Meeting Four — Test Enhanced Learning (20th September)

Please note that this paper is not science-specific. The Sci­ence Teach­ing Jour­nal Club would like to invite all teach­ers, not just sci­ence teach­ers, to par­tic­i­pate in our fourth meet­ing; please invite your friends and col­leagues.

Roedi­ger, Henry and Jef­frey Karpicke, “Test-Enhanced Learn­ing”, Psy­cho­log­i­cal Sci­ence 17(3): 249–255. doi: 10.1111/j.1467–9280.2006.01693.x (.PDF 121kB)

ABSTRACT: Tak­ing a mem­ory test not only assesses what one knows, but also enhances later reten­tion, a phe­nom­e­non known as the test­ing effect. We stud­ied this effect with edu­ca­tion­ally rel­e­vant mate­ri­als and inves­ti­gated whether test­ing facil­i­tates learn­ing only because tests offer an oppor­tu­nity to restudy mate­r­ial. In two exper­i­ments, stu­dents stud­ied prose pas­sages and took one or three imme­di­ate free-recall tests, with­out feed­back, or restud­ied the mate­r­ial the same num­ber of times as the stu­dents who received tests. Stu­dents then took a final reten­tion test five min­utes, two days, or one week later. When the final test was given after five min­utes, repeated study­ing improved recall rel­a­tive to repeated test­ing. How­ever, on the delayed tests, prior test­ing pro­duced sub­stan­tially greater reten­tion than study­ing, even though repeated study­ing increased stu­dents’ con­fi­dence in their abil­ity to remem­ber the mate­r­ial. Test­ing is a pow­er­ful means of improv­ing learn­ing, not just assess­ing it.

Dis­cus­sion Points:

• The paper sug­gests that the act of being tested, not just sim­ply revis­ing for a test, can improve learn­ing (“the test­ing effect”). Does this result sur­prise you?
• The authors argue that “test­ing in all lev­els of edu­ca­tion is mis­guided” and that “if stu­dents know they will be tested reg­u­larly … they will study more and will space their study­ing through­out the [term] rather than con­cen­trat­ing it just before exams”. Do you agree? Should we be includ­ing more test­ing in our teach­ing plans?
• The results from the paper showed that imme­di­ate test­ing pro­duced bet­ter long-term reten­tion than repeated study­ing. Should we do away with infre­quent end-of-unit tests in favour of more fre­quent end-of-lesson tests?
• Does this paper sug­gest that we should we make more use of the spac­ing effect in our teaching?

Some key phrases from the paper:

“We believe that the neglect of test­ing in all lev­els of edu­ca­tion is mis­guided. To state an obvi­ous point, if stu­dents know they will be tested reg­u­larly (say, once a week, or even every class period), they will study more and will space their study­ing through­out the semes­ter rather than con­cen­trat­ing it just before exams. How­ever, more impor­tant for present pur­poses, test­ing has a pow­er­ful pos­i­tive effect on future reten­tion. If stu­dents are tested on mate­r­ial and suc­cess­fully recall or rec­og­nize it, they will remem­ber it bet­ter in the future than if they had not been tested. This phe­nom­e­non, called the test­ing effect, has been stud­ied spo­rad­i­cally over a long period of time, but is not well known out­side cog­ni­tive psychology.”

“Imme­di­ate test­ing after read­ing a prose pas­sage pro­moted bet­ter long-term reten­tion than repeat­edly study­ing the pas­sage. This out­come occurred even though the tests included no feedback.”

“The pos­i­tive effects of test­ing were dra­matic: Stu­dents in the repeated-testing con­di­tion recalled much more after a week than did stu­dents in the repeated-study con­di­tion, even though stu­dents in the for­mer con­di­tion read the pas­sage [far fewer] times. Test­ing has a pow­er­ful effect on long-term retention.”

“Many study con­di­tions and strate­gies that pro­duce rapid learn­ing and short-term ben­e­fits lead to poor long-term per­for­mance. Our results show that test­ing ver­sus study­ing is another case in point: Test­ing clearly intro­duced a desir­able dif­fi­culty dur­ing learning.”

“This out­come on the imme­di­ate tests in the present exper­i­ments reveals just how pow­er­ful the test­ing effect is: Despite the ben­e­fits of repeated study shortly after learn­ing, repeated test­ing pro­duces strong pos­i­tive effects on a delayed test.”

“We sus­pect that tests will pro­duce strong effects when they occur rel­a­tively soon after learn­ing and per­mit rel­a­tively high lev­els of performance.”