etam, Al­pha GPC, CDP Citi­co­l­ine, Resver­a­trol, Kre-Al­ka­lyn & Cre­a­tine Mono­hy­drate, etc.), DNB did noth­ing. How­ev­er, in terms of sub­jec­tively im­proved fo­cus (count­ing the near-cer­tain pos­si­bil­ity that the afore­men­tioned changes also in­flu­enced it), QNB* did the most for me, that is, al­low­ing me to ab­sorb in­for­ma­tion for longer pe­ri­ods of time and main­tain this effort much later into the evening while mit­i­gat­ing the dele­te­ri­ous effects of fa­tigue and al­low­ing me to feel rested after un­usu­ally shorter pe­ri­ods of sleep.”


One of the wor­ries oc­ca­sion­ally cited is that DNB train­ing mostly serves to in­crease one’s fo­cus on the task one is think­ing about. Which is great in most con­texts but, the fear goes, the abil­ity to fo­cus on one thing is the abil­ity to ex­clude (‘in­hibit’) thoughts on all other top­ics - which is cru­cial to cre­ativ­i­ty. Work­ing mem­ory and abil­ity to shift at­ten­tion has a strong cor­re­la­tion with be­ing able to solve s with lat­eral think­ing, but as with the WM-IQ link, that does­n’t say what hap­pens when one in­ter­venes on one side of the cor­re­la­tion (cor­re­la­tion is not cau­sa­tion):

In­di­vid­u­als may have diffi­culty in keep­ing in mind al­ter­na­tives be­cause mul­ti­ple pos­si­bil­i­ties can ex­ceed their work­ing mem­ory ca­pac­ity (Byrne, 2005; John­son-Laird and Byrne, 1991; 2002). They also need to be able to switch their at­ten­tion be­tween the al­ter­na­tive pos­si­bil­i­ties to reach a so­lu­tion. On this ac­count, key com­po­nent skills re­quired in in­sight prob­lem solv­ing in­clude at­ten­tion switch­ing and work­ing mem­ory skill­s….At­ten­tion and work­ing mem­ory may be cru­cial for differ­ent as­pects of suc­cess­ful in­sight prob­lem solv­ing. Plan­ning a num­ber of moves in ad­vance may be im­por­tant to solve in­sight prob­lems such as the well-known nine-dot prob­lem (Chron­i­cle, Ormerod and Mac­Gre­gor, 2001). At­ten­tion may play a role in help­ing peo­ple to de­cide what el­e­ments of a prob­lem to fo­cus on or in help­ing them to di­rect the search for rel­e­vant in­for­ma­tion in­ter­nally and ex­ter­nal­ly.

…In­di­vid­u­als who are good at solv­ing in­sight prob­lems are also good at switch­ing at­ten­tion. Cor­rect per­for­mance on the in­sight prob­lems was as­so­ci­ated with cor­rect per­for­mance on the vi­sual el­e­va­tor task (r=.515, p<.01). Cor­rect per­for­mance on the in­sight prob­lems was as­so­ci­ated with cor­rect per­for­mance on the plus-mi­nus prob­lems (r=-.511, n=32, p<.001)…­Con­sis­tent with this ac­count in­di­vid­u­als who are bet­ter at stor­ing and pro­cess­ing in­for­ma­tion in work­ing mem­ory are bet­ter at solv­ing in­sight prob­lems. [cor­re­la­tion with prob­lem score: r=.39 for digit span, r=.511 for sen­tence span]73

The ma­jor piece of ex­per­i­men­tal ev­i­dence is Takeuchi 2011 & Var­tan­ian 2013, treated at length in the fol­low­ing sub­sec­tion and well worth con­sid­er­a­tion; the rest of this sec­tion will dis­cuss other lines of ev­i­dence.

is re­lated to changes caused by n-back­ing (see the Mc­Nab re­cep­tor study & for a gen­eral re­view, Söderqvist et al 2011), and in­crease in dopamine has been shown to cause a nar­row­ing of fo­cus/as­so­ci­a­tions in tasks74. There are other re­lated cor­re­la­tions on this; for ex­am­ple, Cas­sim­jee 201075 re­port that “…the tem­pera­ment di­men­sion of was in­versely re­lated to per­for­mance ac­cu­racy on the LNB2 (Let­ter-N-Back).” But as ever, ; this re­sult might not mean any­thing about some­one de­lib­er­ately in­creas­ing per­for­mance ac­cu­racy by prac­tice - we might take it to mean just that nar­row un­in­ter­ested peo­ple had a small ad­van­tage at n-back­ing when they first be­gan. Cas­sim­jee 2010 cites 2 other stud­ies sug­gest­ing what this cor­re­la­tion means: “…par­tic­i­pants with higher im­pul­siv­ity may lack the at­ten­tional re­sources to re­tain crit­i­cal in­for­ma­tion and in­hibit ir­rel­e­vant in­for­ma­tion. The ac­ti­va­tion of re­ac­tive con­trol, which is a sys­tem that mon­i­tors, mod­u­lates and reg­u­lates re­ac­tive as­pects of tem­pera­ment, is in­hib­ited in in­di­vid­u­als high in nov­elty seek­ing…” This sug­gests the per­for­mance differ­ence is a weak­ness that can be strength­ened, not to a fun­da­men­tal trade-off.

Re­ports from n-back­ers are mixed. One neg­a­tive re­port is from john21012101:

I’ve done the dual n-back task avidly for over a month and while I find it makes me men­tally sharper, that comes a high cost - the loss of cre­ativ­ity and lat­eral think­ing. In fact, I ex­pe­ri­ence what is called as se­vere di­rected at­ten­tion fa­tigue (see www.troutfoot.com/attn/dafintro.html).

…and even short booster ses­sions se­verely im­pair cre­ativ­ity to the point that one be­comes very men­tally flat, sin­gle-mind­ed, and I’d even say zom­bie-ish.

Ashir­go, chin­mi04, & putomayo begged to differ in the same thread, with biped plump­ing for a null re­sult.

There are some the­o­ret­i­cal rea­sons to be­lieve DNB is­n’t caus­ing gains at the ex­pense of cre­ativ­i­ty, as there is that Jaeggi study show­ing Gf gains, and Gf is mildly cor­re­lated with cre­ativ­i­ty, ac­cord­ing to ex­i­gentsky:

"Fur­ther­more, if the pre­lim­i­nary re­sults hold and du­al-n-back ac­tu­ally in­creased Gf, it should ac­tu­ally con­tribute to cre­ativ­ity for most peo­ple. After all, stud­ies have shown that cre­ativ­ity (ac­cord­ing to stan­dard tests) and IQ are sig­nifi­cantly cor­re­lated to a cer­tain point (~120 on most). While both tests are im­per­fect and in­com­plete, they do give a gen­eral pic­ture.

I have not felt a de­crease in my cre­ativ­ity and am skep­ti­cal of the idea that du­al-n-back harms it. If the pur­ported mech­a­nism is in­creas­ing , that would be an even big­ger break­through than in­creas­ing IQ. The for­mer is still largely con­sid­ered im­mutable."

Vlad has some more de­tails on those cor­re­la­tions:

“Last but not least, there was this re­search”Re­la­tion­ship of in­tel­li­gence and cre­ativ­ity in gifted and non-gifted stu­dents“, which I stud­ied be­cause of this to­day, and they found pos­i­tive cor­re­la­tion IQ vs ver­bal and fig­ural cre­ative processes (flu­en­cy, flex­i­bil­i­ty, ob­ject de­sign­ing, spe­cific traits, in­sight…). And this mild cor­re­la­tion (of 0.3 - 0.5), did not differ for differ­ent IQ lev­els (higher IQs had mild higher cre­ativ­i­ty, lower IQs had mild lower cre­ativ­ity - al­ways mild re­la­tion­ship, so ex­cep­tions too, but in gen­eral more IQ meant more cre­ativ­i­ty).”

On the other hand, Vlad also points out that:

"…there are few the­o­ries how WM works, and one of the most ex­plain­ing is, that WM and at­ten­tion are tied closely to­gether (Ash al­ways em­pha­sizes this and he is right :). This should work through the fact, that higher WM means more sources for in­hi­bi­tion of dis­trac­tion. So, the more WM, the bet­ter you can con­cen­trate. They tested this with cock­tail party effect: in gen­er­al, only 33% of per­sons catch their name from ir­rel­e­vant back­ground noise, while con­cen­trat­ing on some task. Now they found, that only 20% of high WM peo­ple caught their name, but 65% of low WM. On the other side, con­tem­po­rary re­searches some­times differ be­tween WM, STM, pri­mary / sec­ondary WM, even LTM… But the point is, at­ten­tion works at least partly as a fil­ter, and it gets bet­ter with higher WM.

Now the is­sue with cre­ativ­i­ty. I find this in­ter­est­ing, be­cause I think some­body here wor­ried al­ready about be­ing sub­jec­tively less cre­ative than be­fore BW train­ing, and I got this feel­ing few times too.

…Ev­ery cre­ator must deeply con­cen­trate on his work. Maybe there are differ­ent kinds of cre­ativ­i­ty: “ADHD” cre­ativ­i­ty, mean­ing­ful cre­ativ­i­ty, brain­storm­ing cre­ativ­i­ty, ap­pre­ci­a­tion of art, and so on.

Btw after train­ing dnb, I got this in­ter­est in art - I down­loaded lots of clas­si­cal and other artis­tic pic­tures (n­ever be­fore), and re­ally en­joyed choos­ing which I like. Or have you ever seen “the hours”? I fell in love with that movie and even started to read things from vir­ginia woolf"

As well, Pheonexia points out that & Mc­Nab 2009 demon­strated in­creases in var­i­ous things re­lated to be­cause of DNB, and that there is one study that “Dopamine ag­o­nists dis­rupt vi­sual la­tent in­hi­bi­tion in nor­mal males us­ing a with­in-sub­ject par­a­digm”.

Takeuchi 2011

, Takeuchi et al 2011:

Train­ing work­ing mem­ory (WM) im­proves per­for­mance on un­trained cog­ni­tive tasks and al­ters func­tional ac­tiv­i­ty. How­ev­er, WM train­ing’s effects on gray mat­ter mor­phol­ogy and a wide range of cog­ni­tive tasks are still un­known. We in­ves­ti­gated this is­sue us­ing vox­el-based mor­phom­e­try (VBM), var­i­ous psy­cho­log­i­cal mea­sures, such as non-trained WM tasks and a cre­ativ­ity task, and in­ten­sive adap­tive train­ing of WM us­ing men­tal cal­cu­la­tions (IATWMMC), all of which are typ­i­cal WM tasks. IATWMMC was as­so­ci­ated with re­duced re­gional gray mat­ter vol­ume in the bi­lat­eral fron­to-pari­etal re­gions and the left su­pe­rior tem­po­ral gyrus. It im­proved ver­bal let­ter span and com­plex arith­metic abil­i­ty, but de­te­ri­o­rated cre­ativ­i­ty. These re­sults con­firm the train­ing-in­duced plas­tic­ity in psy­cho­log­i­cal mech­a­nisms and the plas­tic­ity of gray mat­ter struc­tures in re­gions that have been as­sumed to be un­der strong ge­netic con­trol.

Takeuchi 2011 has many points of in­ter­est:

  • these sub­jects are re­ally high qual­ity stu­dents and grad stu­dents - which is why a num­ber of them hit the RAPM ceil­ing (!); and it’s im­plied they are all stu­dents. To­hoku is­n’t Tokyo U, but it’s still re­ally good, Wikipedia telling me “It is the third old­est Im­pe­r­ial Uni­ver­sity in Japan and is a mem­ber of the Na­tional Seven Uni­ver­si­ties. It is con­sid­ered as one of the top uni­ver­si­ties in Japan, and one of the top 50 uni­ver­si­ties in the world.”

  • While high qual­i­ty, there aren’t that many of them; Jaeggi 2008 had 35 sub­jects do­ing WM train­ing, while this one has 18 do­ing the adap­tive and an­other 18 do­ing non-adap­tive, and the last of the 55 were pure con­trol. So a lit­tle more than half as many; this is re­flected in some of the weak re­sults, so while rather dis­turbing, this is­n’t a de­fin­i­tive refu­ta­tion or any­thing.

  • the WM task sub­jects did not see any rel­a­tive IQ gains, or much of a gain at all; the IATWMMC (adap­tive arith­metic) group went from 27.3±1 to 31.3±0.7, and the placebo group (non-adap­tive arith­metic) went from 29.1±0.9 to 32.0±0.8. This does­n’t show any no­tice­able differ­ence, the au­thors de­scrib­ing the IQ as ‘prob­a­bly void’.

  • 20 hours of train­ing is more than twice as much train­ing as Jaeggi 2008’s longest group76, so one should not dis­miss this solely on the grounds ‘if only they had trained more’

  • adap­tive arith­metic does­n’t seem like much of a WM task; they did do some n-back­ing (men­tioned briefly) dur­ing the fMRI pre/­post, but not clear why they chose arith­metic over n-back. On the other hand, don’t many n-back­ers use the arith­metic mod­es…?

  • the adap­tive­ness is re­ally im­por­tant; they say the group do­ing non-adap­tive arith­metic was the same as the no-in­ter­ven­tion group on every mea­sure! In­clud­ing ‘a com­plex arith­metic task’

  • one of the key quotes:

    Be­hav­ioral re­sults com­par­ing the com­bined con­trol group, and the IATWMMC group showed a sig­nifi­cantly larger pre- to post- test in­crease for per­for­mance of a com­plex arith­metic task (P = 0.049), for per­for­mance of the let­ter span task (P = 0.002), and for re­verse Stroop in­ter­fer­ence (P = 0.008) in the IATWMMC group. The IATWMMC group showed a sig­nifi­cantly larger pre- to post- test de­crease in cre­ativ­ity test per­for­mance (P = 0.007) (for all the re­sults of the psy­cho­log­i­cal mea­sures, see Ta­ble 1). Also the IATWMMC group showed a sta­tis­ti­cal trend of in­crease in the men­tal ro­ta­tion task (P = 0.064).

  • About the only good news for n-back­ers is that the re­sults were not huge enough to eas­ily sur­vive mul­ti­ple-com­par­i­son cor­rec­tion

    We per­formed sev­eral psy­cho­log­i­cal tests and did not cor­rect for the num­ber of com­par­isons be­tween sta­tis­ti­cal tests, as is al­most al­ways the case with this kind of study. When cor­rected us­ing the Bon­fer­roni cor­rec­tion, even after re­mov­ing the prob­a­bly void tests (RAPM and WAIS arith­metic), the sta­tis­ti­cal value for the effect of IATWMMC on the cre­ativ­ity tests mar­gin­ally sur­passed the thresh­old of P = 0.05 (P = 0.06). Thus, the re­sults should be in­ter­preted with cau­tion un­til repli­cat­ed.

Vartanian 2013

“Work­ing Mem­ory Train­ing Is As­so­ci­ated with Lower Pre­frontal Cor­tex Ac­ti­va­tion in a Di­ver­gent Think­ing Task”; em­pha­sis added:

Work­ing mem­ory (WM) train­ing has been shown to lead to im­prove­ments in WM ca­pac­ity and fluid in­tel­li­gence. Given that di­ver­gent think­ing loads on WM and fluid in­tel­li­gence, we tested the hy­poth­e­sis that WM train­ing would im­prove per­for­mance and mod­er­ate neural func­tion in the Al­ter­nate Uses Task (AUT)-a clas­sic test of di­ver­gent think­ing. We tested this hy­poth­e­sis by ad­min­is­ter­ing the AUT in the func­tional mag­netic res­o­nance imag­ing scan­ner fol­low­ing a short reg­i­men of WM train­ing (ex­per­i­men­tal con­di­tion), or en­gage­ment in a choice re­ac­tion time task not ex­pected to en­gage WM (ac­tive con­trol con­di­tion). Par­tic­i­pants in the ex­per­i­men­tal group ex­hib­ited sig­nifi­cant im­prove­ment in per­for­mance in the WM task as a func­tion of train­ing, as well as a sig­nifi­cant gain in fluid in­tel­li­gence. Al­though the two groups did not differ in their per­for­mance on the AUT, ac­ti­va­tion was sig­nifi­cantly lower in the ex­per­i­men­tal group in ven­tro­lat­eral pre­frontal and dor­so­lat­eral pre­frontal cor­tex-two brain re­gions known to play dis­so­cia­ble and crit­i­cal roles in di­ver­gent think­ing. Fur­ther­more, gain in fluid in­tel­li­gence me­di­ated the effect of train­ing on brain ac­ti­va­tion in ven­tro­lat­eral pre­frontal cor­tex. These re­sults in­di­cate that a short reg­i­men of WM train­ing is as­so­ci­ated with lower pre­frontal ac­ti­va­tion - a marker of neural effi­ciency - in di­ver­gent think­ing.

Non-IQ or non-DNB gains

This sec­tion is for stud­ies that tested non-DNB WM in­ter­ven­tions on IQ, or DNB in­ter­ven­tions on non-IQ prop­er­ties, and mis­cel­la­neous.

Chein 2010

“Ex­pand­ing the mind’s work­space: Train­ing and trans­fer effects with a com­plex work­ing mem­ory span task” (FLOSS im­ple­men­ta­tion); from the in­tro­duc­tion:

In the present study, a novel work­ing mem­ory (WM) train­ing par­a­digm was used to test the mal­leabil­ity of WM ca­pac­ity and to de­ter­mine the ex­tent to which the ben­e­fits of this train­ing could be trans­ferred to other cog­ni­tive skills. Train­ing in­volved ver­bal and spa­tial ver­sions of a com­plex WM span task de­signed to em­pha­size si­mul­ta­ne­ous stor­age and pro­cess­ing re­quire­ments. Par­tic­i­pants who com­pleted 4 weeks of WM train­ing demon­strated sig­nifi­cant im­prove­ments on mea­sures of tem­po­rary mem­o­ry. These WM train­ing ben­e­fits gen­er­al­ized to per­for­mance on the Stroop task and, in a novel find­ing, pro­moted sig­nifi­cant in­creases in read­ing com­pre­hen­sion. The re­sults are dis­cussed in re­la­tion to the hy­poth­e­sis that WM train­ing affects do­main-gen­eral at­ten­tion con­trol mech­a­nisms and can thereby elicit far-reach­ing cog­ni­tive ben­e­fits. Im­pli­ca­tions in­clude the use of WM train­ing as a gen­eral tool for en­hanc­ing im­por­tant cog­ni­tive skills.

While WM train­ing yielded many valu­able ben­e­fits such as in­creased read­ing com­pre­hen­sion, it did not im­prove IQ as mea­sured by an un­speeded Ad­vanced Pro­gres­sive Ma­tri­ces (APM) IQ test;

How­ev­er, such power lim­i­ta­tions do not read­ily ac­count for our fail­ure to repli­cate a trans­fer of WM train­ing ben­e­fits to mea­sures of fluid in­tel­li­gence (as was ob­served by Jaeggi et al., 2008), since we did not find even a trend for im­prove­ment in trained par­tic­i­pants on Raven’s APM. Be­yond sta­tis­ti­cal ex­pla­na­tions, differ­ences in the train­ing par­a­digms used for the two stud­ies may ex­plain the differ­ences in trans­fer effects. The train­ing pro­gram used by Jaeggi et al. (2008) in­volved 400 tri­als per train­ing ses­sion, with a dual n-back train­ing par­a­digm de­signed to em­pha­size bind­ing processes and task man­age­ment. Con­verse­ly, our train­ing par­a­digm in­cluded only 32 tri­als per ses­sion and more heav­ily em­pha­sized main­te­nance in the face of dis­trac­tion. Fi­nal­ly, the seem­ingly con­flict­ing re­sults may be due to differ­ences in in­tel­li­gence test ad­min­is­tra­tion. As was pointed out in a re­cent cri­tique (Moody, 2009), Jaeggi et al. (2008) used atyp­i­cal speeded pro­ce­dures in ad­min­is­ter­ing their tests of fluid in­tel­li­gence, and these al­ter­ations may have con­founded the ap­par­ent effect of WM train­ing on in­tel­li­gence.

Colom 2010

“Im­prove­ment in work­ing mem­ory is not re­lated to in­creased in­tel­li­gence scores” (full text) trained 173 stu­dents on WM tasks (such as the ) with ran­dom­ized diffi­cul­ties, and found no linked IQ im­prove­ment; the IQ tests were “the Ad­vanced Pro­gres­sive Ma­tri­ces Test (APM) along with the ab­stract rea­son­ing (DAT-AR), ver­bal rea­son­ing (DAT-VR), and spa­tial re­la­tions (DAT-SR) sub­tests from the Differ­en­tial Ap­ti­tude Test Bat­tery”. None were speeded as in Jaeggi 2008. Ab­stract:

The ac­knowl­edged high re­la­tion­ship be­tween work­ing mem­ory and in­tel­li­gence sug­gests com­mon un­der­ly­ing cog­ni­tive mech­a­nisms and, per­haps, shared bi­o­log­i­cal sub­strates. If this is the case, im­prove­ment in work­ing mem­ory by re­peated ex­po­sure to chal­leng­ing span tasks might be re­flected in in­creased in­tel­li­gence scores. Here we re­port a study in which 288 uni­ver­sity un­der­grad­u­ates com­pleted the odd num­bered items of four in­tel­li­gence tests on time 1 and the even num­bered items of the same tests one month later (time 2). In be­tween, 173 par­tic­i­pants com­pleted three ses­sions, sep­a­rated by ex­actly one week, com­pris­ing ver­bal, nu­mer­i­cal, and spa­tial short­-term mem­ory (STM) and work­ing mem­ory (WMC) tasks im­pos­ing high pro­cess­ing de­mands (STM-WMC group). 115 par­tic­i­pants also com­pleted three ses­sions, sep­a­rated by ex­actly one week, but com­pris­ing ver­bal, nu­mer­i­cal, and spa­tial sim­ple speed tasks (pro­cess­ing speed, PS, and at­ten­tion, ATT) with very low pro­cess­ing de­mands (PS-ATT group). The main find­ing re­veals in­creased scores from the pre-test to the post-test in­tel­li­gence ses­sion (more than half a stan­dard de­vi­a­tion on av­er­age). How­ev­er, there was no differ­en­tial im­prove­ment on in­tel­li­gence be­tween the STM-WMC and PS-ATT groups.

Com­men­ta­tors on the ML dis­cus­sion crit­i­cized the study for:

  1. Not us­ing DNB it­self
  2. ap­par­ently lit­tle train­ing time on the WM tasks (3 ses­sions over weeks, each of un­clear du­ra­tion)
  3. the ran­dom­iza­tion of diffi­culty (as op­posed to DNB’s adap­tive­ness)
  4. the large in­crease in scores on the WM tasks over the 3 ses­sions (sug­gest­ing grow­ing fa­mil­iar­ity than real chal­lenge & growth)
  5. and the sta­tis­ti­cal ob­ser­va­tion that if IQ gains were lin­ear with train­ing and started small then 173 par­tic­i­pants is not enough to ob­serve with con­fi­dence any im­prove­ments.

Loosli et al 2011

“Work­ing mem­ory train­ing im­proves read­ing processes in typ­i­cally de­vel­op­ing chil­dren”, Loosli, Buschkuehl, Per­rig, and Jaeg­gi:

The goal of this study was to in­ves­ti­gate whether a brief cog­ni­tive train­ing in­ter­ven­tion re­sults in a spe­cific per­for­mance in­crease in the trained task, and whether there are trans­fer effects to other non­trained mea­sures. A com­put­er­ized, adap­tive work­ing mem­ory in­ter­ven­tion was con­ducted with 9- to 11-year-old typ­i­cally de­vel­op­ing chil­dren. The chil­dren con­sid­er­ably im­proved their per­for­mance in the trained work­ing mem­ory task. Ad­di­tion­al­ly, com­pared to a matched con­trol group, the ex­per­i­men­tal group sig­nifi­cantly en­hanced their read­ing per­for­mance after train­ing, pro­vid­ing fur­ther ev­i­dence for shared processes be­tween work­ing mem­ory and read­ing.

This is show­ing con­nec­tion to use­ful tasks, but not for show­ing any gain to IQ. The differ­ence in score im­prove­ment be­tween groups was small, half a point, and the train­ing pe­riod fairly short; the au­thors write:

Due to the short train­ing time, we did not ex­pect large effects on Gf (cf. Jaeggi et al., 2008), also since two other stud­ies that trained ADHD chil­dren ob­served trans­fer effects on Gf only after 5 weeks in­volv­ing ses­sions of 40 min­utes each (K­ling­berg et al., 2002, 2005).

In ad­di­tion, the same group failed to show trans­fer on Gf with a shorter train­ing (Thorell et al., 2008). Thus, con­sid­er­ing that our train­ing in­ter­ven­tion was merely 10 ses­sions long, our lack of trans­fer to Gf is hardly sur­pris­ing; al­though there is now re­cent ev­i­dence that trans­fer to Gf is pos­si­ble with very lit­tle train­ing time (; poster). Our re­sults, how­ev­er, are com­pa­ra­ble to those of Chein and Mor­ri­son (2010), who also trained their par­tic­i­pants on a com­plex WM task and found no trans­fer to Gf.

(Sim­i­lar stud­ies have also found im­prove­ment in read­ing skills after WM train­ing, eg Dahlin 2011 and Shi­ran & Breznitz 2011, but I do not be­lieve oth­ers used n-back or looked for pos­si­ble IQ gain­s.)

Nutley 2011

“Gains in fluid in­tel­li­gence after train­ing non-ver­bal rea­son­ing in 4-year-old chil­dren: a con­trolled, ran­dom­ized study”, Sis­sela Bergman Nut­ley et al:

Fluid in­tel­li­gence (Gf) pre­dicts per­for­mance on a wide range of cog­ni­tive ac­tiv­i­ties, and chil­dren with im­paired Gf often ex­pe­ri­ence aca­d­e­mic diffi­cul­ties. Pre­vi­ous at­tempts to im­prove Gf have been ham­pered by poor con­trol con­di­tions and sin­gle out­come mea­sures77. It is thus still an open ques­tion whether Gf can be im­proved by train­ing. This study in­cluded 4-year-old chil­dren (N = 101) who per­formed com­put­er­ized train­ing (15 min/­day for 25 days) of ei­ther non-ver­bal rea­son­ing, work­ing mem­o­ry, a com­bi­na­tion of both, or a placebo ver­sion of the com­bined train­ing. Com­pared to the placebo group, the non-ver­bal rea­son­ing train­ing group im­proved sig­nifi­cantly on Gf when analysed as a la­tent vari­able of sev­eral rea­son­ing tasks. Smaller gains on prob­lem solv­ing tests were seen in the com­bi­na­tion train­ing group. The group train­ing work­ing mem­ory im­proved on mea­sures of work­ing mem­o­ry, but not on prob­lem solv­ing tests. This study shows that it is pos­si­ble to im­prove Gf with train­ing, which could have im­pli­ca­tions for early in­ter­ven­tions in chil­dren.


  1. The WM tasks were not n-back:

    “The WM train­ing was the same as de­scribed in Thorell et al. (2009) de­vel­oped by Cogmed Sys­tems Inc. There were seven differ­ent ver­sions of vi­suo-s­pa­tial WM tasks, out of which three were trained every day on a ro­tat­ing sched­ule. Briefly, the tasks all con­sisted of a num­ber of an­i­mated fig­ures pre­sented in differ­ent set­tings (e.g. swim­ming in a pool, rid­ing on a roller­coast­er). Some of the fig­ures (s­tart­ing with two fig­ures and then in­creas­ing in num­ber de­pend­ing on the child’s per­for­mance) made a sound and changed colour dur­ing a short time pe­ri­od. The task then con­sisted of re­mem­ber­ing which fig­ures had changed colour and in what or­der this had oc­curred.”

  2. The mag­ni­tude of Gf in­crease was not sus­pi­ciously large:

    “The NVR train­ing group showed trans­fer both when this was es­ti­mated with sin­gle tests, as well as when Gf was mea­sured as a la­tent vari­able. The mag­ni­tude of this im­prove­ment was ap­prox­i­mately 8% (com­pared to the placebo group) which is com­pa­ra­ble with pre­vi­ously re­ported gains of Gf of 5-13.5% (Hamers et al., 1998; Jaeggi et al., 2008; Klauer & Willmes, 2002; Stankov, 1986).”

  3. There are some pos­si­ble coun­ter-ar­gu­ments to gen­er­al­iz­ing the lack of Gf gains in the WM-only group, mostly re­lated to the young age:

    "This could mean that WM is not a lim­it­ing fac­tor for 4-year-old chil­dren solv­ing rea­son­ing prob­lems such as Raven’s CPM and Block De­sign. The mod­er­ate cor­re­la­tions be­tween the Grid Task and the rea­son­ing tests (be­tween 0.3 and 0.6, see Ta­ble 1) point to the some­what coun­ter­in­tu­itive con­clu­sion that cor­re­la­tion be­tween two un­der­ly­ing abil­i­ties is not a suffi­cient pre­dic­tor to de­ter­mine amount of trans­fer of train­ing effects be­tween these abil­i­ties. A sim­i­lar con­clu­sion was drawn after the lack of train­ing effects on WM after train­ing in­hibitory func­tions (Thorell et al., 2009). In that study WM ca­pac­ity cor­re­lated with per­for­mance on the in­hibitory tasks at base­line (R = 0.3). An imag­ing study also showed that per­for­mance on a WM grid task and in­hibitory tasks ac­ti­vate over­lap­ping parts of the cor­tex (). In­hibitory train­ing im­proved per­for­mance on the trained tasks, yet there was no trans­fer seen on WM tasks. The prin­ci­ples gov­ern­ing the type of cog­ni­tive train­ing that will trans­fer are still un­clear and pose an im­por­tant ques­tion for fu­ture stud­ies.

    One way to find these prin­ci­ples may be through un­der­stand­ing the neural mech­a­nisms of train­ing. For ex­am­ple, WM train­ing in 4-year-olds might have a more pro­nounced effect on the pari­etal lobe, com­pared to the less ma­ture frontal lobe. If the trans­fer to Gf is de­pen­dent on pre­frontal func­tions, it may ex­plain the lack of trans­fer from WM train­ing to Gf in 4-year-olds. In other words, trans­fer effects may differ with the pro­gres­sion of de­vel­op­men­t."

Zhao et al 2011

“Effect of up­dat­ing train­ing on fluid in­tel­li­gence in chil­dren”, Chi­nese Sci­ence Bul­letin:

Re­cent stud­ies have in­di­cated that work­ing mem­ory (WM) train­ing can im­prove fluid in­tel­li­gence. How­ev­er, these ear­lier stud­ies con­fused the im­pact of WM stor­age and cen­tral ex­ec­u­tive func­tion on the effects of train­ing. The cur­rent study used the run­ning mem­ory task to train the up­dat­ing abil­ity of [33] 9-11 year-old chil­dren us­ing a dou­ble-blind con­trolled de­sign. The re­sults re­vealed that chil­dren’s fluid in­tel­li­gence was sig­nifi­cantly im­proved by mem­o­ry-up­dat­ing train­ing. Over­all, our find­ings sug­gest that the in­crease in fluid in­tel­li­gence achieved with WM train­ing is re­lated to im­prov­ing cen­tral ex­ec­u­tive func­tion.

Roughan & Hadwin 2011

“The im­pact of work­ing mem­ory train­ing in young peo­ple with so­cial, emo­tional and be­hav­ioural diffi­cul­ties”, Laura Roughan & Julie A. Had­win 2011:

This study ex­am­ined the im­pact of a work­ing mem­ory (WM) train­ing pro­gramme on mea­sures of WM, IQ, be­hav­ioural in­hi­bi­tion, self­-re­port test and trait anx­i­ety and teacher re­ported emo­tional and be­hav­ioural diffi­cul­ties and at­ten­tional con­trol be­fore and after WM train­ing and at a 3 month fol­low-up. The WM train­ing group (N=7) showed sig­nifi­cantly bet­ter post-train­ing on mea­sures of IQ, in­hi­bi­tion, test anx­i­ety and teacher-re­ported be­hav­iour, at­ten­tion and emo­tional symp­toms, com­pared with a non-in­ter­ven­tion pas­sive con­trol group (N=8). Group differ­ences in WM were also ev­i­dent at fol­low-up. The re­sults in­di­cated that WM train­ing has some po­ten­tial to be used to re­duce the de­vel­op­ment of school re­lated diffi­cul­ties and as­so­ci­ated men­tal health prob­lems in young peo­ple. Fur­ther re­search us­ing larger sam­ple sizes and mon­i­tor­ing over a longer time pe­riod is needed to repli­cate and ex­tend these re­sults.

The WM train­ing was done us­ing Cogmed; it’s un­clear whether the Cogmed tasks use DNB or not (they seem to have sim­i­lar tasks avail­able in it, at least), but the study did find IQ gains:

Con­sid­er­ing T1 T2 IQ differ­ence scores, the analy­sis re­vealed a sig­nifi­cant group effect with a large ES (F(1,14) = 10.37, p<.01, n = 0.44); the in­ter­ven­tion group showed in­creased IQ differ­ence scores (N = 7, mean=5.36, SD = 6.52, range= -2.5 to 17.5) com­pared with the con­trol group (N = 7, mean=-6.35, SD = 7.21, range = -15 to 5). T1 T3 analy­ses in­di­cated that the T1 T3 differ­ence was not sig­nifi­cant (see Fig. 1).

Note the means as com­pared with the stan­dard de­vi­a­tion; these are very trou­bled young peo­ple.

Brehmer et al 2012

“Work­ing-mem­ory train­ing in younger and older adults: train­ing gains, trans­fer, and main­te­nance”:

Work­ing mem­ory (WM), a key de­ter­mi­nant of many high­er-order cog­ni­tive func­tions, de­clines in old age. Cur­rent re­search at­tempts to de­velop process-spe­cific WM train­ing pro­ce­dures, which may lead to gen­eral cog­ni­tive im­prove­ment. Adap­tiv­ity of the train­ing as well as the com­par­i­son of train­ing gains to per­for­mance changes of an ac­tive con­trol group are key fac­tors in eval­u­at­ing the effec­tive­ness of a spe­cific train­ing pro­gram. In the present study, 55 younger adults (20-30 years of age) and 45 older adults (60-70 years of age) re­ceived 5 weeks of com­put­er­ized train­ing on var­i­ous spa­tial and ver­bal WM tasks. Half of the sam­ple re­ceived adap­tive train­ing (i.e., in­di­vid­u­ally ad­justed task diffi­cul­ty), whereas the other half-worked on the same task ma­te­r­ial but on a low task diffi­culty level (ac­tive con­trol­s). Per­for­mance was as­sessed us­ing cri­te­ri­on, near-trans­fer, and far-trans­fer tasks be­fore train­ing, after 5 weeks of in­ter­ven­tion, as well as after a 3-month fol­low-up in­ter­val. Re­sults in­di­cate that (a) adap­tive train­ing gen­er­ally led to larger train­ing gains t/han low-level prac­tice, (b) train­ing and trans­fer gains were some­what greater for younger than for older adults in some tasks, but com­pa­ra­ble across age groups in other tasks, (c) far-trans­fer was ob­served to a test on sus­tained at­ten­tion and for a self­-rat­ing scale on cog­ni­tive func­tion­ing in daily life for both young and old, and (d) train­ing gains and trans­fer effects were main­tained across the 3-month fol­low-up in­ter­val across age.

Used Cogmed, which Jaeggi says is not dual n-back.


One fas­ci­nat­ing psy­chol­ogy re­sult is that strongly peo­ple can im­prove their mem­ory (and pos­si­bly N-back per­for­mance) by sim­ply tak­ing 30 sec­onds and flick­ing (“”) their eyes left and right (for a sum­ma­ry, see “A quick eye­-ex­er­cise can im­prove your per­for­mance on mem­ory tests (but only if you’re right-hand­ed)”).

Ver­sion 4.5 of Brain Work­shop in­tro­duced a sac­cad­ing fea­ture: a dot al­ter­nates sides of the screen and one is to fol­low it with one’s eyes. You ac­ti­vate it by press­ing ‘e’ while in fullscreen mode (set­ting WINDOW_FULLSCREEN = True in the con­fig­u­ra­tion file). It may or may not a bad idea to al­ter­nate rounds of N-back with rounds of sac­cad­ing. At my re­quest, sac­cad­ing logs are now kept by BW so at some point in the fu­ture, it should be pos­si­ble to re­quest logs from users and see whether sac­cad­ing in gen­eral cor­re­lates with N-back per­for­mance; I per­son­ally ran­dom­ized use of sac­cad­ing, but saw no ben­e­fits (see next sec­tion).

Ashirgo writes that her pre­vi­ous ad­vice en­com­passes this eye­-move­ment re­sult; Pheonexia re­ports that after try­ing the sac­cad­ing be­fore a BW ses­sion, he “per­formed bet­ter than I ever have be­fore”.

The most re­cent study on this effect seems to be “Eye move­ments en­hance mem­ory for in­di­vid­u­als who are strongly right-handed and harm it for in­di­vid­u­als who are not”. It says:

Sub­jects who make repet­i­tive sac­cadic eye move­ments be­fore a mem­ory test sub­se­quently ex­hibit su­pe­rior re­trieval in com­par­i­son with sub­jects who do not move their eyes. It has been pro­posed that eye move­ments en­hance re­trieval by in­creas­ing in­ter­ac­tion of the left and right cere­bral hemi­spheres. To test this, we com­pared the effect of eye move­ments on sub­se­quent re­call (Ex­per­i­men­t1) and recog­ni­tion (Ex­per­i­men­t2) in two groups thought to differ in base­line de­gree of hemi­spheric in­ter­ac­tion-in­di­vid­u­als who are strongly right-handed (SR) and in­di­vid­u­als who are not (n­SR). For SR sub­jects, who nat­u­rally may ex­pe­ri­ence less hemi­spheric in­ter­ac­tion than nSR sub­jects, eye move­ments en­hanced re­trieval. In con­trast, de­pend­ing on the mea­sure, eye move­ments were ei­ther in­con­se­quen­tial or even detri­men­tal for nSR sub­jects. These re­sults par­tially sup­port the hemi­spheric in­ter­ac­tion ac­count, but de­mand an amend­ment to ex­plain the harm­ful effects of eye move­ments for nSR in­di­vid­u­als.

(Note that very im­por­tant caveat: this is a use­ful tech­nique only for strongly right-handed peo­ple; weak right­ies and left­ies are out­right harmed by this tech­nique.)

See also “In­ter­hemi­spheric In­ter­ac­tion and Sac­cadic Hor­i­zon­tal Eye Move­ments: Im­pli­ca­tions for Episodic Mem­o­ry, EMDR, and PTSD; “The effi­cacy and psy­chophys­i­o­log­i­cal cor­re­lates of du­al-at­ten­tion tasks in eye move­ment de­sen­si­ti­za­tion and re­pro­cess­ing (EMDR)”; “Hor­i­zon­tal sac­cadic eye move­ments en­hance the re­trieval of land­mark shape and lo­ca­tion in­for­ma­tion”; “Re­duced mis­in­for­ma­tion effects fol­low­ing sac­cadic bi­lat­eral eye move­ments”; “Is sac­cade-in­duced re­trieval en­hance­ment a po­ten­tial means of im­prov­ing eye­wit­ness ev­i­dence?”


Brain Work­shop now has log­ging of sac­cad­ing im­ple­ment­ed; this was added at my re­quest to make ex­per­i­ment­ing with sac­cad­ing eas­ier, since you can’t com­pare scores un­less you know when you were sac­cad­ing or not. After this was added (thanks Jonathan etc), I be­gan to ran­dom­ize each day to sac­cad­ing or not-sac­cad­ing be­fore rounds with a coin flip. Blind­ing is im­pos­si­ble, so I did noth­ing about that. After 158 rounds over roughly 35 days be­tween 10 Sep­tem­ber and 2012-11-05, the re­sult is: no differ­ence. Not even close. So ap­par­ently though I am strongly right-handed as the orig­i­nal study’s mem­ory effect re­quired, sac­cad­ing makes no differ­ence to my n-back per­for­mance.


My BW data had to be parsed by hand and some Emacs macros be­cause I could­n’t fig­ure out a nice clean pro­gram­matic way to parse it and spit out scores divvied by whether they were on a sac­cade on or off day (so if you want to repli­cate my analy­sis, you’ll have to do that your­self). The analy­sis78 us­ing BEST re­veals a differ­ence of less than 1% right (+0.4%) per round, and the es­ti­mates of effect size are neg­a­tive al­most as often as they are pos­i­tive:

Bayesian MCMC es­ti­mates of differ­ence in sac­cad­ing and non-sac­cad­ing scores

Since there’s hardly any ev­i­dence even though this looks like plenty of data, I think I’ll stop do­ing sac­cad­ing. I can only speak for my­self, so I would be pleased if other right-handed n-back­ers could adopt a sim­i­lar pro­ce­dure and see whether per­haps I am an ex­cep­tion.


“Sleep Ac­cel­er­ates the Im­prove­ment in Work­ing Mem­ory Per­for­mance”, Kuriyama 2008:

Work­ing mem­ory (WM) per­for­mance, which is an im­por­tant fac­tor for de­ter­min­ing prob­lem-solv­ing and rea­son­ing abil­i­ty, has been firmly be­lieved to be con­stant. How­ev­er, re­cent find­ings have demon­strated that WM per­for­mance has the po­ten­tial to be im­proved by repet­i­tive train­ing. Al­though var­i­ous skills are re­ported to be im­proved by sleep, the ben­e­fi­cial effect of sleep on WM per­for­mance has not been clar­i­fied. Here, we show that im­prove­ment in WM per­for­mance is fa­cil­i­tated by post­train­ing nat­u­ral­is­tic sleep. A spa­tial vari­ant of the n-back WM task was per­formed by 29 healthy young adults who were as­signed ran­domly to three differ­ent ex­per­i­men­tal groups that had differ­ent time sched­ules of repet­i­tive n-back WM task ses­sions, with or with­out in­ter­ven­ing sleep. In­ter­group and in­ter­s­es­sion com­par­isons of WM per­for­mance (ac­cu­racy and re­sponse time) pro­files showed that n-back ac­cu­racy after post­train­ing sleep was sig­nifi­cantly im­proved com­pared with that after the same pe­riod of wake­ful­ness, in­de­pen­dent of sleep tim­ing, sub­jec­t’s vig­i­lance lev­el, or cir­ca­dian in­flu­ences. On the other hand, re­sponse time was not in­flu­enced by sleep or repet­i­tive train­ing sched­ules. The present study in­di­cates that im­prove­ment in n-back ac­cu­ra­cy, which could re­flect WM ca­pac­i­ty, es­sen­tially ben­e­fits from post­train­ing sleep.

(In this test, the base­line/un­prac­ticed per­for­mance of the two groups was the same; but the sched­ule in which sub­jects trained at 10 PM and went to bed re­sulted in greater im­prove­ments in per­for­mance than sched­ules in which sub­jects trained when they got up at 8 AM and went to bed ~10 PM.)

Lucid dreaming

, pi­o­neer of writes79

“Why then is CNS ac­ti­va­tion nec­es­sary for lu­cid dream­ing? Ev­i­dently the high level of cog­ni­tive func­tion in­volved in lu­cid dream­ing re­quires a cor­re­spond­ingly high level of neu­ronal ac­ti­va­tion. In terms of Antrobus’s (1986) adap­ta­tion of An­der­son’s (1983) ACT* model of cog­ni­tion to dream­ing, work­ing mem­ory ca­pac­ity is pro­por­tional to cog­ni­tive ac­ti­va­tion, which in turn is pro­por­tional to cor­ti­cal ac­ti­va­tion. Be­com­ing lu­cid re­quires an ad­e­quate level of work­ing mem­ory to ac­tive the presleep in­ten­tion to rec­og­nize that one is dream­ing. This level of ac­ti­va­tion is ap­par­ently not al­ways avail­able dur­ing sleep but nor­mally only dur­ing pha­sic REM.”

has ap­par­ently spec­u­lated80 that WM and the pre­frontal cor­tex is par­tially de-ac­ti­vated dur­ing REM sleep and this is why dream­ers do not re­al­ize they are dream­ing - the same re­gion that n-back tasks ac­ti­vate.81 The sug­ges­tion then goes that n-back train­ing will en­able greater dream recog­ni­tion & re­call, which are cru­cial skills for any would-be lu­cid dream­er. A num­ber of peo­ple have re­ported only dreams and lu­cid dreams as the re­sult of n-back train­ing (eg. Boris & Michael).

On the other hand, I have seen anec­do­tal re­ports that any in­tense men­tal ex­er­cise or learn­ing causes in­creased dream­ing, even if the ex­er­cise is do­main-spe­cific (eg. the fa­mous ) or just mem­o­riza­tion (as in use of Mnemosyne for ), and LaBerge also re­marks (pg 165 of Ex­plor­ing the World of Lu­cid Dream­ing):

Most peo­ple as­sume that a ma­jor func­tion of sleep­ing and dream­ing is rest and re­cu­per­a­tion. This pop­u­lar con­cep­tion has been up­held by re­search. Thus, for hu­mans, phys­i­cal ex­er­cise leads to more sleep, es­pe­cially delta sleep. Growth hor­mone, which trig­gers growth in chil­dren and the re­pair of stressed tis­sues, is re­leased in delta sleep. On the other hand, men­tal ex­er­cise or emo­tional stress ap­pears to re­sult in in­creases in REM sleep and dream­ing.


Gen­eral cog­ni­tive fac­tors like work­ing mem­ory and pro­cess­ing speed (& per­cep­tual pro­cess­ing82) are traits that peak in early adult hood and then de­cline over a life­time; the fol­low­ing im­age was adapted by Giz­modo from a study of age-re­lated de­cline, “Mod­els of vi­su­ospa­tial and ver­bal mem­ory across the adult life span”83. The units are , units of stan­dard de­vi­a­tions (so for the 80 year olds to be two full units be­low the 20 year olds in­di­cates a pro­found fall in the av­er­ages84); the first im­age is from Park et al 2002:

Graph of mul­ti­ple men­tal traits due to age-re­lated de­cline (in stan­dard de­vi­a­tions) Schaie 1996, In­tel­lec­tual De­vel­op­ment…

A in the Cam­bridge brain-train­ing study found “Age, was by far the most sig­nifi­cant pre­dic­tor of per­for­mance, with the mean scores of in­di­vid­u­als in their 60s ~1.7 SDs be­low those in their early 20s (Fig­ure 4a). (Note, in in­tel­li­gence test­ing, 1 SD is equiv­a­lent to 15 IQ points).” These de­clines in rea­son­ing affect valu­able re­al-world ac­tiv­i­ties like per­sonal fi­nance85, and sim­ple every­day ques­tions:

from Agar­wal et al 2009, “The Age of Rea­son: Fi­nan­cial De­ci­sions over the Life-Cy­cle with Im­pli­ca­tions for Reg­u­la­tion”

These re­sults may be sur­pris­ing be­cause some stud­ies did not find such dra­matic de­cli­nes, but ap­par­ently part of the de­cline can be hid­den by prac­tice effects86, and they are con­sis­tent with other re­sults like the life­long changes in ( in & 87, the lat­ter de­cline pos­si­bly ). Lon­gi­tu­di­nal stud­ies are pes­simistic, find­ing de­clines early on, in one’s 40s (Sing-Manoux et al 2011). The degra­da­tion of white mat­ter and its effects on episodic mem­ory re­trieval have been us­ing . An­other 2011 study test­ing 2000 in­di­vid­u­als be­tween 18 and 60 found that “Top per­for­mances in some of the tests were ac­com­plished at the age of 22. A no­table de­cline in cer­tain mea­sures of ab­stract rea­son­ing, brain speed and in puz­zle-solv­ing be­came ap­par­ent at 27.”88 (Of course, like the pre­vi­ous study, a cor­re­la­tion over many in­di­vid­u­als of vary­ing ages is not as good as hav­ing a se­ries of per­for­mance mea­sure­ments for one ag­ing in­di­vid­ual. But time will cure that fault, hope­ful­ly.) The of this Salt­house study says:

…Re­sults from three meth­ods of es­ti­mat­ing retest effects in this pro­ject, to­gether with re­sults from stud­ies com­par­ing non-hu­man an­i­mals raised in con­stant en­vi­ron­ments and from stud­ies ex­am­in­ing neu­ro­bi­o­log­i­cal vari­ables not sus­cep­ti­ble to retest effects, con­verge on a con­clu­sion that some as­pects of age-re­lated cog­ni­tive de­cline be­gin in healthy ed­u­cated adults when they are in their 20s and 30s.

From the op­ti­mistic per­spec­tive, Salt­house tested For­tune 500 CEOs and found that their mem­ber­ship by av­er­age age did­n’t start drop­ping un­til their 60s, sug­gest­ing that they re­mained rea­son­ably men­tally sharp or were, in prac­tice, com­pen­sat­ing for the many in­sults of age;89 this way of think­ing has ob­vi­ous flaws for the rest of us.

There are a num­ber of re­sults in­di­cat­ing that the el­der­ly, per­haps be­cause they have so much sev­erer cog­ni­tive deficits than the young, re­spond bet­ter to treat­ment. (This is com­mon in nootrop­ics, find­ing that some­thing does not work in the young but does in the el­der­ly: eg. cre­a­tine.) IQ gains in young adults are diffi­cult and min­i­mal even in Jaeggi 2008, but older adults im­prove about as much as young adults in Brehmer et al 2012 and in­struct­ing older adults to think aloud dur­ing an IQ test boosts scores (yet not younger adults)90, and train­ing >65-year olds in one adap­tive WM task sim­i­lar to SNB lead to gains of ~6 IQ points on the which were still present 8 months lat­er; “Work­ing Mem­ory Train­ing in Older Adults: Ev­i­dence of Trans­fer and Main­te­nance Effects” & Car­retti et al 2012 makes for in­ter­est­ing read­ing91:

Few stud­ies have ex­am­ined work­ing mem­ory (WM) train­ing-re­lated gains and their trans­fer and main­te­nance effects in older adults. This present re­search in­ves­ti­gates the effi­cacy of a ver­bal WM train­ing pro­gram in adults aged 65-75 years, con­sid­er­ing spe­cific train­ing gains on a ver­bal WM (cri­te­ri­on) task as well as trans­fer effects on mea­sures of vi­su­ospa­tial WM, short­-term mem­o­ry, in­hi­bi­tion, pro­cess­ing speed, and fluid in­tel­li­gence. Main­te­nance of train­ing ben­e­fits was eval­u­ated at 8-month fol­low-up. Trained older adults showed higher per­for­mance than did con­trols on the cri­te­rion task and main­tained this ben­e­fit after 8 months. Sub­stan­tial gen­eral trans­fer effects were found for the trained group, but not for the con­trol one. Trans­fer main­te­nance gains were found at fol­low-up, but only for fluid in­tel­li­gence and pro­cess­ing speed tasks. The re­sults are dis­cussed in terms of cog­ni­tive plas­tic­ity in older adults.

For more on ag­ing and the brain, rec­om­mends read­ing

Hed­den T, Gabrieli JD. Nat Rev Neu­ro­science. 2004 Feb;5(2):87-96. ‘In­sights into the ag­ing mind: a view from cog­ni­tive neu­ro­science’. PMID 14735112, which is avail­able as full text from this link: http://brainy­be­hav­ior.­com/blog/w­p-con­tent/u­pload­s/2007/11/ag­ing­brain.pdf. I can­not rec­om­mend this pa­per highly enough. Ad­di­tion­al­ly, the Salt Cog­ni­tive Ag­ing Lab­o­ra­to­ry, which over­sees the Vir­ginia Cog­ni­tive Ag­ing Project (VCAP) at the Uni­ver­sity of Vir­ginia, is the pre­mier fa­cil­ity in the US (and ar­guably the world) un­der­tak­ing ac­tive, lon­gi­tu­di­nal stud­ies of ag­ing. The VCAP study has done com­pre­hen­sive cog­ni­tive as­sess­ments in adults rang­ing from 18 to 98 years of age. Ap­prox­i­mately 3,800 adults have par­tic­i­pated in their three­-ses­sion (6-8 hour) as­sess­ment at least on­ce, with about 1,600 par­tic­i­pat­ing at least twice, and about 450 of them par­tic­i­pat­ing three or more times. The data from this project have served as the ba­sis for a ver­i­ta­ble cor­nu­copia of sci­en­tific pub­li­ca­tions which are avail­able in the Re­sources Sec­tion of their web­site http://­fac­ul­ty.vir­gini­a.e­du/co­gage/links/pub­li­ca­tions/. Nearly 200 pa­pers on the cog­ni­tive im­pact of ag­ing are avail­able free of charge on their web­site. It is nec­es­sary to reg­is­ter with your name and email ad­dress to ac­cess the pa­pers, but it is well worth it.


Oth­ers to fol­low up on:

There are sev­eral stud­ies show­ing that work­ing mem­ory and in­tel­li­gence are strongly re­lat­ed. How­ev­er, work­ing mem­ory tasks re­quire si­mul­ta­ne­ous pro­cess­ing and stor­age, so the causes of their re­la­tion­ship with in­tel­li­gence are cur­rently a mat­ter of dis­cus­sion. The present study ex­am­ined the si­mul­ta­ne­ous re­la­tion­ships among short­-term mem­ory (STM), work­ing mem­ory (WM), and gen­eral in­tel­li­gence (g). Two hun­dred and eight par­tic­i­pants per­formed six ver­bal, quan­ti­ta­tive, and spa­tial STM tasks, six ver­bal, quan­ti­ta­tive, and spa­tial WM tasks, and eight tests mea­sur­ing flu­id, crys­tal­lized, spa­tial, and quan­ti­ta­tive in­tel­li­gence. Es­pe­cial care is taken to avoid mis­rep­re­sent­ing the re­la­tions among the con­structs be­ing stud­ied be­cause of spe­cific task vari­ance. Struc­tural equa­tion mod­el­ing (SEM) re­sults re­vealed that (a) WM and g are (al­most) iso­mor­phic con­structs, (b) the iso­mor­phism van­ishes when the stor­age com­po­nent of WM is par­tialed out, and (c) STM and WM (with its stor­age com­po­nent par­tialed out) pre­dict g.

  • Colom et al. “Gen­eral in­tel­li­gence and mem­ory span: Ev­i­dence for a com­mon neu­roanatomic frame­work”; Cog­ni­tive Neu­ropsy­chol­ogy, Vol­ume 24, Is­sue 2007-12-08 , pages 867 - 878

Gen­eral in­tel­li­gence (g) is highly cor­re­lated with work­ing-mem­ory ca­pac­ity (WMC). It has been ar­gued that these cen­tral psy­cho­log­i­cal con­structs should share com­mon neural sys­tems. The present study ex­am­ines this hy­poth­e­sis us­ing struc­tural mag­netic res­o­nance imag­ing to de­ter­mine any over­lap in brain ar­eas where re­gional grey mat­ter vol­umes are cor­re­lated to mea­sures of gen­eral in­tel­li­gence and to mem­ory span. In nor­mal vol­un­teers (N = 48) the re­sults (p < .05, cor­rected for mul­ti­ple com­par­isons) in­di­cate that a com­mon anatomic frame­work for these con­structs im­pli­cates mainly frontal grey mat­ter re­gions be­long­ing to Brod­mann area (BA) 10 (right su­pe­rior frontal gyrus and left mid­dle frontal gyrus) and, to a lesser de­gree, the right in­fe­rior pari­etal lob­ule (BA 40). These find­ings sup­port the nu­clear role of a dis­crete pari­eto-frontal net­work.



There are many free im­ple­men­ta­tions in Flash etc. on­line:







See also Lu­cas Charles’s Au­gust 2011 re­view of 6 An­droid DNB apps.






Offline N-back

You can play N-back in the real world, with­out a com­put­er, if you like. See the ML thread “Non-elec­tronic game ver­sion of N-back task” and the Snap­Back rules. Jonathan Toomim points out that N-back can be eas­ily done with a deck of cards alone, and the FAQ’s au­thor sug­gests a sim­ple men­tal arith­metic rou­tine suit­able for med­i­ta­tion that is much like SNB.

What else can I do?

may in­crease WM, al­though it re­mains un­clear whether the per­for­mance gains per­sist after­wards. See Bog­gio et al 2005, Fregni et al 2005, Ohn et al 2007, Bog­gio et al 2008, Jo et al 2009, An­drews et al 2011, Za­ehle et al 2011, Berry­hill & Jones 2012, Tseng et al 2012, Mar­tin et al 2014, Matzen & Trumbo 2014, Car­valho et al 2014, Moreno et al 2015, , de Put­ter et al 2015, Choe et al 2016 (but also Mar­shall et al 2005, Steen­ber­gen et al 2015, Hoy et al 2015, van Wes­sel et al 2015, Rethans et al 2015, , Nils­son et al 2017, , an in­for­mal in­com­plete tDCS-DNB ex­per­i­ment, and the meta-analy­sis Hill et al 2015).

Fo­rum mem­bers have rec­om­mended a num­ber of other things for gen­eral men­tal fit­ness:


(see for the au­thor’s own ex­pe­ri­ences with them), may help boost per­for­mance. The re­la­tion of caffeine to learn­ing & mem­ory is com­pli­cat­ed; for now, see the thread on it or my Nootrop­ics page.


A use­ful phar­ma­ceu­ti­cal is ; The­Q17 men­tions that “Per­son­al­ly, I have found pirac­etam be quite use­ful in help­ing me stay alert and fo­cused dur­ing long study hours or do­ing re­dun­dant tasks.” Other mem­bers also swear by pirac­etam+­choline.

The au­thor of this FAQ re­ports that pirac­etam and helped re­duced men­tal fa­tigue and gave a small (~10%) in­crease in his D4B score.


Re­ece writes

I’ve tried [a chem­i­cal ex­tracted from an herb] (ac­tu­ally been us­ing it for about a year now) and it is quite effec­tive for both lu­cid dream­ing and in­creas­ing dream re­call if taken shortly be­fore bed, not to men­tion the other ben­e­fits you’d ex­pect from a po­tent in­hibitor. I haven’t had any­thing in the way of neg­a­tive side effects when I’ve stuck to a 5 day/week dosage of 200m­cg.

I’ve never tried pirac­etam, how­ever felt like a placebo when com­pared to the ben­e­fits I’ve re­ceived from hu­perzine A. At larger dos­es, I’ve found hu­perzine A to be far more pow­er­ful than any nootropic I’ve ever tried (haven’t tried any pre­scrip­tion meds such as ), how­ever the side effects such as blurry vi­sion and light-head­ed­ness weren’t some­thing I could tol­er­ate.

He fur­ther com­pared their effects:

I found Oxirac­etam to have a some­what “speedy” effec­t—you would cer­tainly know you took some­thing if some­one slipped that in your drink! As for effects, Oxirac­etam seemed to help most with ver­bal flu­ency (au­di­tory work­ing mem­o­ry?) and cre­ativ­i­ty. Hu­perzine helped more with work­ing mem­ory al­though it did­n’t have some of the in­ter­est­ing effects Oxirac­etam had on cre­ativ­i­ty, nor the speedy rush that some­times seemed like a pow­er­ful mo­ti­va­tor to get work done.

(Reece did not take the oxirac­etam with any choline sup­ple­ments, which is usu­ally rec­om­mend­ed.)


In the realm of un­usual sup­ple­ments to n-back­ing, we can in­clude , but the ev­i­dence is too weak to say much.

See Also

  • The au­thor’s own Brain Work­shop sta­tis­tics can be found here


Flaws in mainstream science (and psychology)

2013 dis­cus­sion of how sys­temic bi­ases in sci­ence, par­tic­u­larly med­i­cine and psy­chol­o­gy, have re­sulted in a re­search lit­er­a­ture filled with false pos­i­tives and ex­ag­ger­ated effects, called ‘the Repli­ca­tion Cri­sis’.

Long-s­tand­ing prob­lems in stan­dard sci­en­tific method­ol­ogy have ex­ploded as the “”: the dis­cov­ery that many re­sults in fields as di­verse as psy­chol­o­gy, eco­nom­ics, med­i­cine, bi­ol­o­gy, and so­ci­ol­ogy are in fact false or quan­ti­ta­tively highly in­ac­cu­rately mea­sured. I cover here a hand­ful of the is­sues and pub­li­ca­tions on this large, im­por­tant, and rapidly de­vel­op­ing topic up to about 2013, at which point the Repli­ca­tion Cri­sis be­came too large a topic to cover more than cur­so­ri­ly.

The cri­sis is caused by meth­ods & pub­lish­ing pro­ce­dures which in­ter­pret ran­dom noise as im­por­tant re­sults, far too small datasets, se­lec­tive analy­sis by an an­a­lyst try­ing to reach ex­pect­ed/de­sired re­sults, pub­li­ca­tion bi­as, poor im­ple­men­ta­tion of ex­ist­ing best-prac­tices, non­triv­ial lev­els of re­search fraud, soft­ware er­rors, philo­soph­i­cal be­liefs among re­searchers that false pos­i­tives are ac­cept­able, ne­glect of known con­found­ing like ge­net­ics, and skewed in­cen­tives (fi­nan­cial & pro­fes­sion­al) to pub­lish ‘hot’ re­sults.

Thus, any in­di­vid­ual piece of re­search typ­i­cally es­tab­lishes lit­tle. Sci­en­tific val­i­da­tion comes not from small p-val­ues, but from dis­cov­er­ing a reg­u­lar fea­ture of the world which dis­in­ter­ested third par­ties can dis­cover with straight­for­ward re­search done in­de­pen­dently on new data with new pro­ce­dures—repli­ca­tion.

Split out to .

  1. By , I mean fluid in­tel­li­gence, not crys­tal­lized in­tel­li­gence, since it’s un­likely that any generic train­ing would teach you Lati­nate vo­cab­u­lary terms or mid­dle-school geom­e­try. For those who ob­ject to the en­tire idea, please see Wikipedia or for a bal­anced overview what IQ can pre­dict and the ex­cep­tions, see Stern­berg et al’s 2001 re­view, “The Pre­dic­tive Value of IQ”.↩︎

  2. After a large amount of train­ing, a task may be­come learned and cease to stress the bot­tle­neck: eg “Vir­tu­ally Per­fect Time Shar­ing in Du­al-task Per­for­mance: Un­cork­ing the Cen­tral Cog­ni­tive Bot­tle­neck”.↩︎

  3. See for ex­am­ple “Do work­ing mem­ory and sus­cep­ti­bil­ity to in­ter­fer­ence pre­dict in­di­vid­ual differ­ences in fluid in­tel­li­gence?”, Borella 2006; WM pre­dicts IQ bet­ter than strong fo­cus/at­ten­tion, with the cor­re­la­tion com­ing mostly from fo­cus with only a small load­ing on ex­ec­u­tive con­trol (Chud­er­ski & Necka 2012).↩︎

  4. “Brain net­works for work­ing mem­ory and fac­tors of in­tel­li­gence as­sessed in males and fe­males with fMRI and DTI, Tang 2010; it found that “in­di­vid­ual differ­ences in ac­ti­va­tion dur­ing the n-back task were cor­re­lated to the gen­eral in­tel­li­gence fac­tor (g), as well as to dis­tilled es­ti­mates (re­mov­ing g) of speed of rea­son­ing, nu­mer­i­cal abil­i­ty, and spa­tial abil­i­ty, but not to mem­ory”. PDF avail­able in Group Files.

    A more re­cent re­sult is the fMRI study Chein 2011, “Do­main-gen­eral mech­a­nisms of com­plex work­ing mem­ory span”, which ab­stract says “For both ver­bal and spa­tial ver­sions of the task, com­plex work­ing mem­ory span per­for­mance in­creased the ac­tiv­ity in lat­eral pre­frontal, an­te­rior cin­gu­late, and pari­etal cor­tices dur­ing the En­cod­ing, Main­te­nance, and Co­or­di­na­tion phase of task per­for­mance. Mean­while, over­lap­ping ac­tiv­ity in an­te­rior pre­frontal and me­dial tem­po­ral lobe re­gions was as­so­ci­ated with both ver­bal and spa­tial re­call from work­ing mem­o­ry.”↩︎

  5. eg. “Rea­son­ing=­work­ing mem­ory ≠ at­ten­tion”, Buehner & Krummb & Pick 2005; more back­ground is avail­able on pg 10/92 of “Work­ing mem­o­ry, fluid in­tel­li­gence, and sci­ence learn­ing”. But see the meta-analy­ses in Ack­er­man et al 2005 which find that WM ≠ IQ.↩︎

  6. from Jaeggi et al 2010:

    The find­ings of Study 1 con­firm other find­ings from the lit­er­a­ture (Jaeg­gi, Buschkuehl, Per­rig, & Meier, 2010; Kane, Con­way, Mi­u­ra, & Colflesh, 2007): Con­sis­tent with our hy­pothe­ses, both n-back task vari­ants were highly cor­re­lat­ed, and both were best pre­dicted by Gf.

    In gen­er­al, ma­trix rea­son­ing tasks seem to be bet­ter pre­dic­tors for both the sin­gle and the dual n-back tasks than a mea­sure of work­ing mem­ory ca­pac­i­ty. As the re­li­a­bil­ity es­ti­mates were ap­pro­pri­ate for the n-back tasks, the lack of cor­re­la­tion be­tween the n-back tasks and the mea­sure of work­ing mem­ory ca­pac­ity can­not be at­trib­uted to in­suffi­cient re­li­a­bil­ity (Jaeg­gi, Buschkuehl, Per­rig, & Meier, 2010). Rather, it seems that per­for­mance for the two tasks re­lies on differ­ent sources of vari­ance, which might re­sult from the differ­ent mem­ory processes that are in­volved in the two tasks: whereas the n-back task re­lies on pas­sive recog­ni­tion process­es, per­for­mance in work­ing mem­ory ca­pac­ity tasks re­quires ac­tive and strate­gic re­call processes (Kane, Con­way, Mi­u­ra, & Colflesh, 2007).

  7. “Work­ing mem­ory ca­pac­ity and fluid abil­i­ties: Ex­am­in­ing the cor­re­la­tion be­tween Op­er­a­tion Span and Raven”, Unsworth, In­tel­li­gence 2005:

    How­ev­er, as shown in Fig. 2, the cor­re­la­tions be­tween so­lu­tion ac­cu­racy for each item and Ospan, al­though fluc­tu­at­ing wide­ly, does not ap­pear to in­crease in any sys­tem­atic man­ner as diffi­culty in­creas­es. In­deed, the cor­re­la­tion be­tween Ospan and ac­cu­racy on the first prob­lem was as high as with prob­lem 24 (i.e., prob­lem 1 r=0.26, prob­lem 24 r=0.26). These re­sults are strik­ingly sim­i­lar to those of Salt­house (1993) who showed roughly the same pat­tern of cor­re­la­tions be­tween so­lu­tion ac­cu­racy and a WM com­pos­ite. Both sets of re­sults sug­gest that there is not a clear re­la­tion­ship be­tween item vari­a­tions in diffi­culty on Raven and mea­sures of WM.

    …Although there seems to be ad­e­quate vari­abil­ity for quar­tile 4, this low cor­re­la­tion is prob­a­bly due to the fact that not as many sub­jects at­tempted these prob­lems. In­deed, 80% of par­tic­i­pants at­tempted the first 27 prob­lems, but only 47% of par­tic­i­pants fin­ished the test. Thus, only quar­tiles 1-3 should be in­ter­pret­ed. With this in mind, the re­sults demon­strate that the cor­re­la­tion be­tween so­lu­tion ac­cu­racy and Ospan does not in­crease as diffi­culty in­creases but in­stead re­mains fairly con­stant across in­creas­ing lev­els of diffi­cul­ty.

    …One re­viewer was con­cerned that only high work­ing mem­ory ca­pac­ity in­di­vid­u­als would fin­ish the test. How­ev­er, of those par­tic­i­pants clas­si­fied as high work­ing mem­ory (one stan­dard de­vi­a­tion above the mean on Ospan), only 25% of them ac­tu­ally fin­ished the test, whereas 71% of those clas­si­fied as low work­ing mem­ory (one stan­dard de­vi­a­tion be­low the mean on Ospan) fin­ished the test. This re­sults in some­what lower scores for these 76 in­di­vid­u­als on the two mea­sures as com­pared the full sam­ple (i.e. M Ospan=11.12, S.D.=5.90; M Raven=17.50, S.D.=7.59).

  8. “Does work­ing mem­ory train­ing gen­er­al­ize?”, Ship­stead et al 2010; ab­stract:

    Re­cent­ly, at­tempts have been made to al­ter the ca­pac­ity of work­ing mem­ory (WMC) through ex­ten­sive prac­tice on adap­tive work­ing mem­ory tasks that ad­just diffi­culty in re­sponse to user per­for­mance. We dis­cuss the de­sign cri­te­ria re­quired to claim va­lid­ity as well as gen­er­al­iz­abil­ity and how re­cent stud­ies do or do not sat­isfy those cri­te­ria. It is con­cluded that, as of yet, the re­sults are in­con­sis­tent and this is likely dri­ven by in­ad­e­quate con­trols and in­effec­tive mea­sure­ment of the cog­ni­tive abil­i­ties of in­ter­est.

  9. See Min­ear & Shah 2008:

    Per­for­mance on task switch­ing, a par­a­digm com­monly used to mea­sure ex­ec­u­tive func­tion, has been shown to im­prove with prac­tice. How­ev­er, no study has tested whether these ben­e­fits are spe­cific to the tasks learned or are trans­fer­able to new sit­u­a­tions. We re­port ev­i­dence of trans­fer­able im­prove­ment in a cued, ran­domly switch­ing par­a­digm as mea­sured by mix­ing cost, but we re­port no con­sis­tent im­prove­ment for switch cost. Im­prove­ment in mix­ing costs arises from a rel­a­tive re­duc­tion in time to per­form both switch and non­switch tri­als that im­me­di­ately fol­low switch tri­als, im­pli­cat­ing the abil­ity to re­cover from un­ex­pected switches as the source of im­prove­ment. These re­sults add to a grow­ing num­ber of stud­ies demon­strat­ing gen­er­al­iz­able im­prove­ment with train­ing on ex­ec­u­tive pro­cess­ing.

  10. “Guest Column: Can We In­crease Our In­tel­li­gence?”; Sam Wang & San­dra Aamodt; The New York Times

    Differ­ences in work­ing mem­ory ca­pac­ity ac­count for 50-70% of in­di­vid­ual differ­ences in fluid in­tel­li­gence (ab­stract rea­son­ing abil­i­ty) in var­i­ous meta-analy­ses, sug­gest­ing that it is one of the ma­jor build­ing blocks of I.Q. (Ack­er­man et al; Kane et al; Süss et al 2002) This idea is in­trigu­ing be­cause work­ing mem­ory can be im­proved by train­ing.

    See also the 2012 NYT fol­lowup, “Can You Make Your­self Smarter?”↩︎

  11. Is this right? I have no idea. But it is a cu­ri­ous col­lec­tion of stud­ies and an in­ter­est­ing pro­posed mod­el: Hat­ton 1997:

    For years I sub­scribed to such a prin­ci­ple: that mod­u­lar­iza­tion, or struc­tural de­com­po­si­tion, is a good de­sign con­cept and there­fore al­ways im­proves sys­tems. This be­lief is so wide­spread as to be al­most un­chal­lenge­able. It is re­spon­si­ble for the im­por­tant pro­gram­ming lan­guage con­cept of com­pi­la­tion mod­el­s-which are ei­ther sep­a­rate, with guar­an­teed in­ter­face con­sis­tency (such as C++, Ada, and Mod­u­la-2), or in­de­pen­dent, whereby a sys­tem is built in pieces and glued to­gether later (C and For­tran, for ex­am­ple). It is a very at­trac­tive con­cept with strong roots in the “di­vide and con­quer” prin­ci­ple of tra­di­tional en­gi­neer­ing. How­ev­er, this con­ven­tional wis­dom may be wrong. Only those com­po­nents that fit best into hu­man short­-term mem­ory cache seem to use it effec­tive­ly, thereby pro­duc­ing the low­est fault den­si­ties. Big­ger and smaller av­er­age com­po­nent sizes ap­pear to de­grade re­li­a­bil­i­ty.

    …It is easy to get the im­pres­sion from these case his­to­ries that de­vel­op­ing soft­ware sys­tems with low fault den­si­ties is ex­ceed­ingly diffi­cult. In fact, analy­sis of the lit­er­a­ture re­veals graphs such as that de­pen­dent faults per KLOC will ap­proach an as­ymp­tote as time in­creas­es. In re­al­i­ty, only this as­ymp­tote makes sense for com­par­ing the re­li­a­bil­ity of differ­ent sys­tems. So, given that the as­ymp­tote can never be reached, the faults per KLOC and the rate of change of this value are re­quired to com­pare such sys­tems effec­tive­ly. Of course, real sys­tems are sub­ject to con­tin­ual non­cor­rec­tive change, so things be­come rather more com­plex. No no­tion of rate of change of faults per KLOC was avail­able for any of the data in this study, al­though both ma­ture and im­ma­ture sys­tems were pre­sent, with the same be­hav­ior ob­served. This would sug­gest that the ob­served de­fect be­hav­ior is present through the life cy­cle, sup­port­ing even fur­ther the con­jec­ture that it is a macro­scopic prop­er­ty. If only im­ma­ture sys­tems had been present in the stud­ies, it could have been ar­gued that smaller com­po­nents may get ex­er­cised more. This does not seem to be the case.

    A fur­ther re­lated point, also ob­served in the NAG li­brary study, is that when com­po­nent fault den­si­ties are plot­ted as a func­tion of size, the us­age of each com­po­nent must be taken into ac­count. The mod­els dis­cussed in this ar­ti­cle are es­sen­tially as­ymp­tot­ic, and the fault den­si­ties they pre­dict are there­fore an en­ve­lope to which com­po­nent fault den­si­ties will tend only as they are used suffi­ciently to be­gin to flush out faults. An un­used com­po­nent has com­plex­ity but no faults, by de­fi­n­i­tion. The lit­er­a­ture re­ports ap­par­ently near-ze­ro-de­fect sys­tems that have turned out on closer in­spec­tion to have been un­used. shown in Fig­ure 2. This data was com­piled from NASA God­dard data by the Uni­ver­sity of Mary­land’s Soft­ware En­gi­neer­ing Lab­o­ra­to­ry, as quoted in the De­cem­ber 1991 spe­cial edi­tion of Busi­ness Week. First of all, in spite of NASA’s enor­mous re­sources and tal­ent pool, the av­er­age was still five to six faults per KLOC. Other stud­ies have re­ported sim­i­lar fault den­si­ties.4,8 More telling is the ob­ser­va­tion that in Fig­ure 2, im­prove­ment has been achieved mostly by im­prov­ing the bad process­es, not the good ones. This fact sug­gests that con­sis­ten­cy, a process is­sue, has im­proved much more than ac­tual fault den­si­ty, a prod­uct is­sue. The sim­ple con­clu­sion is that the av­er­age across many lan­guages and de­vel­op­ment efforts for “good” soft­ware is around six faults per KLOC, and that with our best tech­niques, we can achieve 0.5-1 fault per KLOC. Per­fec­tion will al­ways elude us, of course, but the in­tractabil­ity of achiev­ing sys­tem­at­i­cally bet­ter fault den­si­ties than have been achieved so far also sug­gests that some other lim­i­ta­tion may be at work.

    THE PROPOSED MODEL …Re­cov­ery code scram­bling is an im­por­tant fac­tor in my pro­posed mod­el. The ev­i­dence sug­gests that any­thing that fits in a short­-term or cache mem­ory is eas­ier to un­der­stand and less fault­-prone; pieces that are too large over­flow, in­volv­ing use of the more er­ror-prone re­cov­ery code mech­a­nism used for long-term stor­age. Thus, if a pro­gram­mer is work­ing with a com­po­nent of com­plex­ity Ω, and that com­po­nent fits en­tirely into the cache or short­-term mem­o­ry, which in turn can be ma­nip­u­lated with­out re­course to back­-up or long-term mem­o­ry, the in­cre­men­tal in­crease in bugs or dis­or­der dE due to an in­cre­men­tal in­crease of com­plex­ity of dΩ is sim­ply dE = (1/Ω) dΩ.

    This re­sem­bles the ar­gu­ment lead­ing to Boltz­man­n’s law re­lat­ing en­tropy to com­plex­i­ty, where the ana­logue of equipar­ti­tion of en­ergy in a phys­i­cal sys­tem is mir­rored by the ap­par­ently equal dis­tri­b­u­tion of re­hearsal ac­tiv­ity in the short­-term mem­o­ry. In other words, be­cause no part of the cache is fa­vored and the cache ac­cu­rately ma­nip­u­lates sym­bols, the in­cre­men­tal in­crease in dis­or­der is in­versely pro­por­tional to the ex­ist­ing com­plex­i­ty, mak­ing the ideal case when pieces just fit into cache. It is as­sumed with­out loss of gen­er­al­ity that both E and Ω are con­tin­u­ously val­ued vari­ables. What hap­pens when we en­counter com­plex­ity greater than Ω′ (the com­plex­ity which will just fit into the cache)? The in­crease in dis­or­der will cor­re­spond to the com­plex­ity in the (now-full) cache con­tents, plus a con­tri­bu­tion pro­por­tional to the num­ber of times the cache mem­ory must be re­loaded from the long-term mem­o­ry. In other words, dE = (1/2Ω)’ (1 + Ω/Ω’) * dΩ

    The fac­tor of 1/2 matches Equa­tion 1 when Ω = Ω′, that is, when the com­plex­ity of the pro­gram is about to over­flow the cache mem­o­ry. The sec­ond term is di­rectly pro­por­tional to the cache over­flow effect and mim­ics the scram­bling of the re­cov­ery codes. In­te­grat­ing Equa­tions 1 and 2 sug­gests that E = log Ω for Ω ≤ Ω′ and E = 1/2 * (Ω/Ω’ + Ω^2/2*Ω’^2) for Ω > Ω’

    …The Ada data and the as­sem­bly and macro-assem­bly data pro­vide strong em­pir­i­cal sup­port for this be­hav­ior, with about 200 to 400 lines cor­re­spond­ing to the com­plex­ity Ω′ at which cache mem­ory over­flows into long-term mem­o­ry. That such dis­parate lan­guages can pro­duce ap­prox­i­mately the same tran­si­tion point from log­a­rith­mic to qua­dratic be­hav­ior sup­ports the view that Ω is not the un­der­ly­ing al­go­rith­mic com­plex­ity but the sym­bolic com­plex­ity of the lan­guage im­ple­men­ta­tion, given that a line of Ada would be ex­pected to gen­er­ate five or more lines of as­sem­bly. This is di­rectly anal­o­gous to the ob­ser­va­tion that it is fit, rather than the ac­tual in­for­ma­tion con­tent of the cache that is rel­e­van­t.9

    …To sum­ma­rize, if a sys­tem is de­com­posed into pieces much smaller than the short­-term mem­ory cache, the cache is used in­effi­ciently be­cause the in­ter­face of such a com­po­nent with its neigh­bors is not “re­hearsed” ex­plic­itly into the cache in the same way, and the re­sult­ing com­po­nents tend to ex­hibit higher de­fect den­si­ties. If com­po­nents ex­ceed the cache size, they are less com­pre­hen­si­ble be­cause the re­cov­ery codes con­nect­ing com­pre­hen­sion with long-term mem­ory break down. Only those com­po­nents that match the cache size well use it effec­tive­ly, thereby pro­duc­ing the low­est fault den­si­ties.

    …Sup­pose that a par­tic­u­lar func­tion­al­ity re­quires 1,000 “lines” to im­ple­ment, where a “line” is some mea­sure of com­plex­i­ty. The im­me­di­ate im­pli­ca­tion of the ear­lier dis­cus­sion is that, to be re­li­able, we should im­ple­ment it as five 200-line com­po­nents (each fit­ting in cache) rather than as 50 20-line com­po­nents. The for­mer would lead to per­haps 5 log_10(200) = 25 bugs while the lat­ter would lead to 50 × log_10(20) = 150 bugs. This ap­par­ently in­escapable but un­pleas­ant con­clu­sion runs com­pletely counter to con­ven­tional wis­dom. …The ad­di­tional un­re­li­a­bil­ity caused by split­ting up the sys­tem might be due to sim­ple in­ter­face in­con­sis­ten­cies. The Basil­i-Per­ri­cone study con­sid­ered this a pos­si­ble ex­pla­na­tion, as did Moller-Paul­ish. How­ev­er, it was not a fac­tor in the Hat­ton-Hop­kins study, since the in­ter­nally reusable com­po­nents in the NAG li­brary (largely ex­ter­nally used reusable com­po­nents) had high in­ter­face con­sis­ten­cy. Fur­ther­more, it is un­likely to ex­plain the Comp­ton-With­row data be­cause Ada man­dates in­ter­face con­sis­tency in lan­guage im­ple­men­ta­tions. (This may be re­spon­si­ble for the differ­ence in small com­po­nents in Fig­ure 4.)

  12. Jeff At­wood; “Mul­ti­ple Mon­i­tors and Pro­duc­tiv­ity”, “The Pro­gram­mer’s Bill of Rights”, “Join­ing the Pres­ti­gious Three Mon­i­tor Club”, “Does More Than One Mon­i­tor Im­prove Pro­duc­tiv­i­ty?” etc.↩︎

  13. See ci­ta­tion roundup at the Skep­tics Stack­Ex­change.↩︎

  14. Jeff At­wood, “The Large Dis­play Para­dox”↩︎

  15. See his blog posts, pri­mar­ily “Pro­gram­ming’s Dirt­i­est Lit­tle Se­cret”. One dis­sent­ing view­point is John D. Cook’s “How much does typ­ing speed mat­ter?”, which takes an per­spec­tive—s­ince typ­ing speeds don’t vary by more than an or­der of mag­ni­tude or two or take up much time for the most part, you can’t ex­pect the over­all pro­duc­tiv­ity boost of faster typ­ing to be too big (though it could still be well worth your while). Nev­er­the­less, I think the most com­pelling ar­gu­ment for learn­ing typ­ing well or us­ing a good in­put de­vice in gen­eral is to sim­ply spend some time us­ing a key­board with a bro­ken key or a worn-out mouse/­track­ball which once in a while misclicks: even an ob­jec­tively ‘small’ er­ror rate is enough to drive one batty and de­stroy ‘flow’. I’m re­minded of Dan Lu­u’s ar­ti­cles on : specifi­cal­ly, & & .↩︎

  16. Page 457, Coders at Work:

    Seibel: “Is there any­thing you would have done differ­ently about learn­ing to pro­gram? Do you have any re­grets about the sort of path you took or do you wish you had done any­thing ear­lier?”

    : “Oh, sure, sure. In high school I wish I’d taken typ­ing. I suffer from poor typ­ing yet to­day, but who knew. I did­n’t plan any­thing or do any­thing. I have no dis­ci­pline. I did what I wanted to do next, pe­ri­od, all the time. If I had some fore­sight or plan­ning or some­thing, there are things, like typ­ing, I would have done when I had the chance.”

  17. When I was younger, I rea­soned that early in life is the best time to learn to read fast since one reaps the great­est gains over the longest pos­si­ble pe­riod (I still agree with my for­mer rea­son­ing) and so did a great deal of read­ing on and the re­lated aca­d­e­mic lit­er­a­ture, and spent more than a few hours work­ing with tachis­to­cop­ic-style soft­ware. My ul­ti­mate con­clu­sion was that it was a good use of my time as it bumped my WPM up to ~400-500 WPM from the or­di­nary 300 WPM, but the tech­niques were not go­ing to give any use­ful abil­ity be­yond that as greater speed be­comes an in­di­ca­tion one is read­ing too easy ma­te­r­ial or one should be us­ing more so­phis­ti­cated search ca­pa­bil­i­ties. In par­tic­u­lar, weren’t very use­ful for non-prac­tice read­ing and were least use­ful on deep or heav­i­ly-hy­per­linked con­tent. “Pho­tore­ad­ing”, how­ev­er, is sim­ply a scam or very shal­low skim­ming. Un­for­tu­nate­ly, I omit­ted to take notes on spe­cific stud­ies or pro­grams, though, be­ing too young to care about be­ing able to ex­plain & de­fend my be­liefs later - but that is just as well since by now, all the web­sites would be gone, pro­grams bi­trot­ten, and links bro­ken. Read­ers will just have to do their own re­search on the topic if they care (much eas­ier in this age of Wikipedi­a). One start­ing point: Scott Young.↩︎

  18. From the in­ter­view an­thol­ogy (2009), pg 114:

    Pe­ter Seibel: “Do you think that pro­gram­ming is at all bi­ased to­ward be­ing young?”

    : "I used to think so. A few years ago I had , but I did­n’t know it. I thought I was just get­ting tired and old, and I got to the point where it was so diffi­cult to con­cen­trate that I could­n’t pro­gram any­more be­cause I just could­n’t keep enough stuff in my head. A lot of pro­gram­ming is you keep stuff in your head un­til you can get it writ­ten down and struc­tured prop­er­ly. And I just could­n’t do it.

    I had lost that abil­ity and I thought it was just be­cause I was get­ting old­er. For­tu­nate­ly, I got bet­ter and it came back and so I’m pro­gram­ming again. I’m do­ing it well and maybe a lit­tle bit bet­ter now be­cause I’ve learned how not to de­pend so much on my mem­o­ry. I’m bet­ter at doc­u­ment­ing my code now than I used to be be­cause I’m less con­fi­dent that I’ll re­mem­ber next week why I did this. In fact, some­times I’ll be go­ing through my stuff and I’m amazed at stuff that I had writ­ten: I don’t re­mem­ber hav­ing done it and it’s ei­ther re­ally ei­ther aw­ful or bril­liant. I had no idea I was ca­pa­ble of that."

    From pg 154:

    Seibel: “How do you de­sign code?”

    : "A lot of pro­to­typ­ing. I used to do sort of high­-level pseudocode, and then I’d start fill­ing in bot­tom up. I do less of the high­-level pseudocode be­cause I can usu­ally hold it in my head and just do bot­tom-up un­til it joins.

    Often I’m work­ing with ex­ist­ing pieces of code adding some new sub­sys­tem or some­thing on the side and I can al­most do it bot­tom-up. When I get in trou­ble in the mid­dle I do still write pseudo-code and just start work­ing bot­tom up un­til I can com­plete it. I try not to let that take too long be­cause you’ve got to be able to test it; you’ve got to be able to see it run and step through it and make sure it’s do­ing what it’s sup­posed to be do­ing."

    From pg 202, a co­gent re­minder that ’tis a good wind that blows no ill (and that as William T. Pow­ers wrote some­where on the CSGNet ML, “Some peo­ple revel in com­plex­i­ty, and what’s worse, they have the brain power to deal with vast sys­tems of ar­cane equa­tions. This abil­ity can be a hand­i­cap be­cause it leads to over­look­ing sim­ple so­lu­tions.”):

    Seibel: “Speak­ing of writ­ing in­tri­cate code, I’ve no­ticed that peo­ple who are too smart, in a cer­tain di­men­sion any­way, make the worst code. Be­cause they can ac­tu­ally fit the whole thing in their head they can write these great reams of spaghetti code.”

    : “I agree with you that peo­ple who are both smart enough to cope with enor­mous com­plex­ity and lack em­pa­thy with the rest of us may fall prey to that. They think, ‘I can un­der­stand this and I can use it, so it has to be good.’”

    From pg 236:

    : “I read some­where, that you have to have a good mem­ory to be a rea­son­able pro­gram­mer. I be­lieve that to be true.”

    Seibel: “Bill Gates once claimed that he could still go to a black­board and write out big chunks of the code to the BASIC that he writ­ten for the Al­tair, a decade or so after he had orig­i­nally writ­ten it. Do you think you can re­mem­ber your old code that way?”

    Arm­strong: “Yeah. Well, I could re­con­struct some­thing. Some­times I’ve just com­pletely lost some old code and it does­n’t worry me in the slight­est.”

    From page 246:

    : “Yeah, that’s right. So es­sen­tially we wrote out our types by draw­ing them on large sheets of pa­pers with ar­rows. That was our type sys­tem. That was a pretty large pro­gram-in fact it was over am­bi­tious; we never com­pleted it.”

    Seibel: “Do you think you learned any lessons from that fail­ure?”

    Pey­ton Jones: “That was prob­a­bly when I first be­came aware that writ­ing a re­ally big pro­gram you could end up with prob­lems of scale-you could­n’t keep enough of it in your head at the same time. Pre­vi­ously all the things I had writ­ten, you could keep the whole thing in your head with­out any trou­ble. So it was prob­a­bly the first time I’d done any se­ri­ous at­tempt at long-s­tand­ing doc­u­men­ta­tion.”

    Seibel: “But even that was­n’t enough, in this case…”

    From page 440:

    [:] “The sec­ond rea­son I like Python is that-and maybe this is just the way my brain has changed over the years-I can’t keep as much stuff in my head as I used to. It’s more im­por­tant for me to have stuff in front of my face. So the fact that in Smalltalk you effec­tively can­not put more than one method on the screen at a time dri­ves me nuts. As far as I’m con­cerned the fact that I edit Python pro­grams with Emacs is an ad­van­tage be­cause I can see more than ten lines’ worth at a time.”

  19. From an in­ter­view given by to Dikran Karagueuzian, the di­rec­tor of CSLI Pub­li­ca­tions:

    I could­n’t keep up with all my teach­ing at Stan­ford though, I’m not on sab­bat­i­cal but I found that do­ing soft­ware was much, was much harder than writ­ing books and do­ing re­search pa­pers. It takes an­other level of com­mit­ment that you have to have so much in your head at the time when you’re do­ing soft­ware, that, that I had to take leave of ab­sence from Stan­ford from my, from my or­di­nary teach­ing for sev­eral quar­ters dur­ing this pe­ri­od.

  20. The best pro­gram­mers seem to suffer few dis­trac­tions and the worst had many, al­though it is hard to in­fer causal­ity from this strik­ing cor­re­la­tion. From “The Rise of the New Group­think”, Su­san Cain, The New York Times (draw­ing on the 1987 book or per­haps the re­lated ex­cerpts “Why Mea­sure Per­for­mance”):

    Pri­vacy also makes us pro­duc­tive. In a fas­ci­nat­ing study known as the Cod­ing War Games, con­sul­tants Tom De­Marco and Tim­o­thy Lis­ter com­pared the work of more than 600 com­puter pro­gram­mers at 92 com­pa­nies. They found that peo­ple from the same com­pa­nies per­formed at roughly the same level - but that there was an enor­mous per­for­mance gap be­tween or­ga­ni­za­tions. What dis­tin­guished pro­gram­mers at the top-per­form­ing com­pa­nies was­n’t greater ex­pe­ri­ence or bet­ter pay. It was how much pri­va­cy, per­sonal work­space and free­dom from in­ter­rup­tion they en­joyed. 62% of the best per­form­ers said their work­space was suffi­ciently pri­vate com­pared with only 19% of the worst per­form­ers. 76% of the worst pro­gram­mers but only 38% of the best said that they were often in­ter­rupted need­less­ly.

  21. “Who is Likely to Ac­quire Pro­gram­ming Skills?”, Shute 1991; Shute mea­sured WM for stu­dents learn­ing and of course found that higher WM cor­re­lated with faster learn­ing, but de­spite us­ing the g-loaded , un­for­tu­nately she ap­par­ently did not mea­sure against IQ di­rect­ly, so pos­si­bly it’s just IQ cor­re­lat­ing with the pro­gram­ming skill:

    Fol­low­ing in­struc­tion, an on­line bat­tery of cri­te­rion tests was ad­min­is­tered mea­sur­ing pro­gram­ming knowl­edge and skills ac­quired from the tu­tor. Re­sults showed that a large amount (68%) of the out­come vari­ance could be pre­dicted by a work­ing-mem­ory fac­tor, spe­cific word prob­lem solv­ing abil­i­ties (i.e., prob­lem iden­ti­fi­ca­tion and se­quenc­ing of el­e­ments) and some learn­ing style mea­sures (i.e., ask­ing for hints and run­ning pro­gram­s).

  22. In “Why An­gry Birds is so suc­cess­ful and pop­u­lar: a cog­ni­tive tear­down of the user ex­pe­ri­ence”, writer Charles L. Mauro sin­gles out se­lec­tive stress­ing of work­ing mem­ory as key to ’s man­age­ment of the diffi­culty of its puz­zles:

    It is a well-known fact of cog­ni­tive sci­ence that hu­man short­-term mem­ory (S­M), when com­pared to other at­trib­utes of our mem­ory sys­tems, is ex­ceed­ingly lim­it­ed….Where things get in­ter­est­ing is the point where poor user in­ter­face de­sign im­pacts the de­mand placed on SM. For ex­am­ple, a user in­ter­face de­sign so­lu­tion that re­quires the user to view in­for­ma­tion on one screen, store it in short­-term mem­o­ry, and then reen­ter that same in­for­ma­tion in a data field on an­other screen seems like a triv­ial task. Re­search shows that it is diffi­cult to do ac­cu­rate­ly, es­pe­cially if some other form of stim­u­lus flows be­tween the mem­o­riza­tion of the data from the first screen and be­fore the user en­ters the data in the sec­ond. This dis­rup­tive data flow can be in al­most any form, but as a gen­eral rule, any­thing that is en­gag­ing, such as con­ver­sa­tion, noise, mo­tion, or worst of all, a com­bi­na­tion of all three, is likely to to­tally erase SM. When you en­counter this type of data flow be­fore you com­plete trans­fer of data us­ing short­-term mem­o­ry, chances are very good that when you go back to re­trieve im­por­tant in­for­ma­tion from short­-term mem­o­ry, it is gone!

    An­gry Birds is a sur­pris­ingly smart man­ager of the play­er’s short­-term mem­o­ry.

    By sim­ple ma­nip­u­la­tion of the user in­ter­face, An­gry Birds de­sign­ers cre­ated sig­nifi­cant short­-term mem­ory loss, which in turn in­creases game play com­plex­ity but in a way that is not per­ceived by the player as neg­a­tive and adds to the ad­dic­tive na­ture of the game it­self. The sub­tle, yet pow­er­ful con­cept em­ployed in An­gry Birds is to bend short­-term mem­ory but not to ac­tu­ally break it. If you do break SM, make sure you give the user a very sim­ple, fast way to ac­cu­rately re­load. There are many ex­am­ples in the An­gry Birds game model of this prin­ci­ple in ac­tion….

    One of the main ben­e­fits of play­ing An­gry Birds on the iPad [rather than the smaller iPhone] is the abil­ity to pinch down the win­dow size so you can keep the en­tire game space (birds & pigs in hous­es) in full view all the time. Keep­ing all as­pects of the game’s in­ter­face in full view pre­vents short­-term mem­ory loss and im­proves the rate at which you ac­quire skills nec­es­sary to move up to a higher game lev­el. Side note: If you want the ul­ti­mate An­gry Birds ex­pe­ri­ence use a POGO pen on the iPad with the dis­play pinched down to view the en­tire game space. This gives you finer con­trol, bet­ter tar­get­ing and rapidly chang­ing game play. The net im­pact in cog­ni­tive terms is a vastly su­pe­rior skill ac­qui­si­tion pro­file. How­ev­er, you will also find that the game is less in­ter­est­ing to play over ex­tended pe­ri­ods. Why does this hap­pen?

  23. “Re­ex­am­in­ing the Fault Den­si­ty-Com­po­nent Size Con­nec­tion”, Les Hat­ton (ex­tended ex­cerpts):

    For years I sub­scribed to such a prin­ci­ple: that mod­u­lar­iza­tion, or struc­tural de­com­po­si­tion, is a good de­sign con­cept and there­fore al­ways im­proves sys­tems. This be­lief is so wide­spread as to be al­most un­chal­lenge­able. It is re­spon­si­ble for the im­por­tant pro­gram­ming lan­guage con­cept of com­pi­la­tion mod­el­s-which are ei­ther sep­a­rate, with guar­an­teed in­ter­face con­sis­tency (such as C++, Ada, and ), or in­de­pen­dent, whereby a sys­tem is built in pieces and glued to­gether later (C and For­tran, for ex­am­ple). It is a very at­trac­tive con­cept with strong roots in the “di­vide and con­quer” prin­ci­ple of tra­di­tional en­gi­neer­ing. How­ev­er, this con­ven­tional wis­dom may be wrong. Only those com­po­nents that fit best into hu­man short­-term mem­ory cache seem to use it effec­tive­ly, thereby pro­duc­ing the low­est fault den­si­ties. Big­ger and smaller av­er­age com­po­nent sizes ap­pear to de­grade re­li­a­bil­i­ty.

    …The Ada data and the as­sem­bly and macro-assem­bly data pro­vide strong em­pir­i­cal sup­port for this be­hav­ior, with about 200 to 400 lines cor­re­spond­ing to the com­plex­ity Ω′ at which cache mem­ory over­flows into long-term mem­o­ry. That such dis­parate lan­guages can pro­duce ap­prox­i­mately the same tran­si­tion point from log­a­rith­mic to qua­dratic be­hav­ior sup­ports the view that Ω is not the un­der­ly­ing al­go­rith­mic com­plex­ity but the sym­bolic com­plex­ity of the lan­guage im­ple­men­ta­tion, given that a line of Ada would be ex­pected to gen­er­ate five or more lines of as­sem­bly. This is di­rectly anal­o­gous to the ob­ser­va­tion that it is fit, rather than the ac­tual in­for­ma­tion con­tent of the cache that is rel­e­vant.9

    …To sum­ma­rize, if a sys­tem is de­com­posed into pieces much smaller than the short­-term mem­ory cache, the cache is used in­effi­ciently be­cause the in­ter­face of such a com­po­nent with its neigh­bors is not “re­hearsed” ex­plic­itly into the cache in the same way, and the re­sult­ing com­po­nents tend to ex­hibit higher de­fect den­si­ties. If com­po­nents ex­ceed the cache size, they are less com­pre­hen­si­ble be­cause the re­cov­ery codes con­nect­ing com­pre­hen­sion with long-term mem­ory break down. Only those com­po­nents that match the cache size well use it effec­tive­ly, thereby pro­duc­ing the low­est fault den­si­ties.

    …Sup­pose that a par­tic­u­lar func­tion­al­ity re­quires 1,000 “lines” to im­ple­ment, where a “line” is some mea­sure of com­plex­i­ty. The im­me­di­ate im­pli­ca­tion of the ear­lier dis­cus­sion is that, to be re­li­able, we should im­ple­ment it as five 200-line com­po­nents (each fit­ting in cache) rather than as 50 20-line com­po­nents. The for­mer would lead to per­haps bugs while the lat­ter would lead to bugs. This ap­par­ently in­escapable but un­pleas­ant con­clu­sion runs com­pletely counter to con­ven­tional wis­dom. …The ad­di­tional un­re­li­a­bil­ity caused by split­ting up the sys­tem might be due to sim­ple in­ter­face in­con­sis­ten­cies. The Basil­i-Per­ri­cone study con­sid­ered this a pos­si­ble ex­pla­na­tion, as did Moller-Paul­ish. How­ev­er, it was not a fac­tor in the Hat­ton-Hop­kins study, since the in­ter­nally reusable com­po­nents in the NAG li­brary (largely ex­ter­nally used reusable com­po­nents) had high in­ter­face con­sis­ten­cy. Fur­ther­more, it is un­likely to ex­plain the Comp­ton-With­row data be­cause Ada man­dates in­ter­face con­sis­tency in lan­guage im­ple­men­ta­tions. (This may be re­spon­si­ble for the differ­ence in small com­po­nents in Fig­ure 4.)

  24. “Walk­ing is free, but Amer­i­cans spent $13 mil­lion on brain-fit­ness soft­ware and games last year [2009]…”; from Newsweek↩︎

  25. See for ex­am­ple Na­ture’s cov­er­age of the Cam­bridge study, “No gain from brain train­ing: Com­put­er­ized men­tal work­outs don’t boost men­tal skills, study claims”; or Dis­cover’s blog dis­cus­sion.↩︎

  26. Newsweek:

    Train­ing your mem­o­ry, rea­son­ing, or speed of pro­cess­ing im­proves that skill, found a large gov­ern­men­t-spon­sored study called Ac­tive. Un­for­tu­nate­ly, there is no trans­fer: im­prov­ing pro­cess­ing speed does not im­prove mem­o­ry, and im­prov­ing mem­ory does not im­prove rea­son­ing. Sim­i­lar­ly, do­ing cross­word puz­zles will im­prove your abil­ity to…do cross­words. “The re­search so far sug­gests that cog­ni­tive train­ing ben­e­fits only the task used in train­ing and does not gen­er­al­ize to other tasks,” says Columbi­a’s Stern.

  27. “This Is Your Brain. Ag­ing. Sci­ence is re­shap­ing what we know about get­ting old­er. (The news is bet­ter than you think.)”, Newsweek:

    Do­ing cross­word puz­zles would seem to be ideal brain ex­er­cise since avid puz­zlers do them daily and say it keeps them men­tally sharp, es­pe­cially with vo­cab­u­lary and mem­o­ry. But this may be con­fus­ing cause and effect. It is mostly peo­ple who are good at fig­ur­ing out “Dole’s run­ning mate” who do cross­words reg­u­lar­ly; those who aren’t, don’t. In a re­cent study, Salt-house and col­leagues found “no ev­i­dence” that peo­ple who do cross­words have “a slower rate of age-re­lated de­cline in rea­son­ing.” As he put it in a 2006 analy­sis, there is “lit­tle sci­en­tific ev­i­dence that en­gage­ment in men­tally stim­u­lat­ing ac­tiv­i­ties al­ters the rate of men­tal ag­ing,” an idea that is “more of an op­ti­mistic hope than an em­pir­i­cal re­al­i­ty.” (P.S.: Bob Dole’s 1996 VP choice was Jack Kem­p.)

  28. Mu­sic cor­re­lates with in­creased SAT scores, which has been cited as a jus­ti­fi­ca­tion for teach­ing stu­dents mu­sic, but it ex­hibit a com­mon pat­tern for claims of far trans­fer: it ap­pears in sim­ple analy­ses, dis­ap­pears in ran­dom­ized ex­per­i­ments (eg ), and fi­nally a thor­ough analy­sis in­clud­ing a wide range of co­vari­ates like El­pus 2013 finds the cor­re­la­tion dis­ap­pears be­cause it was due to some con­found like the high­er-per­form­ing stu­dents also be­ing wealth­i­er. The back­ground for mu­sic:

    An en­tire spe­cial is­sue of the Jour­nal of Aes­thetic Ed­u­ca­tion (JAE) in 2000, ti­tled “The Arts and Aca­d­e­mic Achieve­ment: What the Ev­i­dence Shows”, was ded­i­cated to ex­am­in­ing the aca­d­e­mic per­for­mance of arts and non-arts stu­dents. In that vol­ume, Win­ner and Cooper (2000) meta-an­a­lyzed some 31 pub­lished and un­pub­lished stud­ies, yield­ing 66 sep­a­rate effect sizes ex­am­in­ing the gen­eral re­search ques­tion of whether arts ed­u­ca­tion, broadly de­fined, pos­i­tively in­flu­enced aca­d­e­mic achieve­ment. Re­sults of the meta-analy­sis showed that arts ed­u­ca­tion was mod­er­ately pos­i­tively as­so­ci­ated with higher achieve­ment in math, ver­bal, and com­pos­ite math­-ver­bal out­comes. In the same jour­nal is­sue, Vaughan and Win­ner (2000) sought to an­a­lyze the link be­tween arts course work and SAT scores specifi­cal­ly. Us­ing data from 12 years of na­tional SAT means re­ported by the Col­lege Board in the an­nual Pro­files of Col­lege Bound Se­niors re­port, Vaughan and Win­ner found that stu­dents who self­-re­ported on the SAT’s Stu­dent De­scrip­tive Ques­tion­naire that they had pur­sued arts course work outscored stu­dents who re­ported they had not taken any arts course work. Meta-analy­ses of mu­sic stu­dents’ per­for­mance on ver­bal (But­zlaff, 2000) and math­e­mat­i­cal (Vaugh­an, 2000) stan­dard­ized tests were some­what in­con­clu­sive: Al­though pos­i­tive as­so­ci­a­tions were found in the cor­re­la­tional re­search lit­er­a­ture, meta-analy­ses of re­sults from the few ex­per­i­men­tal stud­ies lo­cated in the lit­er­a­ture showed lit­tle to no in­flu­ence of mu­sic on ver­bal or math test scores…In British Columbia, Canada (Gouzoua­sis, Guhn, & Kishor, 2007), re­sults of an ob­ser­va­tional study in­di­cated an as­so­ci­a­tion be­tween mu­sic en­roll­ment and higher sub­jec­t-area stan­dard­ized test scores among high school stu­dents. The re­sults of a ran­dom­ized ex­per­i­ment in Mon­tre­al, Canada, showed no effects of pi­ano in­struc­tion on sub­jec­t-area stan­dard­ized tests among el­e­men­tary school chil­dren from low so­cioe­co­nomic back­grounds (Costa-Giomi, 2004).

  29. A spe­cific ex­am­ple: —chess-play­ing had su­pe­rior chess­board re­call than adults, but adults still had bet­ter re­call of num­bers. Ex­actly as ex­pected from train­ing with no trans­fer.↩︎

  30. Na­tional mem­ory cham­pion Ta­tiana Coo­ley: “I’m in­cred­i­bly ab­sen­t-mind­ed. I live by Post-its.” Or the Wash­ing­ton Post, re­view­ing Joshua Fo­er’s 2011 :

    Foer sets out to meet the leg­endary “Brain­man,” who learned Span­ish in a sin­gle week­end, could in­stantly tell if any num­ber up to 10,000 was prime, and saw dig­its in col­ors and shapes, en­abling him to hold long lists of them in mem­o­ry. The au­thor also tracks down “Rain Man” , the fa­mous sa­vant whose as­ton­ish­ing abil­ity to re­cite all of Shake­speare’s works, re­pro­duce scores from a vast canon of clas­si­cal mu­sic and re­tain the con­tents of 9,000 books was im­mor­tal­ized in the star­ring Dustin Hoff­man. When Foer is told that the Rain Man had an IQ of merely 87 - that he was ac­tu­ally miss­ing a part of his brain; that mem­ory cham­pi­ons have no more in­tel­li­gence than you or I; that build­ing a mem­ory is a mat­ter of ded­i­ca­tion and train­ing - he de­cides to try for the U.S. him­self. Here is where the book veers sharply from sci­ence jour­nal­ism to a mem­oir of a sin­gu­lar ad­ven­ture.

  31. “Get­ting a Grip on Drink­ing Be­hav­ior: Train­ing Work­ing Mem­ory to Re­duce Al­co­hol Abuse”, Houben et al 2011:

    Al­co­hol abuse dis­rupts core ex­ec­u­tive func­tions, in­clud­ing work­ing mem­ory (WM)-the abil­ity to main­tain and ma­nip­u­late goal-rel­e­vant in­for­ma­tion. When ex­ec­u­tive func­tions like WM are weak­ened, drink­ing be­hav­ior gets out of con­trol and is guided more strongly by au­to­matic im­puls­es. This study in­ves­ti­gated whether train­ing WM re­stores con­trol over drink­ing be­hav­ior. Forty-eight prob­lem drinkers per­formed WM train­ing tasks or con­trol tasks dur­ing 25 ses­sions over at least 25 days. Be­fore and after train­ing, we mea­sured WM and drink­ing be­hav­ior. Train­ing WM im­proved WM and re­duced al­co­hol in­take for more than 1 month after the train­ing. Fur­ther, the in­di­rect effect of train­ing on al­co­hol use through im­proved WM was mod­er­ated by par­tic­i­pants’ lev­els of au­to­matic im­puls­es: In­creased WM re­duced al­co­hol con­sump­tion in par­tic­i­pants with rel­a­tively strong au­to­matic pref­er­ences for al­co­hol. These find­ings are con­sis­tent with the the­o­ret­i­cal frame­work and demon­strate that train­ing WM may be an effec­tive strat­egy to re­duce al­co­hol use by in­creas­ing con­trol over au­to­matic im­pulses to drink al­co­hol.

  32. , Bickel et al 2011; WM tasks were digit span, re­verse digit span, and a list-of-word­s-match­ing task. De­creas­ing their does not ac­tu­ally show any re­duced drug abuse or bet­ter odds of re­ha­bil­i­ta­tion, but it is hope­ful.↩︎

  33. “Self­-Dis­ci­pline Out­does IQ in Pre­dict­ing Aca­d­e­mic Per­for­mance of Ado­les­cents”, Duck­worth 2006; ab­stract:

    In a lon­gi­tu­di­nal study of 140 eighth-grade stu­dents, self­-dis­ci­pline mea­sured by self­-re­port, par­ent re­port, teacher re­port, and mon­e­tary choice ques­tion­naires in the fall pre­dicted fi­nal grades, school at­ten­dance, stan­dard­ized achieve­men­t-test scores, and se­lec­tion into a com­pet­i­tive high school pro­gram the fol­low­ing spring. In a repli­ca­tion with 164 eighth graders, a be­hav­ioral de­lay-of-grat­i­fi­ca­tion task, a ques­tion­naire on study habits, and a group-ad­min­is­tered IQ test were added. Self­-dis­ci­pline mea­sured in the fall ac­counted for more than twice as much vari­ance as IQ in fi­nal grades, high school se­lec­tion, school at­ten­dance, hours spent do­ing home­work, hours spent watch­ing tele­vi­sion (in­verse­ly), and the time of day stu­dents be­gan their home­work. The effect of self­-dis­ci­pline on fi­nal grades held even when con­trol­ling for first-mark­ing-pe­riod grades, achieve­men­t-test scores, and mea­sured IQ. These find­ings sug­gest a ma­jor rea­son for stu­dents falling short of their in­tel­lec­tual po­ten­tial: their fail­ure to ex­er­cise self­-dis­ci­pline.

  34. This is prob­a­bly not sur­pris­ing, since even in adults, those with higher WMs are bet­ter at con­trol­ling their emo­tions when asked to do so; ab­stract of “Work­ing mem­ory ca­pac­ity and spon­ta­neous emo­tion reg­u­la­tion: High ca­pac­ity pre­dicts self­-en­hance­ment in re­sponse to neg­a­tive feed­back”:

    Al­though pre­vi­ous ev­i­dence sug­gests that work­ing mem­ory ca­pac­ity (WMC) is im­por­tant for suc­cess at emo­tion reg­u­la­tion, that ev­i­dence may re­veal sim­ply that peo­ple with higher WMC fol­low in­struc­tions bet­ter than those with lower WMC. The present study tested the hy­poth­e­sis that peo­ple with higher WMC more effec­tively en­gage in spon­ta­neous emo­tion reg­u­la­tion fol­low­ing neg­a­tive feed­back, rel­a­tive to those with lower WMC. Par­tic­i­pants were ran­domly as­signed to re­ceive ei­ther no feed­back or neg­a­tive feed­back about their emo­tional in­tel­li­gence. They then com­pleted a dis­guised mea­sure of self­-en­hance­ment and a self­-re­port mea­sure of affect. Ex­per­i­men­tal con­di­tion and WMC in­ter­acted such that higher WMC pre­dicted more self­-en­hance­ment and less neg­a­tive affect fol­low­ing neg­a­tive feed­back. This re­search pro­vides novel in­sight into the con­se­quences of in­di­vid­ual differ­ences in WMC and il­lus­trates that cog­ni­tive ca­pac­ity may fa­cil­i­tate the spon­ta­neous self­-reg­u­la­tion of emo­tion.

  35. “In­ves­ti­gat­ing the pre­dic­tive roles of work­ing mem­ory and IQ in aca­d­e­mic at­tain­ment”, Al­loway 2010:

    …The find­ings in­di­cate that chil­dren’s work­ing mem­ory skills at 5 years of age were the best pre­dic­tor of lit­er­acy and nu­mer­acy 6 years lat­er. IQ, in con­trast, ac­counted for a smaller por­tion of unique vari­ance to these learn­ing out­comes. The re­sults demon­strate that work­ing mem­ory is not a proxy for IQ but rather rep­re­sents a dis­so­cia­ble cog­ni­tive skill with unique links to aca­d­e­mic at­tain­ment. Crit­i­cal­ly, we find that work­ing mem­ory at the start of for­mal ed­u­ca­tion is a more pow­er­ful pre­dic­tor of sub­se­quent aca­d­e­mic suc­cess than IQ….

    Less strik­ing but still rel­e­vant is “Work­ing Mem­o­ry, but Not IQ, Pre­dicts Sub­se­quent Learn­ing in Chil­dren with Learn­ing Diffi­cul­ties”, Al­loway 2009:

    The pur­pose of the present study was to com­pare the pre­dic­tive power of work­ing mem­ory and IQ in chil­dren iden­ti­fied as hav­ing learn­ing diffi­cul­ties…Chil­dren aged be­tween 7 and 11 years were tested at Time 1 on mea­sures of work­ing mem­o­ry, IQ, and learn­ing. They were then retested 2 years later on the learn­ing mea­sures. The find­ings in­di­cated that work­ing-mem­ory ca­pac­ity and do­main-spe­cific knowl­edge at Time 1, but not IQ, were sig­nifi­cant pre­dic­tors of learn­ing at Time 2.

  36. “Com­put­er­ized Train­ing of Work­ing Mem­ory in Chil­dren With ADHD - A Ran­dom­ized, Con­trolled Trial”, Kling­berg et al 2005; ab­stract:

    …For the span-board task, there was a sig­nifi­cant treat­ment effect both post-in­ter­ven­tion and at fol­low-up. In ad­di­tion, there were sig­nifi­cant effects for sec­ondary out­come tasks mea­sur­ing ver­bal WM, re­sponse in­hi­bi­tion, and com­plex rea­son­ing. Par­ent rat­ings showed sig­nifi­cant re­duc­tion in symp­toms of inat­ten­tion and hy­per­ac­tiv­i­ty/im­pul­siv­i­ty, both post-in­ter­ven­tion and at fol­low-up. Con­clu­sion­s:This study shows that WM can be im­proved by train­ing in chil­dren with ADHD. This train­ing also im­proved re­sponse in­hi­bi­tion and rea­son­ing and re­sulted in a re­duc­tion of the par­en­t-rated inat­ten­tive symp­toms of ADHD.

    See also Green et al 2012.↩︎

  37. “Train­ing and trans­fer effects of ex­ec­u­tive func­tions in preschool chil­dren”, Thorell et al 2009↩︎

  38. “Differ­en­tial effects of rea­son­ing and speed train­ing in chil­dren” (the list of rea­son­ing games, page 5, does not seem to in­clude any di­rect ana­logues to n-back):

    The goal of this study was to de­ter­mine whether in­ten­sive train­ing can ame­lio­rate cog­ni­tive skills in chil­dren. Chil­dren aged 7 to 9 from low so­cioe­co­nomic back­grounds par­tic­i­pated in one of two cog­ni­tive train­ing pro­grams for 60 min­utes ⁄ day and 2 days ⁄ week, for a to­tal of 8 weeks. Both train­ing pro­grams con­sisted of com­mer­cially avail­able com­put­er­ized and non- com­put­er­ized games. Rea­son­ing train­ing em­pha­sized plan­ning and re­la­tional in­te­gra­tion; speed train­ing em­pha­sized rapid vi­sual de­tec­tion and rapid mo­tor re­spons­es. Stan­dard as­sess­ments of rea­son­ing abil­ity - the Test of Non-Ver­bal In­tel­li­gence (TONI-3) and cog­ni­tive speed (Cod­ing B from WISC IV) - were ad­min­is­tered to all chil­dren be­fore and after train­ing. Nei­ther group was ex­posed to these stan­dard­ized tests dur­ing train­ing. Chil­dren in the rea­son­ing group im­proved sub­stan­tially on TONI (Co­hen’s d = 1.51), ex­hibit­ing an av­er­age in­crease of 10 points in Per­for­mance IQ, but did not im­prove on Cod­ing. By con­trast, chil­dren in the speed group im­proved sub­stan­tially on Cod­ing (d = 1.15), but did not im­prove on TONI. Counter to wide­spread be­lief, these re­sults in­di­cate that both fluid rea­son­ing and pro­cess­ing speed are mod­i­fi­able by train­ing.

  39. See again Stern­berg et al’s 2001 re­view, “The Pre­dic­tive Value of IQ”:

    Ev­i­dence from stud­ies of the nat­ural course of de­vel­op­ment: Some get more in­tel­li­gent, oth­ers get less in­tel­li­gent. The Berke­ley Guid­ance Study (Honzik, Mac­far­lane, & Al­len, 1948) in­ves­ti­gated the sta­bil­ity of IQ test per­for­mance over 12 years. The au­thors re­ported that nearly 60% of the sam­ple changed by 15 IQ points or more from 6 to 18 years of age. A sim­i­lar re­sult was found in the Fels study (Son­tag, Bak­er, & Nel­son, 1958): Nearly two thirds of the chil­dren changed more than 15 IQ points from age 3 to age 10. Re­searchers also in­ves­ti­gated the so-called in­tel­li­gence la­bil­ity score, which is a child’s stan­dard de­vi­a­tion from his or her own grand mean IQ. Bay­ley (1949), in the Berke­ley Growth study, de­tected very large in­di­vid­ual differ­ences in la­bil­ity across the span of 18 years. Rees and Palmer (1970) com­bined the data from five large-s­cale lon­gi­tu­di­nal stud­ies, se­lect­ing those par­tic­i­pants who had scores at both age 6 and age 12 or at both age 12 and age 17. They found that about 30% of the se­lected par­tic­i­pants changed by 10 or more IQ points.

    Stern­berg et al also dis­cusses the dra­matic IQ gains pos­si­ble dur­ing in­fancy when adoptees mov­ing from a bad en­vi­ron­ment (Third or Sec­ond World or­phan­ages) to good ones (First World home­s), but also the dis­cour­ag­ing ex­am­ples of early in­ter­ven­tion pro­grams in the USA where ini­tial IQ gains often fade away over the years.↩︎

  40. See (me­dia cov­er­age: “IQ Is­n’t Set In Stone, Sug­gests Study That Finds Big Jumps, Dips In Teens”);

    Neu­roimag­ing al­lows us to test whether un­ex­pected lon­gi­tu­di­nal fluc­tu­a­tions in mea­sured IQ are re­lated to brain de­vel­op­ment. Here we show that ver­bal and non-ver­bal IQ can rise or fall in the teenage years, with these changes in per­for­mance val­i­dated by their close cor­re­la­tion with changes in lo­cal brain struc­ture. A com­bi­na­tion of struc­tural and func­tional imag­ing showed that ver­bal IQ changed with grey mat­ter in a re­gion that was ac­ti­vated by speech, whereas non-ver­bal IQ changed with grey mat­ter in a re­gion that was ac­ti­vated by fin­ger move­ments. By us­ing lon­gi­tu­di­nal as­sess­ments of the same in­di­vid­u­als, we ob­vi­ated the many sources of vari­a­tion in brain struc­ture that con­found cross-sec­tional stud­ies. This al­lowed us to dis­so­ci­ate neural mark­ers for the two types of IQ and to show that gen­eral ver­bal and non-ver­bal abil­i­ties are closely linked to the sen­so­ri­mo­tor skills in­volved in learn­ing.

    It’s worth not­ing that sub­stan­tial changes in the brain con­tinue to take place to­wards the end of ado­les­cence and early adult­hood, and at least some are about re­duc­ing one’s men­tal flex­i­bil­i­ty; from Na­tional Ge­o­graphic, “Beau­ti­ful Brains: Moody. Im­pul­sive. Mad­den­ing. Why do teenagers act the way they do? Viewed through the eyes of evo­lu­tion, their most ex­as­per­at­ing traits may be the key to suc­cess as adults”:

    Mean­while, in times of doubt, take in­spi­ra­tion in one last dis­tinc­tion of the teen brain-a fi­nal key to both its clum­si­ness and its re­mark­able adapt­abil­i­ty. This is the pro­longed plas­tic­ity of those late-de­vel­op­ing frontal ar­eas as they slowly ma­ture. As noted ear­lier, these ar­eas are the last to lay down the fatty in­su­la­tion-the brain’s white mat­ter-that speeds trans­mis­sion. And at first glance this seems like bad news: If we need these ar­eas for the com­plex task of en­ter­ing the world, why aren’t they run­ning at full speed when the chal­lenges are most daunt­ing?

    The an­swer is that speed comes at the price of flex­i­bil­i­ty. While a myelin coat­ing greatly ac­cel­er­ates an ax­on’s band­width, it also in­hibits the growth of new branches from the ax­on. Ac­cord­ing to Dou­glas Fields, an NIH neu­ro­sci­en­tist who has spent years study­ing myelin, “This makes the pe­riod when a brain area lays down myelin a sort of cru­cial pe­riod of learn­ing-the wiring is get­ting up­grad­ed, but once that’s done, it’s harder to change.”

    The win­dow in which ex­pe­ri­ence can best rewire those con­nec­tions is highly spe­cific to each brain area. Thus the brain’s lan­guage cen­ters ac­quire their in­su­la­tion most heav­ily in the first 13 years, when a child is learn­ing lan­guage. The com­pleted in­su­la­tion con­sol­i­dates those gain­s-but makes fur­ther gains, such as sec­ond lan­guages, far harder to come by. So it is with the fore­brain’s myeli­na­tion dur­ing the late teens and early 20s. This de­layed com­ple­tion-a with­hold­ing of readi­ness-height­ens flex­i­bil­ity just as we con­front and en­ter the world that we will face as adults.

  41. Jaeg­gi, S. M., Seew­er, R., Nirkko, A. C., Eck­stein, D., Schroth, G., Groner, R., et al, (2003). “Does ex­ces­sive mem­ory load at­ten­u­ate ac­ti­va­tion in the pre­frontal cor­tex? Load­-de­pen­dent pro­cess­ing in sin­gle and dual tasks: func­tional mag­netic res­o­nance imag­ing study”, Neu­roim­age 19(2) 210-225.↩︎

  42. 22 = 4; 4-1 = 3. For DNB, the 3 re­sponses are:

    1. au­dio match
    2. vi­sual match
    3. au­dio & vi­sual matches
  43. or 23 - 1↩︎

  44. or 24 - 1↩︎

  45. or 25 - 1↩︎

  46. Spread­ing one’s efforts over a va­ri­ety of ac­tiv­i­ties is not nec­es­sar­ily a good thing, and can be sub­-op­ti­mal; con­sider the char­ity ex­am­ple (“Giv­ing Your All”, Steven E. Lands­burg):

    Peo­ple con­stantly ig­nore my good ad­vice by con­tribut­ing to the Amer­i­can Heart As­so­ci­a­tion, the Amer­i­can Can­cer So­ci­ety, CARE, and pub­lic ra­dio all in the same year—as if they were think­ing, “OK, I think I’ve pretty much wrapped up the prob­lem of heart dis­ease; now let’s see what I can do about can­cer.”

  47. see eg. Mc­Nab or West­er­berg.↩︎

  48. I’m not the only one to no­tice this. ‘y offs et’ men­tions dur­ing a dis­cus­sion of TNB that:

    It’s in­ter­est­ing how do­ing n-back proves that time is rel­a­tive and based upon our per­cep­tion of its pass­ing

    When I’m do­ing well, the next in­stance comes with metronome ex­act­ness as ex­pected from a ma­chine. When I’m re­set­ting after a tricky dou­ble-back, the next in­stance al­ways comes way too quick­ly, as if a sec­ond had been re­moved. The same per­cep­tion hap­pens on an upped lev­el, and it is so per­sis­tent. It’s like some time had van­ished.

    For the longest time I thought the pro­gram had a bug, be­ing the mere hu­man.

  49. ↩︎

  50. Sleep affects IQ, not just vig­i­lance or en­er­gy: , John­stone et al 2010; ab­stract:

    Fluid in­tel­li­gence in­volves novel prob­lem-solv­ing and may be sus­cep­ti­ble to poor sleep. This study ex­am­ined re­la­tion­ships be­tween ado­les­cent sleep, fluid in­tel­li­gence, and aca­d­e­mic achieve­ment. Par­tic­i­pants were 217 ado­les­cents (42% male) aged 13 to 18 years (mean age, 14.9 years; SD = 1.0) in grades 9-11. Fluid in­tel­li­gence was pre­dicted to me­di­ate the re­la­tion­ship be­tween ado­les­cent sleep and aca­d­e­mic achieve­ment. Stu­dents com­pleted on­line ques­tion­naires of self­-re­ported sleep, fluid in­tel­li­gence (Let­ter Sets and Num­ber Se­ries), and self­-re­ported grades. To­tal sleep time was not sig­nifi­cantly re­lated to fluid in­tel­li­gence nor aca­d­e­mic achieve­ment (both p > 0.05); how­ev­er, sleep diffi­culty (e.g. diffi­culty ini­ti­at­ing sleep, un­re­fresh­ing sleep) was re­lated to both (P < 0.05)…

    Fur­ther, we can eas­ily de­lude our­selves about our own men­tal states:

    Still, while it’s tempt­ing to be­lieve we can train our­selves to be among the five-hour group - we can’t, Dinges says - or that we are nat­u­rally those five-hour sleep­ers, con­sider a key find­ing from Van Don­gen and Dinges’s study: after just a few days, the four- and six-hour group re­ported that, yes, they were slightly sleepy. But they in­sisted they had ad­justed to their new state. Even 14 days into the study, they said sleepi­ness was not affect­ing them. In fact, their per­for­mance had tanked. In other words, the sleep­-de­prived among us are lousy judges of our own sleep needs.

  51. NASA Naps: NASA-supported sleep re­searchers are learn­ing new and sur­pris­ing things about naps.”, 2005-06-03:

    “To our amaze­ment, work­ing mem­ory per­for­mance ben­e­fited from the naps, [but] vig­i­lance and ba­sic alert­ness did not ben­e­fit very much,” says Dinges.

  52. has been shown to im­prove men­tal fit­ness. One small study with old di­a­bet­ics found im­prove­ment in work­ing mem­o­ry/ex­ec­u­tive func­tion caused by an aer­o­bic ex­er­cise reg­i­men, and an­other found in­creased brain vol­ume and in­creased hip­pocam­pal vol­ume & BDNF se­cre­tion in healthy old peo­ple; a found ben­e­fits in 8 of 11 aer­o­bic in­ter­ven­tions in the el­der­ly. And ex­er­cise im­proves work­ing mem­ory (or at least cor­re­lated with in­tel­li­gence & ed­u­ca­tion in twin­s), and there is some sug­ges­tive ev­i­dence that or may help as well. One pos­si­ble mech­a­nism (in rats, any­way) is in­creases in chem­i­cal en­ergy stor­age in the brain. For fur­ther read­ing, see the re­view & re­views cited in and .↩︎

  53. See for ex­am­ple “Zinc sta­tus and cog­ni­tive func­tion of preg­nant women in South­ern Ethiopia” or “Zinc sup­ple­men­ta­tion im­proved cog­ni­tive per­for­mance and taste acu­ity in In­dian ado­les­cent girls”↩︎

  54. “Acute hy­po­glycemia im­pairs non­ver­bal in­tel­li­gence: im­por­tance of avoid­ing ceil­ing effects in cog­ni­tive func­tion test­ing.”. While we’re at it, blood sugar seems to be closely linked to at­ten­tion/­self-con­trol/­self-dis­ci­pline (see LW dis­cus­sions: “The Phys­i­ol­ogy of Willpower”, “Willpow­er: not a lim­ited re­source?”, “What would you do if blood glu­cose the­ory of willpower was true?”, Vladimir/ Golovin, and “Su­per­stim­uli and the Col­lapse of West­ern Civ­i­liza­tion”). For a roundup of all the re­search, read Baumeis­ter & Tier­ney’s 2011 book Willpower.↩︎

  55. Quotes from “Do You Suffer From De­ci­sion Fa­tigue?”, NYT, it­self quot­ing from Baumeis­ter & Tier­ney 2011:

    Once you’re men­tally de­plet­ed, you be­come re­luc­tant to make trade-offs, which in­volve a par­tic­u­larly ad­vanced and tax­ing form of de­ci­sion mak­ing. In the rest of the an­i­mal king­dom, there aren’t a lot of pro­tracted ne­go­ti­a­tions be­tween preda­tors and prey. To com­pro­mise is a com­plex hu­man abil­ity and there­fore one of the first to de­cline when willpower is de­plet­ed. You be­come what re­searchers call a cog­ni­tive mis­er, hoard­ing your en­er­gy. If you’re shop­ping, you’re li­able to look at only one di­men­sion, like price: just give me the cheap­est. Or you in­dulge your­self by look­ing at qual­i­ty: I want the very best (an es­pe­cially easy strat­egy if some­one else is pay­ing). De­ci­sion fa­tigue leaves you vul­ner­a­ble to mar­keters who know how to time their sales, as Jonathan Lev­av, the Stan­ford pro­fes­sor, demon­strated in ex­per­i­ments in­volv­ing tai­lored suits and new cars.

    Most of us in Amer­ica won’t spend a lot of time ag­o­niz­ing over whether we can afford to buy soap, but it can be a de­plet­ing choice in rural In­dia. Dean Spears, an econ­o­mist at Prince­ton, offered peo­ple in 20 vil­lages in Ra­jasthan in north­west­ern In­dia the chance to buy a cou­ple of bars of brand-name soap for the equiv­a­lent of less than 20 cents. It was a steep dis­count off the reg­u­lar price, yet even that sum was a strain for the peo­ple in the 10 poor­est vil­lages. Whether or not they bought the soap, the act of mak­ing the de­ci­sion left them with less willpow­er, as mea­sured after­ward in a test of how long they could squeeze a hand grip. In the slightly more afflu­ent vil­lages, peo­ple’s willpower was­n’t affected sig­nifi­cant­ly…To es­tab­lish cause and effect, re­searchers at Baumeis­ter’s lab tried re­fu­el­ing the brain in a se­ries of ex­per­i­ments in­volv­ing lemon­ade mixed ei­ther with sugar or with a diet sweet­en­er. The sug­ary lemon­ade pro­vided a burst of glu­cose, the effects of which could be ob­served right away in the lab; the sug­ar­less va­ri­ety tasted quite sim­i­lar with­out pro­vid­ing the same burst of glu­cose. Again and again, the sugar re­stored willpow­er, but the ar­ti­fi­cial sweet­ener had no effect. The glu­cose would at least mit­i­gate the ego de­ple­tion and some­times com­pletely re­verse it. The re­stored willpower im­proved peo­ple’s self­-con­trol as well as the qual­ity of their de­ci­sions: they re­sisted ir­ra­tional bias when mak­ing choic­es, and when asked to make fi­nan­cial de­ci­sions, they were more likely to choose the bet­ter long-term strat­egy in­stead of go­ing for a quick pay­off. The ego-de­ple­tion effect was even demon­strated with dogs in two stud­ies by Holly Miller and Nathan De­Wall at the Uni­ver­sity of Ken­tucky. After obey­ing sit and stay com­mands for 10 min­utes, the dogs per­formed worse on self­-con­trol tests and were also more likely to make the dan­ger­ous de­ci­sion to chal­lenge an­other dog’s turf. But a dose of glu­cose re­stored their willpow­er. The re­sults of the ex­per­i­ment were an­nounced in Jan­u­ary, dur­ing Heather­ton’s speech ac­cept­ing the lead­er­ship of the So­ci­ety for Per­son­al­ity and So­cial Psy­chol­o­gy, the world’s largest group of so­cial psy­chol­o­gists. In his pres­i­den­tial ad­dress at the an­nual meet­ing in San An­to­nio, Heather­ton re­ported that ad­min­is­ter­ing glu­cose com­pletely re­versed the brain changes wrought by de­ple­tion - a find­ing, he said, that thor­oughly sur­prised him. Heather­ton’s re­sults did much more than pro­vide ad­di­tional con­fir­ma­tion that glu­cose is a vi­tal part of willpow­er; they helped solve the puz­zle over how glu­cose could work with­out global changes in the brain’s to­tal en­ergy use. Ap­par­ently ego de­ple­tion causes ac­tiv­ity to rise in some parts of the brain and to de­cline in oth­ers. Your brain does not stop work­ing when glu­cose is low. It stops do­ing some things and starts do­ing oth­ers. It re­sponds more strongly to im­me­di­ate re­wards and pays less at­ten­tion to long-term prospects.

    …The psy­chol­o­gists gave pre­pro­grammed Black­Ber­rys to more than 200 peo­ple go­ing about their daily rou­tines for a week. The phones went off at ran­dom in­ter­vals, prompt­ing the peo­ple to re­port whether they were cur­rently ex­pe­ri­enc­ing some sort of de­sire or had re­cently felt a de­sire. The painstak­ing study, led by Wil­helm Hof­mann, then at the Uni­ver­sity of Würzburg, col­lected more than 10,000 mo­men­tary re­ports from morn­ing un­til mid­night.

    De­sire turned out to be the norm, not the ex­cep­tion. Half the peo­ple were feel­ing some de­sire when their phones went off - to snack, to goof off, to ex­press their true feel­ings to their bosses - and an­other quar­ter said they had felt a de­sire in the past half-hour. Many of these de­sires were ones that the men and women were try­ing to re­sist, and the more willpower peo­ple ex­pend­ed, the more likely they be­came to yield to the next temp­ta­tion that came along. When faced with a new de­sire that pro­duced some I-wan­t-to-but-I-re­al­ly-should­n’t sort of in­ner con­flict, they gave in more read­ily if they had al­ready fended off ear­lier temp­ta­tions, par­tic­u­larly if the new temp­ta­tion came soon after a pre­vi­ously re­ported one. The re­sults sug­gested that peo­ple spend be­tween three and four hours a day re­sist­ing de­sire. Put an­other way, if you tapped four or five peo­ple at any ran­dom mo­ment of the day, one of them would be us­ing willpower to re­sist a de­sire. The most com­monly re­sisted de­sires in the phone study were the urges to eat and sleep, fol­lowed by the urge for leisure, like tak­ing a break from work by do­ing a puz­zle or play­ing a game in­stead of writ­ing a memo. Sex­ual urges were next on the list of most-re­sisted de­sires, a lit­tle ahead of urges for other kinds of in­ter­ac­tions, like check­ing Face­book. To ward off temp­ta­tion, peo­ple re­ported us­ing var­i­ous strate­gies. The most pop­u­lar was to look for a dis­trac­tion or to un­der­take a new ac­tiv­i­ty, al­though some­times they tried sup­press­ing it di­rectly or sim­ply tough­ing their way through it. Their suc­cess was de­cid­edly mixed. They were pretty good at avoid­ing sleep, sex and the urge to spend mon­ey, but not so good at re­sist­ing the lure of tele­vi­sion or the Web or the gen­eral temp­ta­tion to re­lax in­stead of work.

    …‘Good de­ci­sion mak­ing is not a trait of the per­son, in the sense that it’s al­ways there,’ Baumeis­ter says. ‘It’s a state that fluc­tu­ates.’ His stud­ies show that peo­ple with the best self­-con­trol are the ones who struc­ture their lives so as to con­serve willpow­er. They don’t sched­ule end­less back­-to-back meet­ings. They avoid temp­ta­tions like al­l-y­ou-can-eat buffets, and they es­tab­lish habits that elim­i­nate the men­tal effort of mak­ing choic­es. In­stead of de­cid­ing every morn­ing whether or not to force them­selves to ex­er­cise, they set up reg­u­lar ap­point­ments to work out with a friend. In­stead of count­ing on willpower to re­main ro­bust all day, they con­serve it so that it’s avail­able for emer­gen­cies and im­por­tant de­ci­sion­s….’Even the wis­est peo­ple won’t make good choices when they’re not rested and their glu­cose is low,’ Baumeis­ter points out. That’s why the truly wise don’t re­struc­ture the com­pany at 4 p.m. They don’t make ma­jor com­mit­ments dur­ing the cock­tail hour. And if a de­ci­sion must be made late in the day, they know not to do it on an empty stom­ach. ‘The best de­ci­sion mak­ers,’ Baumeis­ter says, ‘are the ones who know when not to trust them­selves.’

  56. Al­though that said, note that the blood sugar par­a­digm ap­pears to have fallen vic­tim to : , Kurzban et al 2013; “A Meta-Analy­sis of Blood Glu­cose Effects on Hu­man De­ci­sion Mak­ing”, Orquin & Kurzban 2016; “Is Ego-De­ple­tion a Replic­a­ble Effect? A Foren­sic Meta-Analy­sis of 165 Ego De­ple­tion Ar­ti­cles”.↩︎

  57. See “Effects of prior light ex­po­sure on early evening per­for­mance, sub­jec­tive sleepi­ness, and hor­monal se­cre­tion” (cov­er­age), Münch et al 2012:

    …For cog­ni­tive per­for­mance we found a sig­nifi­cant in­ter­ac­tion be­tween light con­di­tions, men­tal load (2- or 3-back task) and the or­der of light ad­min­is­tra­tion. On their first evening, sub­jects per­formed with sim­i­lar ac­cu­racy after both light con­di­tions, but on their sec­ond evening, sub­jects per­formed sig­nifi­cantly more ac­cu­rately after the DL in both n-back ver­sions and com­mit­ted fewer false alarms in the 2-back task com­pared to the AL group. Lower sleepi­ness in the evening was sig­nifi­cantly cor­re­lated with bet­ter cog­ni­tive per­for­mance (p < .05).

  58. “With re­gards to changes in n-back lev­el, I went up about 1 solid level on all the tasks that I trained. That is, I went from 7 to 8 for du­al, 6 to 7 for po­si­tion-sound-col­or, 6 to 7 for po­si­tion-sound-shape, and 4 to 5 on quad. I don’t use any strate­gies.”↩︎

  59. “Train­ing of Work­ing Mem­ory Im­pacts Struc­tural Con­nec­tiv­ity”, Takeuchi 2010 in .↩︎

  60. , “Lies, Damn Lies, and Chi­nese Sci­ence: The Peo­ple’s Re­pub­lic is be­com­ing a tech­no­log­i­cal su­per­pow­er, but who’s check­ing the facts? Sam Geall seeks out the Chi­nese sci­ence cops” (see also Lancet, Na­ture, NYT, Joe Hil­gard):

    This pub­lish-or-per­ish cul­ture has led to un­re­al­is­tic tar­gets at Chi­nese uni­ver­si­ties - and as a pre­dictable con­se­quence, ram­pant pla­gia­rism. In Jan­u­ary, the peer-re­viewed in­ter­na­tional jour­nal Sec­tion E an­nounced the re­trac­tion of more than 70 pa­pers by Chi­nese sci­en­tists who had fal­si­fied da­ta. Three months lat­er, the same pub­li­ca­tion an­nounced the re­moval of an­other 39 ar­ti­cles “as a re­sult of prob­lems with the data sets or in­cor­rect atom as­sign­ments”, 37 of which were en­tirely pro­duced in Chi­nese uni­ver­si­ties. The New Jer­sey-based Cen­te­nary Col­lege closed its affil­i­ated Chi­nese busi­ness school pro­gramme in July after a re­view “re­vealed ev­i­dence of wide­spread pla­gia­rism, among other is­sues, at a level that or­di­nar­ily would have re­sulted in stu­dents’ im­me­di­ate dis­missal from the col­lege.” A gov­ern­ment study, cited by Na­ture, found that about one-third of over 6,000 sci­en­tists sur­veyed at six top Chi­nese in­sti­tu­tions had prac­tised “pla­gia­rism, fal­si­fi­ca­tion or fab­ri­ca­tion”. But it’s not only the em­pha­sis on quan­tity that dam­ages sci­en­tific qual­ity in Chi­na. Pub­li­ca­tion bias - the ten­dency to priv­i­lege the re­sults of stud­ies that show a sig­nifi­cant find­ing, rather than in­con­clu­sive re­sults - is no­to­ri­ously per­va­sive. One sys­tem­atic re­view of acupunc­ture stud­ies from 1998, pub­lished in Con­trolled Clin­i­cal Tri­als, found that every sin­gle clin­i­cal trial orig­i­nat­ing in China was pos­i­tive - in other words, no trial pub­lished in China had found a treat­ment to be in­effec­tive.

    , “Tra­di­tional Chi­nese med­i­cine: Big ques­tions: Jour­nal re­ports of ben­e­fits often lack method­olog­i­cal rigor or de­tails”:

    fo­cuses ex­clu­sively on re­ports pub­lished since 1999 in Chi­nese aca­d­e­mic jour­nals, roughly half of which were spe­cialty pub­li­ca­tions. Clin­i­cians au­thored half of the pa­pers. Al­most 85% of the re­ports fo­cused on herbal reme­dies - any­thing from bulk herbs or pills to “de­coc­tions”. Most of the re­main­ing re­views as­sessed the value of acupunc­ture, al­though about 1% of the re­ports dealt with Tu­ina mas­sage…The pa­pers were re­views, or what are typ­i­cally re­ferred to in West­ern jour­nals as meta-analy­ses…­Many of the pa­pers were in­com­plete, roughly one-third con­tained sta­tis­ti­cal er­rors and oth­ers pro­vided data or com­par­isons that the au­thors termed mis­lead­ing. Fewer than half of the sur­veyed pa­pers de­scribed how the data they were pre­sent­ing had been col­lect­ed, how those data had been an­a­lyzed or how a de­ci­sion had been made about which stud­ies to com­pare. The ma­jor­ity of pa­pers also did not as­sess the risk of bias across stud­ies or offer any in­for­ma­tion on po­ten­tial con­flic­t-of-in­ter­est fac­tors (such as who funded or oth­er­wise offered sup­port for the re­search be­ing re­viewed)….Over­all, “the qual­ity of these re­views is trou­bling,” the Lanzhou re­searchers con­clude in the May 25 PLoS One.

  61. “Pla­gia­rism Plague Hin­ders Chi­na’s Sci­en­tific Am­bi­tion”, NPR:

    In 2008, when her sci­en­tific pub­li­ca­tion, the Jour­nal of Zhe­jiang Uni­ver­si­ty-Science, be­came the first in China to use Cross­Check text analy­sis soft­ware to spot pla­gia­rism, Zhang was pleased to be a trail­blaz­er. But when the first set of re­sults came in, she was up­set and hor­ri­fied. “In al­most 2 years, we find about 31% of pa­pers with un­rea­son­able copy­[ing] and pla­gia­rism,” she says, shak­ing her head. “This is true.” For com­puter sci­ence and life sci­ence pa­pers, that fig­ure went up to al­most 40 per­cent…De­spite the out­pour­ing of Chi­nese pa­pers, Chi­nese re­search is­n’t that in­flu­en­tial glob­al­ly. Thom­son Reuters’ Sci­ence Watch web­site notes that China is­n’t even in the top 20 when mea­sur­ing the num­ber of times a pa­per is cited on a na­tional ba­sis. Sci­enceNet’s Zhao says he fears Chi­nese re­search is still about quan­tity rather than qual­i­ty….How­ev­er, Chi­na’s lead­ers have com­mit­ted to fight­ing sci­en­tific fraud. And Zhang, the jour­nal ed­i­tor, says that one year on, pla­gia­rism at her pub­li­ca­tion has fallen no­tice­ably, to 24% of all sub­mis­sions.

    “Chi­na’s aca­d­e­mic scan­dal: call tol­l-free hot­lines to get your name pub­lished”; “Looks good on pa­per: A flawed sys­tem for judg­ing re­search is lead­ing to aca­d­e­mic fraud”; SAGE Pub­li­ca­tions busts ‘peer re­view and ci­ta­tion ring’, 60 pa­pers re­tracted”; out­sourced meta-analy­sis writ­ing.

    Pour­ing more money in seems to not be help­ing (“Fraud Scan­dals Sap Chi­na’s Dream of Be­com­ing a Sci­ence Su­per­power”), and the fox is in charge of the hen house.↩︎

  62. Ab­stract:

    We in­ves­ti­gated whether and how in­di­vid­ual differ­ences in per­son­al­ity de­ter­mine cog­ni­tive train­ing out­comes. 47 par­tic­i­pants were ei­ther trained on a sin­gle or on a dual n-back task for a pe­riod of 4 weeks. 52 ad­di­tional par­tic­i­pants did not re­ceive any train­ing and served as a no-con­tact con­trol group. We as­sessed neu­roti­cism and con­sci­en­tious­ness as per­son­al­ity traits as well as per­for­mance in near and far trans­fer mea­sures. The re­sults in­di­cated a sig­nifi­cant in­ter­ac­tion of neu­roti­cism and in­ter­ven­tion in terms of train­ing effi­ca­cy. Whereas dual n-back train­ing was more effec­tive for par­tic­i­pants low in neu­roti­cism, sin­gle n-back train­ing was more effec­tive for par­tic­i­pants high in neu­roti­cism. Con­sci­en­tious­ness was as­so­ci­ated with high train­ing scores in the sin­gle n-back and im­prove­ment in near trans­fer mea­sures, but lower far trans­fer per­for­mance, sug­gest­ing that sub­jects scor­ing high in this trait de­vel­oped task-spe­cific skills pre­vent­ing gen­er­al­iz­ing effects. We con­clude by propos­ing that in­di­vid­ual differ­ences in per­son­al­ity should be con­sid­ered in fu­ture cog­ni­tive in­ter­ven­tion stud­ies to op­ti­mize the effi­cacy of train­ing.

  63. Re­search pro­gram­mer Jonathan Graehl writes on the LW dis­cus­sion of Jaeggi 2011:

    …If you sep­a­rated the “ac­tive con­trol” group into high and low im­provers post-hoc just like was done for the n-back group, you might see that the ac­tive con­trol “high im­provers” are even smarter than the n-back “high im­provers”. We should ex­pect some 8-9 year olds to im­prove in in­tel­li­gence or mo­ti­va­tion over the course of a month or two, with­out any in­ter­ven­tion. Ba­si­cal­ly, this re­sult sucks, be­cause of the ar­ti­fi­cial post-hoc di­vi­sion into high- and low- re­spon­ders to n-back train­ing, needed to show a strong “effect”. I’m not cer­tain that the effect is ar­ti­fi­cial; I’d have to spend a lot of time do­ing some kind of sam­pling to show how well the data is ex­plained by my al­ter­na­tive hy­poth­e­sis.

  64. The DNB groups gain ~1 point (ques­tion), and the con­trol group falls ~2 points after start­ing off ~2 points high­er. In other words, if the con­trol group had not fallen so much, the DNB groups would at no point have scored high­er!

    Repli­cat­ing their re­sults, we found a sig­nifi­cant gain in Gf scores in the train­ing group over and above gains on the digit span task F(1, 26) = 3.00, P = 0.05, ηp2 = 0.10. In con­trast, the con­trol group showed a non-sig­nifi­cant de­crease in Gf, F<1, and the crit­i­cal group by time in­ter­ac­tion was sig­nifi­cant, F(1, 40) = 7.47, P = 0.01, ηp2 = 0.16. As can be seen in Fig­ure 3, there was a trend to­ward a sig­nifi­cant group differ­ence in Gf (RPM scores) at pre-train­ing, p≤0.10. This raises the pos­si­bil­ity that the rel­a­tive gains in Gf in the train­ing ver­sus con­trol groups may be to some ex­tent an arte­fact of base­line differ­ences. How­ev­er, the in­ter­ac­tive effect of trans­fer as a func­tion of group re­mained sig­nifi­cant even after more closely match­ing the train­ing and con­trol groups for pre-train­ing RPM scores (by re­mov­ing the high­est scor­ing con­trols) F(1, 30) = 3.66, P = 0.032, ηp2 = 0.10. The ad­justed means (s­tan­dard de­vi­a­tions) for the con­trol and train­ing groups were now 27.20 (1.93), 26.63 (2.60) at pre-train­ing (t(43) = 1.29, P>0.05) and 26.50 (4.50), 27.07 (2.16) at post-train­ing, re­spec­tive­ly. More­over, there was a trend for the gain in Gf to be pos­i­tively cor­re­lated with im­prove­ments in n-back per­for­mance across train­ing r(29) = 0.36 at P = 0.057, sug­gest­ing that such gains were in­deed a func­tion of train­ing….Al­though the Gf trans­fer­able gains we found ap­pear to be some­what re­lated to train­ing gains and the effects re­main when we trim the groups to pro­vide a bet­ter match for pre-train­ing Gf, it is im­por­tant to note that some de­gree of re­gres­sion to the mean may be in­flu­enc­ing the re­sults.

  65. At least, they seem to ad­min­is­ter the whole thing with no men­tion of such a vari­a­tion:

    We as­sessed Gf with the Raven’s Pro­gres­sive Ma­tri­ces (RPM; [35]) - a stan­dard mea­sure in the lit­er­a­ture. Each RPM item pre­sented par­tic­i­pants with a ma­trix of vi­sual pat­terns with one pat­tern miss­ing. The par­tic­i­pant chose how the ma­trix should be com­pleted by se­lect­ing a pat­tern from a se­ries of al­ter­na­tives. We used par­al­lel ver­sions of the RPM (even and un­even num­bered pages), which we coun­ter­bal­anced across par­tic­i­pants and pre- and post-train­ing. The RPM is scored on a scale from 0-30, with each cor­rect ma­trix earn­ing par­tic­i­pants one point.

  66. From the pa­per:

    The fig­ure de­picts a block of the emo­tional ver­sion of the dual n-back task (train­ing task) where n = 1. The top row shows the se­quence across tri­als (A, B, C, D, etc.) of vi­su­ally pre­sented stim­uli in a 4×4 grid (the vi­sual stim­uli were pre­sented on a stan­dard 1280×1024 pixel com­puter dis­play). A pic­ture of a face ap­peared in one of the 16 pos­si­ble grid po­si­tions on each tri­al. Si­mul­ta­ne­ous­ly, with the pre­sen­ta­tion of these vi­sual stim­uli on the com­puter dis­play, par­tic­i­pants heard words over head­phones (sec­ond row in the fig­ure). Par­tic­i­pants were re­quired to in­di­cate, by but­ton press, whether the trial was a ‘tar­get trial’ or not. Tar­gets could be vi­sual or au­di­to­ry. In the ex­am­ple here, Trial C is a vi­sual tar­get. That is, the face in Trial C is pre­sented in the same lo­ca­tion as the face in Trial B (i.e., n = 1 po­si­tions back). Note, the faces are of differ­ent ac­tors. For vi­sual stim­uli par­tic­i­pants were asked to ig­nore the con­tent of the im­age and solely at­tend to the lo­ca­tion in which the im­ages were pre­sent­ed. In the cur­rent ex­am­ple, Trial D was an au­di­tory tar­get trial be­cause ‘Evil’ is the same word as the word pre­sented in Trial C - n po­si­tions back (where n = 1). Each block con­sisted of 20+n tri­als.

    (If you look at Fig­ure 1, ex­am­ple stim­uli words are ‘dead’, ‘hate’, ‘evil’, ‘rape’, ‘slum’, and a pic­ture of a very an­gry male face.)↩︎

  67. The differ­ence does­n’t seem to change progress on n-back in ei­ther group, which is good since if there were differ­ences, that would be trou­bling eg. if the affec­tive n-back group did­n’t in­crease as many lev­els, that would make any fol­low­ing re­sults more du­bi­ous:

    Per­for­mance of the two n-back groups pre- to post- train­ing did not differ sig­nifi­cantly on ei­ther the neu­tral F(1, 27) = 1.02, P>0.05 or affec­tive F (1, 27)<1 n-back tasks. Sim­i­lar­ly, the con­trol group showed a sig­nifi­cantly greater pre- to post-train­ing im­prove­ment on the fea­ture match task they trained on, com­pared with the n-back groups F(1, 42) = 41.09, P<0.001, ηp2 = 0.67.

    And as one would hope, both DNB groups in­creased their WM scores:

    As pre­dict­ed, par­tic­i­pants in the train­ing group showed a sig­nifi­cant im­prove­ment on digit span F(1, 28) = 33.96, p < 0.001, ηp2 = 0.55. How­ev­er, this was not true of con­trols F(1, 15) = 1.89, p = 0.19, ηp2 = 0.11, and the gain was sig­nifi­cantly greater in the train­ing group par­tic­i­pants com­pared to con­trols F(1,43) = 5.92, p = 0.02, ηp2 = 0.12.

  68. Al­loway 2009, “The effi­cacy of work­ing mem­ory train­ing in im­prov­ing crys­tal­lized in­tel­li­gence” (PDF) 7 chil­dren with learn­ing dis­abil­i­ties re­ceived the train­ing for 8 weeks; Gc was mea­sured us­ing the vo­cab­u­lary & math sec­tions of the Wech­sler IQ test.↩︎

  69. The prac­tice effect can last for many years. “In­flu­ence of Age on Prac­tice Effects in Lon­gi­tu­di­nal Neu­rocog­ni­tive Change”, Salt­house 2010:

    of neu­rocog­ni­tive func­tion­ing often re­veal sta­bil­ity or age-re­lated in­creases in per­for­mance among adults un­der about 60 years of age. Be­cause nearly mo­not­o­nic de­clines with in­creas­ing age are typ­i­cally ev­i­dent in , there is a dis­crep­ancy in the in­ferred age trends based on the two types of com­par­ison­s….In­creased age was as­so­ci­ated with sig­nifi­cantly more neg­a­tive lon­gi­tu­di­nal changes with each abil­i­ty. All of the es­ti­mated prac­tice effects were pos­i­tive, but they var­ied in mag­ni­tude across neu­rocog­ni­tive abil­i­ties and as a func­tion of age. After ad­just­ing for prac­tice effects the lon­gi­tu­di­nal changes were less pos­i­tive at younger ages and slightly less neg­a­tive at older ages. Con­clu­sions: It was con­cluded that some, but not all, of the dis­crep­ancy be­tween cross-sec­tional and lon­gi­tu­di­nal age trends in neu­rocog­ni­tive func­tion­ing is at­trib­ut­able to prac­tice effects pos­i­tively bi­as­ing the lon­gi­tu­di­nal trends.

  70. To­fu: “I should also add, my score on the num­ber test jumped dra­mat­i­cally from the first test to the sec­ond test prob­a­bly be­cause I taught my­self how to do long di­vi­sion be­fore the sec­ond test (which was the only study­ing I did for all 3 test­s).”↩︎

  71. Ship­stead, Redick, & En­gle 2012 men­tion an amus­ing study I had­n’t heard of be­fore:

    Green­wald et al. (1991) pro­vided a use­ful demon­stra­tion of the prob­lems as­so­ci­ated with sub­jec­tive re­ports. Par­tic­i­pants in this study re­ceived com­mer­cially pro­duced au­dio­tapes that con­tained sub­lim­i­nal mes­sages in­tended to im­prove ei­ther self­-es­teem or mem­o­ry. Un­known to the par­tic­i­pants, half of the tapes that were de­signed to im­prove mem­ory were re­la­beled “self­-es­teem” and vice ver­sa. At a 5-week posttest, par­tic­i­pants’ scores on sev­eral stan­dard mea­sures of self­-es­teem and mem­ory were im­proved, but this change was in­de­pen­dent of the mes­sage and the la­bel on the au­dio­tape (i.e., par­tic­i­pants showed across the board im­prove­men­t). How­ev­er, in re­sponse to sim­ple ques­tions re­gard­ing per­ceived effects, roughly 50% of par­tic­i­pants re­ported ex­pe­ri­enc­ing im­prove­ments that were con­sis­tent with the la­bel on the au­dio­tape, while only 15% re­ported im­prove­ments in the op­po­site do­main. The self­-re­port mea­sures were nei­ther re­lated to ac­tual im­prove­ments in trans­fer task per­for­mance nor re­lated to the con­tent of the in­ter­ven­tion. In­stead, they were at­trib­ut­able to ex­pec­ta­tion of out­come.

  72. from at 90 days see­ing lit­tle effect, to 2.5 months later pro­duc­ing the sec­ond tes­ta­ment↩︎

  73. “At­ten­tion and Work­ing Mem­ory in In­sight Prob­lem-Solv­ing”, Mur­ray 2011. The study does not seem to have con­trolled for IQ, so it’s hard to say whether the WM/at­ten­tion are re­spon­si­ble for in­creased per­for­mance or not.↩︎

  74. From Sander­berg/­Bostrom 2006:

    Giv­ing , a dopamine pre­cur­sor, to healthy vol­un­teers did not affect di­rect se­man­tic prim­ing (faster recog­ni­tion of words di­rectly se­man­ti­cally re­lated to a pre­vi­ous word, such as “black­-white”) but did in­hibit in­di­rect prim­ing (faster recog­ni­tion of more se­man­ti­cally dis­tant words, such as “sum­mer-s­now”) (Kischka et al. 1996). This was in­ter­preted by the au­thors of the study as dopamine in­hibit­ing the spread of ac­ti­va­tion within the se­man­tic net­work, that is, a fo­cus­ing on the task.

  75. “Tem­pera­ment and char­ac­ter cor­re­lates of neu­ropsy­cho­log­i­cal per­for­mance”, June 2010, Psy­cho­log­i­cal So­ci­ety of South Africa↩︎

  76. Jaeggi 2008’s notes say the daily train­ing was ~25 min­utes; the longest group was 19 days; hours.↩︎

  77. See the crit­i­cal re­view of WM train­ing re­search, “Does work­ing mem­ory train­ing gen­er­al­ize?” (Ship­stead et al 2010).↩︎

  78. The R code:

    on <- c(35,31,27,66,25,38,35,43,60,47,38,58,50,23,50,45,60,37,22,28,50,20,41,42,47,55,47,42,35,
    off <- c(17,43,46,50,36,31,38,33,66,30,68,42,40,29,69,40,41,45,37,18,44,60,31,46,46,45,27,35,45,
    # [1] 158
    mcmcChain = BESTmcmc(off, on)
    postInfo = BESTplot(off, on, mcmcChain) # image
    #            SUMMARY.INFO
    # PARAMETER       mean   median     mode  HDIlow  HDIhigh pcgtZero
    #   mu1       40.96178 40.95536 40.93523 38.1887  43.7104       NA
    #   mu2       41.37400 41.37365 41.39874 38.8068  44.0550       NA
    #   muDiff    -0.41222 -0.41368 -0.45968 -4.2497   3.3593    41.54
    #   sigma1    12.32844 12.27614 12.28283 10.3024  14.4116       NA
    #   sigma2    11.21408 11.15464 10.99812  9.2924  13.1895       NA
    #   sigmaDiff  1.11436  1.10736  0.94511 -1.6011   3.9756    78.73
    #   nu        45.65240 37.49245 22.16426  5.3586 108.1555       NA
    #   nuLog10    1.56504  1.57394  1.61572  0.9956   2.1157       NA
    #   effSz     -0.03528 -0.03519 -0.03547 -0.3588   0.2851    41.54

    For those who pre­fer a reg­u­lar two-sam­ple test:

    #     Wilcoxon rank sum test with continuity correction
    # data:  off and on
    # W = 3004, p-value = 0.7066
    # alternative hypothesis: true location shift is not equal to 0
  79. “The Psy­chophys­i­ol­ogy of Lu­cid Dream­ing”, col­lected in Con­scious Mind, Sleep­ing Brain.↩︎

  80. pg 97 of his book The Dream Drug­store (2001)↩︎

  81. ; ab­stract:

    Func­tional neu­roimag­ing stud­ies car­ried out on healthy vol­un­teers while per­form­ing differ­ent n-back tasks have shown a com­mon pat­tern of bi­lat­eral fron­topari­etal ac­ti­va­tion, es­pe­cially of the dor­so­lat­eral pre­frontal cor­tex (DLPFC). Our ob­jec­tive was to use func­tional mag­netic res­o­nance imag­ing (fMRI) to com­pare the pat­tern of brain ac­ti­va­tion while per­form­ing two sim­i­lar n-back tasks which differed in their pre­sen­ta­tion modal­i­ty. Thir­teen healthy vol­un­teers com­pleted a ver­bal 2-back task pre­sent­ing au­di­tory stim­uli, and a sim­i­lar 2-back task pre­sent­ing vi­sual stim­uli. A con­junc­tion analy­sis showed bi­lat­eral ac­ti­va­tion of fron­topari­etal ar­eas in­clud­ing the DLPFC. The left DLPFC and the su­pe­rior tem­po­ral gyrus showed a greater ac­ti­va­tion in the au­di­tory than in the vi­sual con­di­tion, whereas pos­te­rior brain re­gions and the an­te­rior cin­gu­late showed a greater ac­ti­va­tion dur­ing the vi­sual than dur­ing the au­di­tory task. Thus, brain ar­eas in­volved in the vi­sual and au­di­tory ver­sions of the n-back task showed an im­por­tant over­lap be­tween them, re­flect­ing the supramodal char­ac­ter­is­tics of work­ing mem­o­ry. How­ev­er, the differ­ences found be­tween the two modal­i­ties should be con­sid­ered in or­der to se­lect the most ap­pro­pri­ate task for fu­ture clin­i­cal stud­ies.

  82. Schnei­der, B, Pi­chora-Fuller, MK. “Im­pli­ca­tions of per­cep­tual de­te­ri­o­ra­tion for cog­ni­tive ag­ing Re­search”. In: Craik, FI, Salt­house, TA, ed­i­tors. The hand­book of ag­ing and cog­ni­tion, Psy­chol­ogy Press, 2000. ISBN-10: 080585990X↩︎

  83. Ab­stract: “The au­thors in­ves­ti­gated the dis­tinc­tive­ness and in­ter­re­la­tion­ships among vi­su­ospa­tial and ver­bal mem­ory processes in short­-term, work­ing, and long-term mem­o­ries in 345 adults. Be­gin­ning in the 20s, a con­tin­u­ous, reg­u­lar de­cline oc­curs for pro­cess­ing-in­ten­sive tasks (e.g., speed of pro­cess­ing, work­ing mem­o­ry, and long-term mem­o­ry), whereas ver­bal knowl­edge in­creases across the life span [Be­sides Salt­house, for the ver­bal flu­ency claim see Schaie, K. W. In­tel­lec­tual De­vel­op­ment in Adult­hood: The Seat­tle Lon­gi­tu­di­nal Study. Cam­bridge Uni­ver­sity Press, 1996]. There is lit­tle differ­en­ti­a­tion in the cog­ni­tive ar­chi­tec­ture of mem­ory across the life span. Vi­su­ospa­tial and ver­bal work­ing mem­ory are dis­tinct but highly in­ter­re­lated sys­tems with do­main-spe­cific short­-term mem­ory sub­sys­tems. In con­trast to re­cent neu­roimag­ing data, there is lit­tle ev­i­dence for ded­iffer­en­ti­a­tion of func­tion at the be­hav­ioral level in old com­pared with young adults.” That the neu­roimag­ing shows no change in gen­eral lo­ca­tions of ac­tiv­ity is prob­a­bly in­ter­pretable as the lower per­for­mance be­ing due to gen­eral low-level prob­lems and in­effi­cien­cies of age, and not the el­der­ly’s brains start­ing to ‘un­learn’ spe­cific tasks.↩︎

  84. “The Z-s­core rep­re­sents the age-con­tin­gent mean, mea­sured in units of stan­dard de­vi­a­tion rel­a­tive to the pop­u­la­tion mean. More pre­cise­ly, the Z-s­core is (age-con­tin­gent mean mi­nus pop­u­la­tion mean) / (pop­u­la­tion stan­dard de­vi­a­tion).” –A­gar­wal et al 2009↩︎

  85. , Agar­wal et al 2009 (slides):

    …The preva­lence of de­men­tia ex­plodes after age 60, dou­bling with every 5 years of age.5 In the co­hort above age 85, the preva­lence of de­men­tia ex­ceeds 30%. More­over, many older adults with­out a strict di­ag­no­sis of de­men­tia, still ex­pe­ri­ence sub­stan­tial cog­ni­tive im­pair­ment. For ex­am­ple, the preva­lence of the di­ag­no­sis “cog­ni­tive im­pair­ment with­out de­men­tia” is nearly 30% be­tween ages 80 and 89.6 Draw­ing these facts to­geth­er, among the pop­u­la­tion be­tween ages 80 and 89, about half of the pop­u­la­tion ei­ther has a di­ag­no­sis of de­men­tia or cog­ni­tive im­pair­ment with­out de­men­tia.

    …Third, us­ing a new dataset, we doc­u­ment a link be­tween age and the qual­ity of fi­nan­cial de­ci­sion-mak­ing in debt mar­kets. In a cross-sec­tion of prime bor­row­ers, mid­dle-aged adults bor­row at lower in­ter­est rates and pay fewer fees rel­a­tive to younger and older adults. Av­er­ag­ing across ten credit mar­kets, fee and in­ter­est pay­ments are min­i­mized around age 53. The mea­sured effects are not ex­plained by ob­served risk char­ac­ter­is­tics. Com­bin­ing mul­ti­ple data sets we do not find ev­i­dence that se­lec­tion effects and co­hort effects ex­plain our re­sults. The lead­ing ex­pla­na­tion for the pat­terns that we ob­serve is that ex­pe­ri­ence rises with age, but an­a­lyt­i­cal abil­i­ties de­cline with it.

    …Neu­ro­log­i­cal patholo­gies rep­re­sent one im­por­tant path­way for age effects in older adults. For in­stance, de­men­tia is pri­mar­ily at­trib­ut­able to Alzheimer’s Dis­ease (60%) and vas­cu­lar dis­ease (25%). The preva­lence of de­men­tia dou­bles with every five ad­di­tional years of life­cy­cle age (Ferri et al., 2006; Fratiglioni, De Ronchi, and Agüero-Tor­res, 1999).10 For ex­am­ple, Ta­ble 1 re­ports that the preva­lence of de­men­tia in North Amer­ica rises from 3.3% for adults ages 70–74, to 6.5% for adults ages 75–79, to 12.8% for adults ages 80–84, to 30.1% for adults at least 85 years of age (Ferri et al. 2006). Many older adults also suffer from a less se­vere form of cog­ni­tive im­pair­ment, which is di­ag­nosed as “cog­ni­tive im­pair­ment with­out de­men­tia.” For ex­am­ple, the preva­lence of this di­ag­no­sis rises from 16.0% for adults ages 71-79, to 29.2% for adults ages 80-89.

    • 10: There is also grow­ing lit­er­a­ture that iden­ti­fies age-re­lated changes in the na­ture of cog­ni­tion (see Park and Schwarz, 1999 [Cog­ni­tive Ag­ing: A Primer]; and Den­burg, Tranel, and Bechara 2005). Mather and Carstensen (2005) and iden­tify age-vari­a­tion in cog­ni­tive pref­er­ences. Sub­jects with short time hori­zons or older ages at­tend to neg­a­tive in­for­ma­tion rel­a­tively less than sub­jects with long time hori­zons or younger ages.

    …Fig­ure 4d plots naive and con­trol per­for­mance in the Tele­phone In­ter­view of Cog­ni­tive Sta­tus (TICS) task. This task asks the re­spon­dent ten triv­ial ques­tions and as­signs one point for each cor­rect an­swer: What is the cur­rent year? Mon­th? Day? Day of the week? What do you usu­ally use to cut pa­per? What do you call the kind of prickly plant that grows in the de­sert? Who is the cur­rent pres­i­dent? Vice pres­i­dent? Count back­wards from twenty to ten (twice). At age 63, the av­er­age score is 9.2 out of 10. By age 90, the av­er­age (con­trol) score is 7.5. Fi­nal­ly, we present two mea­sures of prac­ti­cal nu­mer­a­cy. 4e plots naive and con­trol per­for­mance in re­sponse to the ques­tion: If the chance of get­ting a dis­ease is 10 per­cent, how many peo­ple out of 1,000 would be ex­pected to get the dis­ease? At age 53, 79% an­swer cor­rect­ly. By age 90, 50% an­swer cor­rect­ly. Fig­ure 4f plots naive and con­trol per­for­mance in re­sponse to the ques­tion: If 5 peo­ple all have the win­ning num­bers in the lot­tery and the prize is two mil­lion dol­lars, how much will each of them get? We be­lieve that this ques­tion is im­pre­cisely posed, since the log­i­cal an­swer could be ei­ther $2,000,000 or $400,000. How­ev­er, the re­sults are still in­ter­est­ing, since the frac­tion an­swer­ing $400,000 (the offi­cial cor­rect an­swer) drops pre­cip­i­tous­ly. At age 53, 52% an­swer $400,000. By age 90, 10% give this an­swer.

    …For the 1989, 1998, 2001, and 2004 sur­veys, we com­pute the ra­tios of in­come, ed­u­ca­tion, and net worth for bor­row­ers to the pop­u­la­tion as a whole, by age group; re­sults are pre­sented in the on­line ap­pen­dix. We find that within age groups, bor­row­ers al­most al­ways have higher lev­els of in­come and ed­u­ca­tion than the pop­u­la­tion as a whole, and often have higher lev­els of net worth. More­over, older bor­row­ers ap­pear to have rel­a­tively higher lev­els of in­come and ed­u­ca­tion rel­a­tive to their peers than mid­dle-aged bor­row­ers do. Hence these data sug­gest that se­lec­tion effects by age go in the op­po­site di­rec­tion: older bor­row­ers ap­pear to be a bet­ter pool than mid­dle-aged bor­row­ers. We present ad­di­tional re­sults in the on­line ap­pen­dix show­ing that bor­row­ing by age does not ap­pear to vary by race, and that older bor­row­ers do not ap­pear to have dis­pro­por­tion­ately lower in­comes, FICO score, or higher debt lev­els. None of these analy­ses lend sup­port to the idea that sam­ple se­lec­tion effects con­tribute to the U-shape pat­terns that we see in the da­ta.

    …The effects we find have a wide range of dol­lar mag­ni­tudes, re­ported in Ta­ble 4. We es­ti­mate that, for home­-e­quity lines of cred­it, 75-year-olds pay about $265 more each year than 50-year-olds, and 25-year-olds pay about $295 more. For other quan­ti­ties, say, credit card fees, the im­plied age differ­en­tials are small - roughly $10-$20 per year for each kind of fee. The im­por­tance of the U-shaped effects we es­ti­mate goes be­yond the eco­nomic sig­nifi­cance of each in­di­vid­ual choice, how­ev­er: it lies in the fact that the ap­pear­ance of a U-shaped pat­tern of costs in such a wide va­ri­ety of cir­cum­stances points to a phe­nom­e­non that might ap­ply to many ar­eas.

  86. The prac­tice effect can last for many years. “In­flu­ence of Age on Prac­tice Effects in Lon­gi­tu­di­nal Neu­rocog­ni­tive Change”, Salt­house 2010:

    of neu­rocog­ni­tive func­tion­ing often re­veal sta­bil­ity or age-re­lated in­creases in per­for­mance among adults un­der about 60 years of age. Be­cause nearly mo­not­o­nic de­clines with in­creas­ing age are typ­i­cally ev­i­dent in , there is a dis­crep­ancy in the in­ferred age trends based on the two types of com­par­ison­s….In­creased age was as­so­ci­ated with sig­nifi­cantly more neg­a­tive lon­gi­tu­di­nal changes with each abil­i­ty. All of the es­ti­mated prac­tice effects were pos­i­tive, but they var­ied in mag­ni­tude across neu­rocog­ni­tive abil­i­ties and as a func­tion of age. After ad­just­ing for prac­tice effects the lon­gi­tu­di­nal changes were less pos­i­tive at younger ages and slightly less neg­a­tive at older ages. Con­clu­sions: It was con­cluded that some, but not all, of the dis­crep­ancy be­tween cross-sec­tional and lon­gi­tu­di­nal age trends in neu­rocog­ni­tive func­tion­ing is at­trib­ut­able to prac­tice effects pos­i­tively bi­as­ing the lon­gi­tu­di­nal trends.

  87. Per­haps sur­pris­ing­ly, the com­mon wis­dom that peo­ple adopt con­ser­v­a­tive at­ti­tudes as part of the ag­ing process may not be cor­rect, and the ob­served con­ser­vatism of old peo­ple due to their com­ing from a more con­ser­v­a­tive time (ie. the past, as the 20th cen­tury saw a grand sweep of lib­eral be­liefs through First World so­ci­eties); “Pop­u­la­tion Ag­ing, In­tra­co­hort Ag­ing, and So­ciopo­lit­i­cal At­ti­tudes”, Danige­lis et al 2007’s ab­stract (ex­cerpts):

    Pre­vail­ing stereo­types of older peo­ple hold that their at­ti­tudes are in­flex­i­ble or that ag­ing tends to pro­mote in­creas­ing con­ser­vatism in so­ciopo­lit­i­cal out­look. In spite of mount­ing sci­en­tific ev­i­dence demon­strat­ing that learn­ing, adap­ta­tion, and re­assess­ment are be­hav­iors in which older peo­ple can and do en­gage, the stereo­type per­sists. We use U.S. Gen­eral So­cial Sur­vey data from 25 sur­veys be­tween 1972 and 2004 to for­mally as­sess the mag­ni­tude and di­rec­tion of changes in at­ti­tudes that oc­cur within co­horts at differ­ent stages of the life course. We de­com­pose changes in so­ciopo­lit­i­cal at­ti­tudes into the pro­por­tions at­trib­ut­able to co­hort suc­ces­sion and in­tra­co­hort ag­ing for three cat­e­gories of items: at­ti­tudes to­ward his­tor­i­cally sub­or­di­nate groups, civil lib­er­ties, and pri­va­cy. We find that sig­nifi­cant in­tra­co­hort change in at­ti­tudes oc­curs in co­hort­s-in-later-stages (age 60 and old­er) as well as co­hort­s-in-ear­lier-stages (ages 18 to 39), that the change for co­hort­s-in-later-stages is fre­quently greater than that for co­hort­s-in-ear­lier-stages, and that the di­rec­tion of change is most often to­ward in­creased tol­er­ance rather than in­creased con­ser­vatism. These find­ings are dis­cussed within the con­text of pop­u­la­tion ag­ing and de­vel­op­ment.

  88. “Cog­ni­tive De­cline Be­gins In Late 20s, Study Sug­gests”, Sci­ence Daily↩︎

  89. “This Is Your Brain. Ag­ing. Sci­ence is re­shap­ing what we know about get­ting old­er. (The news is bet­ter than you think.)”, Newsweek:

    The [Salt­house] graph shows two roller-coas­t­er­ing lines. One rep­re­sents the pro­por­tion of peo­ple of each age who are in the top 25% on a stan­dard lab test of rea­son­ing abil­i­ty-think­ing. The other shows the pro­por­tion of CEOs of com­pa­nies of each age. Rea­son­ing abil­ity peaks at about age 28 and then plum­mets, trac­ing that well-known plunge that makes those older than 30 (OK, fine, 40) cringe: only 6% of top scor­ers are in their 50s, and only 4% are in their 60s. But the age dis­tri­b­u­tion of CEOs is an al­most per­fect mir­ror im­age: it peaks just be­fore age 60. About half are older than 55. And the num­ber un­der 40 is about ze­ro.

    …Salt-house de­duces more coun­ter­in­tu­itive, and hope­ful, lessons. The first is that in real life, rather than in psych labs, peo­ple rely on men­tal abil­i­ties that stand up very well to age and dis­cover work-arounds for the men­tal skills that do fade.

  90. “How to Gain Eleven IQ Points in Ten Min­utes: Think­ing Aloud Im­proves Raven’s Ma­tri­ces Per­for­mance in Older Adults”, Fox et al 2009:

    “Few stud­ies have ex­am­ined the im­pact of age on re­ac­tiv­ity to con­cur­rent think-aloud (TA) ver­bal re­ports. An ini­tial study with 30 younger and 31 older adults re­vealed that think­ing aloud im­proves older adult per­for­mance on a short form of the Raven’s Ma­tri­ces (Bors & Stokes, 1998, Ed­u­ca­tional and Psy­cho­log­i­cal Mea­sure­ment, 58, p. 382) but did not affect other tasks. In the repli­ca­tion ex­per­i­ment, 30 older adults (mean age = 73.0) per­formed the Raven’s Ma­tri­ces and three other tasks to repli­cate and ex­tend the find­ings of the ini­tial study. Once again older adults per­formed sig­nifi­cantly bet­ter only on the Raven’s Ma­tri­ces while think­ing aloud. Per­for­mance gains on this task were sub­stan­tial (d = 0.73 and 0.92 in Ex­per­i­ments 1 and 2, re­spec­tive­ly), cor­re­spond­ing to a fluid in­tel­li­gence in­crease of nearly one stan­dard de­vi­a­tion.”

  91. Some rel­e­vant ex­cerpts:

    Buschkuehl et al. (2008) pro­posed an adap­tive vi­sual WM train­ing pro­gram to old-old adults: Their re­sults showed sub­stan­tial gains in the WM trained tasks. Short and long-term trans­fer effects were found only for tasks with the same stim­uli con­tent. Sim­i­lar­ly, Li et al. (2008) found in young and older adults spe­cific im­prove­ment in the task prac­ticed-a spa­tial 2 n-back WM task-that in­volved two con­di­tions: one stan­dard, one more de­mand­ing. Trans­fer effects were found on a more de­mand­ing 3 n-back vi­sual task as well as on nu­mer­i­cal n-back tasks. Al­though near trans­fer effects to the same (vi­su­al) and also differ­ent (nu­mer­i­cal) modal­ity were shown, no far trans­fer effects to more com­plex WM tasks (op­er­a­tion and ro­ta­tion span tests) were found. With re­gard to main­te­nance effects, Buschkuehl et al. (2008) failed to find any main­te­nance 1 year after com­ple­tion of train­ing, in com­par­i­son with pretest. In con­trast, Li et al. (2008) showed a main­te­nance of prac­tice gains and of near-trans­fer effects at 3-month fol­low-up; nonethe­less, in con­trast with young adults, older par­tic­i­pants showed a per­for­mance decre­ment from post­prac­tice to fol­low-up.

    …Com­mon mea­sures used in cog­ni­tive ag­ing re­search, and the­o­ret­i­cally re­lated to WM, were cho­sen: short­-term mem­o­ry, fluid in­tel­li­gence, in­hi­bi­tion, and pro­cess­ing speed (Craik & Salt­house, 2000; Ver­haeghen, Steitz, Sli­win­ski, & Cerel­la, 2003). For near­est-trans­fer effects, a vi­su­ospa­tial WM task (Dot Ma­trix task; adapted from Miyake, Fried­man, Ret­tinger, Shah, & Hegar­ty, 2001) was in­clud­ed. This task in­volves processes (e­lab­o­ra­tion and pro­cess­ing phase) sim­i­lar to the one prac­ticed. How­ev­er, the na­ture of the ma­te­r­ial and the sec­ondary re­quire­ment are differ­ent from those of the trained task. The For­ward and Back­ward Digit Span tests were used to as­sess near-trans­fer effects be­cause they are part of the gen­eral mem­ory fac­tor, but the task re­quests were differ­ent from those of the WM tasks (see Bopp & Ver­haeghen, 2005). Be­cause these tasks mea­sure the same nar­row or same broad abil­i­ty, we ex­pect trans­fer effects onto them. To de­ter­mine the pres­ence of far trans­fer effects, we chose clas­sic tasks: the Cat­tell task to mea­sure non­ver­bal rea­son­ing abil­i­ty; the Stroop Color test to in­dex in­hi­bi­tion-re­lated mech­a­nisms; and the Pat­tern Com­par­i­son test to as­sess pro­cess­ing speed. The trans­fer abil­i­ties were cho­sen with con­sid­er­a­tion of their re­la­tion­ship to WM process­es. Work­ing mem­ory im­pair­ment in older adults is gen­er­ally at­trib­uted to gen­eral mech­a­nisms such as in­hi­bi­tion and pro­cess­ing speed (Borella et al., 2008). Fur­ther­more, WM is fre­quently ad­vanced as one of the mech­a­nisms that also ac­counts for age-re­lated differ­ences in in­tel­li­gence tasks (de Rib­aupierre & Lecerf, 2006; Rab­bitt & Lowe, 2000; Schaie & Hert­zog, 1986)…

    The Cat­e­go­riza­tion Work­ing Mem­ory Span task (CWMS; Borella et al. 2008; De Beni, Borel­la, Car­ret­ti, Marigo, & Nava, 2008) is sim­i­lar to the clas­sic WM tasks, such as the Lis­ten­ing Span test (Borella et al., 2008), the only differ­ence be­ing that it in­volves pro­cess­ing lists of words rather than sen­tences, lim­it­ing the role of se­man­tic pro­cess­ing. The ma­te­ri­als con­sisted of 10 sets of words, each set com­pris­ing 20 lists of words, which were or­ga­nized in se­ries of word lists of differ­ent lengths (from 2 to 6). Each list con­tained 5 words of high­-medium fre­quen­cy. Fur­ther­more, the lists con­tained ze­ro, one, or two an­i­mal nouns, present in any po­si­tion, in­clud­ing last. An ex­am­ple list is house, moth­er, dog, word, night. Of the to­tal num­ber of words (200) in the task, 28% were an­i­mal words. Par­tic­i­pants lis­tened to the lists of words au­diorecorded pre­sented at a rate of 1 s per word and had to tap their hand on the ta­ble when­ever they heard an an­i­mal noun (pro­cess­ing phase). The in­ter­val be­tween se­ries of word lists was 2 s (the pre­sen­ta­tion was thus paced by the ex­per­i­menter). At the end of the se­ries, par­tic­i­pants re­called the last word of each string in se­r­ial or­der (main­te­nance phase). Two prac­tice tri­als of 2-word length were given be­fore the ex­per­i­ment start­ed. Words re­called were writ­ten down by the ex­per­i­menter on a pre­pared form. The to­tal num­ber of cor­rectly re­called words was used as the mea­sure of WM per­for­mance (max­i­mum score 20). This score has been demon­strated to show large cor­re­la­tions with vi­su­ospa­tial (Jig­saw Puz­zle test) and ver­bal (Lis­ten­ing Span test) WM tasks (Borella et al., 2008), and mea­sures of fluid in­tel­li­gence (Borella et al., 2006).

    …Cul­ture Fair test, Scale 3 (Cat­tell & Cat­tell, 1963). Scale 3 of the Cat­tell test con­sists of two par­al­lel forms (A and B), each con­tain­ing four sub­tests to be com­pleted in 2.5 to 4 min, de­pend­ing on the sub­test. In the first sub­test, Se­ries, par­tic­i­pants saw an in­com­plete se­ries of ab­stract shapes and fig­ures and had to choose from six al­ter­na­tives that best com­pleted the se­ries. In the sec­ond sub­test, Clas­si­fi­ca­tions, par­tic­i­pants saw 14 prob­lems com­pris­ing ab­stract shapes and fig­ures and had to choose which 2 of the 5 differed from the other 3. In the third sub­test, Ma­tri­ces, par­tic­i­pants were pre­sented with 13 in­com­plete ma­tri­ces con­tain­ing four to nine boxes of ab­stract fig­ures and shapes plus an empty box and six choic­es: Their task was to se­lect the an­swer that cor­rectly com­pleted each ma­trix. In the fi­nal sub­test, Con­di­tions, par­tic­i­pants were pre­sented with 10 sets of ab­stract fig­ures, lines, and a sin­gle dot, along with five al­ter­na­tives: Their task was to as­sess the re­la­tion­ship among the dot, fig­ures, and lines, then choose the al­ter­na­tive in which a dot could be po­si­tioned in the same re­la­tion­ship. The de­pen­dent vari­able was the num­ber of cor­rectly solved items across the four sub­sets (max­i­mum score of 50). One of the two par­al­lel forms (A or B) was ad­min­is­tered at pretest, the other at posttest in coun­ter­bal­anced fash­ion across test­ing ses­sions.

    …Far-trans­fer effect. For the Cat­tell test, re­sults in­di­cated that trained par­tic­i­pants per­formed sig­nifi­cantly bet­ter than did con­trols (Md­iff ϭ 3.22, p Ͻ .001). Posttest and fol­low-up per­for­mances were sig­nifi­cantly bet­ter than on pretest (Md­iff ϭ 3.40, p Ͻ .001, and Md­iff ϭ 2.75, p Ͻ .001, re­spec­tive­ly). No sig­nifi­cant differ­ence was found be­tween posttest and fol­low-up. Post hoc com­par­isons re­vealed that only the trained group showed sig­nifi­cant im­prove­ment in per­for­mance be­tween pretest and both posttest ( p Ͻ .001) and fol­low-up ( p Ͻ .001), al­though posttest per­for­mance was not differ­ent from that of fol­low-up. By con­trast, no sig­nifi­cant differ­ence was found for the con­trol group. The trained group per­formed bet­ter at both posttest and fol­low-up than did the con­trol group ( p Ͻ .001).

    …First, the par­tic­i­pants in­volved in our study were young-old (mean age of 69 years), whereas in Buschkuehl et al.’s (2008) study as well as that of Li et al. (2008), they were old-old adults (mean age of 80.1 and 74.5 years, re­spec­tive­ly). In the con­text of episodic mem­o­ry, the meta-analy­sis by Ver­haeghen et al. (1992) has pointed out that the ben­e­fit of in­ter­ven­tions is neg­a­tively re­lated to par­tic­i­pant age (see also Singer, Lin­den­berg­er, & Bal­tes, 2003). It has been shown that cog­ni­tive plas­tic­ity is re­duced over the adult life span (Jones et al., 2006), with young-old ex­hibit­ing larger train­ing-re­lated gains than old-old (Singer et al., 2003). The im­por­tance of par­tic­i­pant age is ev­i­dent from con­sid­er­ing the re­sults of train­ing fo­cused on ex­ec­u­tive con­trol tasks-for ex­am­ple, task-switch­ing (Buch­ler, Hoy­er, & Cerel­la, 2008; Kar­bach & Kray, 2009; Kramer, Hahn, & Go­pher, 1999), dual tasks (B­herer et al., 2005, 2008), or gen­eral ex­ec­u­tive func­tions (Basak et al., 2008)-for which trans­fer effects emerged with a sam­ple com­pris­ing young-old (age range be­tween 60 and 75 years, mean age be­tween 65 and 71 years; Basak et al., 2008; Bherer et al., 2005, 2008; Kar­bach & Kray, 2009; Kramer et al., 1995). The ques­tion of whether trans­fer effects of WM train­ing can also be de­ter­mined by par­tic­i­pant age range is of in­ter­est and should be ad­dressed in fur­ther re­search.

    Sec­ond, as is men­tioned at the be­gin­ning of this sec­tion, the task and the pro­ce­dure used to train par­tic­i­pants can be con­sid­ered an im­por­tant source of differ­ence. For ex­am­ple, Buschkuehl et al. (2008) re­ported that trained par­tic­i­pants claimed to have gen­er­ated task-spe­cific strate­gies in one of the vari­ants of the WM task in which they were trained, lead­ing to greater train­ing gains (62%) with re­spect to the other two vari­ants (44% and 15%, re­spec­tive­ly). The diffi­culty of trans­fer­ring the gains ob­tained in a spe­cific task to other tasks sug­gests that the WM train­ing by Buschkuehl et al. did not fos­ter an in­crease in flex­i­bil­i­ty, but sim­ply the ten­dency to find a strat­egy to re­call as many items as pos­si­ble but in the con­text of each WM task. In the case of Li et al. (2008), the mod­est trans­fer effects to the WM task can be ex­plained by re­flect­ing on the na­ture of the trained task: n-back task, which in­volves the ma­nip­u­la­tion and main­te­nance of in­for­ma­tion as well as up­dat­ing of tem­po­ral or­der and con­tex­tual in­for­ma­tion and bind­ing processes be­tween stim­uli and cer­tain rep­re­sen­ta­tion (Ober­auer, 2005). Al­though the n-back shares com­mon pro­cess­ing mech­a­nisms with com­plex span tasks, the un­der­ly­ing mech­a­nisms of the n-back are not com­pletely un­der­stood (Schmiedek, Hilde­brandt, Löv­den, Wil­helm, & Lin­den­berg­er, 2009). More­over, the few stud­ies that used it with other WM tasks- com­plex span tasks- have shown vari­able cor­re­la­tions (from very low or nul­l-Kane, Con­way, Mi­u­ra, & Colflesh, 2007; Roberts & Gib­son, 2002-to large-Schmiedek et al., 2009; Shamosh et al., 2008).

  92. Specifi­cal­ly, per­for­mance on ; see (S­lagter 2007); cf.“Study Sug­gests Med­i­ta­tion Can Help Train At­ten­tion” (New York Times).↩︎

  93. “Can Med­i­ta­tion Curb Heart At­tacks?” (New York Times)↩︎

  94. Psy­cho­nomic Bul­letin & Re­view 2008 Aug;15(4):763-71. “Train­ing gen­er­al­ized spa­tial skills.” Wright R, Thomp­son WL, Ga­nis G, New­combe NS, Koss­lyn SM.

    …The present study in­ves­ti­gated whether in­ten­sive long-term prac­tice leads to change that tran­scends stim­u­lus and task pa­ra­me­ters. Thir­ty-one par­tic­i­pants (14 male, 17 fe­male) were tested on three cog­ni­tive tasks: a com­put­er­ized ver­sion of the Shep­ard-Met­zler (1971) men­tal ro­ta­tion task (MRT), a men­tal pa­per-fold­ing task (MPFT), and a ver­bal analo­gies task (VAT). Each in­di­vid­ual then par­tic­i­pated in daily prac­tice ses­sions with the MRT or the MPFT over 21 days. Post­prac­tice com­par­isons re­vealed trans­fer of prac­tice gains to novel stim­uli for the prac­ticed task, as well as trans­fer to the oth­er, non­prac­ticed spa­tial task. Thus, prac­tice effects were process based, not in­stance based. Im­prove­ment in the non­prac­ticed spa­tial task was greater than that in the VAT; thus, im­prove­ment was not merely due to greater ease with com­put­er­ized test­ing.

  95. [A pre­vi­ous ver­sion of this foot­note dis­cussed method­olog­i­cal prob­lems & trans­ferra­bil­ity of an­i­mal re­search; this has been moved else­where.] “First Di­rect Ev­i­dence of Neu­ro­plas­tic Changes Fol­low­ing Brain­wave Train­ing” & “Mind­ful­ness Med­i­ta­tion Train­ing Changes Brain Struc­ture in Eight Weeks”, Sci­ence Daily↩︎