Test-Enhanced Learning

October 11, 2018

Article: Test-Enhanced Learning

Author: Henry L. Roediger III and Jeffrey D. Karpicke

Article Reading Time: 20 minutes




This article describes two experiments undertaken by the authors of the paper which look at the benefits of testing as a learning tool. The idea of the testing effect (or retrieval practice) is well established and the authors wish to "investigate the testing effect under educationally relevant conditions".


Both experiments compare restudying (in this case rereading a text) with testing (in this case a free recall, involving students writing down all they can remember). This is then measured instantly and after a delay. The structures of the experiments are slightly different, but the guiding principles are the same.


In the first experiment it is found that students that studied and then tested themselves performed better after a delay than those that studied twice. This is taken to mean that they have learned the information better (as learning is defined as a change in long term memory).




In experiment 2, a few more things were discovered:

  • Multiple testing was better than a single test for long term retention, even though they had a much reduced exposure to the original text.

  • Testing increased interest in the passage, whereas those who reread it many times found it more boring.

  • Those who only studied were over confident in comparison with those who were tested, possibly due to an idea of fluency of the material.


    My takeaways

I have been looking at the ideas of retrieval practice for a while now, but it was interesting to see some research on it being used in a classroom setting. In particular, a comparison of rereading a text vs testing on the text that shows that immediate performance is better from rereading, but that long-term retention is better in the testing category.


I have been doing weekly quizzes with my S3 class this year, and, anecdotally, these have been helping students to recall ideas and strengthen their memories. Recently I have started doing this with my 5B class, and their reaction has been positive, actually asking for the quizzes.


The idea of free recall is an interesting concept, that would be so easy to implement in class. Just give out a sheet of blank paper, or ask students to turn to a blank page in their book, and get them to write down everything they can remember on a given topic.


And finally, as the authors also mention, both of these experiments are based on testing without feedback. They are not even told if what they recalled was correct or not, and they still remembered more than those who restudied. So imagine how powerful testing could be if we were to then feedback to students and give them opportunities to correct their work.


Please reload





Please reload

Please reload