Repository logo
 

Testing effects for self-generated versus experimenter-generated questions

Date

2020

Authors

Myers, Sarah J., author
Rhodes, Matthew, advisor
Cleary, Anne, committee member
Folkestad, James, committee member

Journal Title

Journal ISSN

Volume Title

Abstract

Those familiar with the testing effect (i.e., the finding that practicing retrieval improves memory) frequently suggest that students test themselves while studying for their classes. However, it is unclear whether students benefit from testing if they are not provided with testing materials. Few studies have examined whether generating one's own test questions improves performance, and none of these studies have given participants a full retrieval opportunity. The proposed experiments bridged this gap between testing effect and question generation research by allowing participants to generate questions and attempt to answer those questions after a delay. In Experiment 1, participants generated test questions over passages and either answered their questions as they created them or after a delay. In Experiment 2, participants either generated questions and answered them after a delay (i.e., self-testing), answered experimenter- generated questions, or restudied the material. Both experiments found no benefits of self-testing compared to the other conditions. In fact, those who self-tested tended to have worse final test performance than the other conditions. Analyses of the questions that participants created suggest that students may benefit more from self-testing when they generate more questions and those questions target material that is on the final test. Although further research is needed to confirm these conclusions (e.g., longer delays between study activities and final test), the current study suggests that testing may not always benefit learning if students must create their own questions.

Description

Rights Access

Subject

Citation

Associated Publications