The Testing Effect for Learning Principles and Procedures from Texts

The authors explored whether a testing effect occurs not only for retention of facts but also for application of principles and procedures. For that purpose, 38 high school students either repeatedly studied a text on probability calculations or studied the text, took a test on the content, restudied the text, and finally took the test a second time. Results show that testing not only leads to better retention of facts than restudying, but also to better application of acquired knowledge (i.e., principles and procedures) in high school statistics. In other words, testing seems not only to benefit fact retention, but also positively affects deeper learning.

Testing After Worked Example Study Does Not Enhance Delayed Problem-Solving Performance Compared to Restudy

Four experiments investigated whether the testing effect also applies to the acquisition of problem-solving skills from worked examples. Experiment 1 (n=120) showed no beneficial effects of testing consisting of isomorphic problem solving or example recall on final test performance, which consisted of isomorphic problem solving, compared to continued study of isomorphic examples. Experiment 2 (n=124) showed no beneficial effects of testing consisting of identical problem solving compared to restudying an identical example. Interestingly, participants who took both an immediate and a delayed final test outperformed those taking only a delayed test. This finding suggested that testing might become beneficial for retention but only after a certain level of schema acquisition has taken place through restudying several examples. However, experiment 2 had no control condition restudying examples instead of taking the immediate test. Experiment 3 (n=129) included such a restudy condition, and there was no evidence that testing after studying four examples was more effective for final delayed test performance than restudying, regardless of whether restudied/tested problems were isomorphic or identical. Experiment 4 (n=75) used a similar design as experiment 3 (i.e., testing/restudy after four examples), but with examples on a different topic and with a different participant population. Again, no evidence of a testing effect was found. Thus, across four experiments, with different types of initial tests, different problem-solving domains, and different participant populations, we found no evidence that testing enhanced delayed test performance compared to restudy. These findings suggest that the testing effect might not apply to acquiring problem-solving skills from worked examples

Answering questions after initial study guides attention during restudy

Though the testing-effect can be boosted by including a restudy phase after answering test questions, we do not know precisely why it does so. One possible explanation is being tested here. The present study measured attention allocation during initial reading and rereading with a remote eye tracker to gain information on the cognitive processes during restudy, with and without prior testing. The results show that at the final study moment, students in the study-test-condition attended longer to information pertaining to the initial test questions as compared to students in the study-only condition (i.e., who did not take the test). No differences in attention allocation were found for information only questioned on a posttest 1 week later. In addition, it was found that performance on the initial test questions heavily affected which information students restudy; students in the study-test-condition paid namely more attention to the answers of questions they answered incorrectly during the initial test than to the answers of the questions they answered correctly on the initial test