Long-term Retention of Prose Following Testing

Testing enhances later recall by highschool students of a history text
psychology
by: P. Duchastel, R. Nungester 2012-01-292013-08-24 finished certainty: log importance: 0


This tran­script was sourced from Am­mons Sci­en­tific.

  • P. Duchas­tel and R. Nungester (1981); “Long-term Re­ten­tion of Prose Fol­low­ing Test­ing”. : Vol­ume 49, Is­sue 2, pg470. doi: 10.2466/pr0.1981.49.2.470

Long-term Retention of Prose Following Testing

P. Duchas­tel and R. Nungester

The Amer­i­can Col­lege, Bryn Mawr

Tak­ing a test on con­tent which has just been stud­ied is known to en­hance later re­ten­tion of the ma­te­r­ial stud­ied1. This test­ing effect is gen­er­ally ex­am­ined through a re­ten­tion test ad­min­is­tered a few (typ­i­cally two) weeks after ini­tial learn­ing and test­ing. The present study ex­am­ined this phe­nom­e­non with re­spect to long-term ben­e­fits which might re­sult from test­ing. This was ac­com­plished through the ad­min­is­tra­tion of a re­ten­tion test 5 months fol­low­ing a typ­i­cal study of effects of test­ing.

In the ini­tial study2, high­-school stu­dents stud­ied a brief his­tory text, then took a test on the pas­sage, spent equiv­a­lent time re­view­ing the pas­sage, or went on to an un­re­lated filler task. A re­ten­tion test given 2 weeks later in­di­cated that the test con­di­tion re­sulted in sig­nifi­cantly bet­ter re­ten­tion than ei­ther the re­view or the con­trol con­di­tions.

Five months lat­er, these same 83 stu­dents were ad­min­is­tered a 12-item mul­ti­ple-choice test (five op­tions per item) sim­i­lar to the re­ten­tion test they had com­pleted ear­li­er. Re­ten­tion after 5 months was less over-all, but the data closely par­al­lel the data ob­tained on the 2-week re­ten­tion test. The re­spec­tive means and stan­dard de­vi­a­tions are as fol­lows: Test group: M = 6.0, σ = 2.3; Re­view group: M = 5.1, σ = 1.7; Con­trol group: M = 4.8, σ = 2.1 The analy­sis of vari­ance of these data gave an F2.80 of 2.4 which was, how­ev­er, only sig­nifi­cant at p < .10.

While the analy­sis did not repli­cate the strong sig­nifi­cant differ­ences found after 2 weeks, it is nev­er­the­less of in­ter­est that the or­der of the group means after 5 months ex­actly fol­lows pre­dic­tion and closely par­al­lels the or­der of the group means after 2 weeks. A study with more power (for in­stance, with more sub­jects) would prob­a­bly es­tab­lish clearly sig­nifi­cant group differ­ences. In con­clu­sion, the data in­di­cate the pos­si­bil­ity that tak­ing a test im­me­di­ately after learn­ing may lead to an ad­van­tage in re­ten­tion (although a small one) even at a much later time. Test­ing may thus be a pow­er­ful way of en­hanc­ing long-term re­ten­tion.


  1. Duchastel, P. “Re­ten­tion of prose ma­te­ri­als: the effect of test­ing”. Jour­nal of Ed­u­ca­tional Re­search, 1979, 72, 399-300.↩︎

  2. Nungester, R. & Duchastel, P. “Test­ing ver­sus re­view: effects on re­ten­tion”. Jour­nal of Ed­u­ca­tional Psy­chol­ogy, in press.↩︎