(From left to right) Seniors Mary Blen­dermann, Giannina Imperial, and Colby Clark pre­sented their research findings on dif­ferent exam question formats at a meeting of the Psy­cho­nomic Society.
Jeri Little | Courtesy

College stu­dents may not enjoy taking tests, but according to a study three Hillsdale College stu­dents pre­sented last weekend, test-taking may be what helps them learn.

On Friday, seniors Mary Blen­dermann, Colby Clark, and Giannina Imperial pre­sented their research in Van­couver, British Columbia, at the 58th Annual Meeting of the Psy­cho­nomic Society, the top inter­na­tional con­ference for cog­nitive psy­chology.

Over the summer, the stu­dents con­ducted research with Jeri Little, who taught at Hillsdale for three years and began this fall as an assistant pro­fessor of psy­chology at Cal­i­fornia State Uni­versity.

They com­pared mul­tiple-choice and matching ques­tions to see which pro­duced a better result. To their knowledge, no one else had studied how matching tests pro­moted learning retention before.

“Our results showed that matching ques­tions are an effi­cient way to promote test-induced learning, and such learning gains may be greater than mul­tiple-choice or cued-recall ques­tions,” Little said.

Their results showed that each type of question aids the learning process, but matching ques­tions can do a better job than mul­tiple-choice ques­tions, because the test-taker must analyze each option, rather than just choosing the one correct answer and moving on.

Blen­dermann and Little also dis­covered that “none of the above” answers helped the learning process only when the answer was an incorrect option.

“The biggest takeaway from research on test-taking is that it’s a good learning tool,” Clark said. “If you had the choice between just re-reading text or quizzing yourself with actual mul­tiple choice ques­tions, always do the test because even if it seems more dif­ficult and it’s more labo­rious, you’re going to learn the material in much more depth, and it will ulti­mately be more effi­cient.”

To inves­tigate the types of test ques­tions, the stu­dents created a research goal, com­bined mul­tiple-choice ques­tions from pre­vious studies with their new ques­tions, and adapted all the ques­tions into matching format.

They created an online survey for par­tic­i­pants, con­tributed to coding and data analysis, and wrote the results in a research paper.

Little, a fellow in the Psy­cho­nomic Society, has attended its con­fer­ences for more than 10 years and invited the stu­dents to come with her this year to present their research.

On Friday evening, the stu­dents took turns over a couple of hours explaining a poster about their study to con­ference attendees, mostly graduate or post­doc­toral researchers. The stu­dents heard encour­agement and sug­ges­tions for further study as feedback.

“People really seemed to care about it,” Clark said. “They had really intel­ligent feedback, and that’s just because of the nature of the con­ference. It’s a very spe­cialized field, so everyone had a good idea of what we were doing and could appre­ciate it.”

For the rest of the weekend, the stu­dents browsed other research posters and lis­tened to lec­tures.

“To anyone involved in research, whether in social or natural sci­ences, I highly rec­ommend attending a research con­ference and pre­senting your work to other like-minded researchers,” Imperial said. “Not only is it a great résumé builder and a great way to get your name and research out there in the public for other sci­en­tists to see, but it’s a fan­tastic way to learn more about the dif­ferent areas of study in your field, and it is just so much fun to go to con­fer­ences like this.”

Clark said the findings of their own study haven’t made him more excited about taking tests in his classes, but it has taught him about their value.

“It doesn’t mean it’s not a pain, but that’s indicative of the fact that it’s actually making you think through the process — and you really know if you know it,” Clark said.