by Atena Bishka, MBA, MSc
Although widely used, multiple-choice tests are commonly frowned upon. They are held in contempt by education pundits, looked down on by subject matter experts, despised by test-takers, and even seen as dubious by training decision makers. Why do multiple-choice tests seem to have taken a permanent seat on the accusation bench? The overall claim is that multiple-choice testing does not measure, and cannot be a good measuring scale of, learning.
At the heart of the criticism is the claim that multiple-choice tests simply focus on recognition processes. Different from cued-recall tests (or short-answer tests), they are incapable of engaging the productive retrieval processes (or later recall of knowledge). Essentially this means that because the multiple-choice tests display the correct answer among alternatives, all a learner has to do is recognize the correct answer. Thus, there is no need for learners to exert mental efforts leading them to sift through information and knowledge that is necessary to figure out and articulate the correct answer by retrieving pertinent information from memory.
Guilty or Not–What Does Research Say?
Does recent research support such charges? Are multiple-choice tests inferior to cued-recall tests or short-answer tests? Is it true that multiple-choice tests do not facilitate later recall of knowledge–a key cognitive capability for learning retention? Do they indeed fail to trigger productive retrieval processes?
An experimental study published at Psychological Science (Little, Bjork, Bjork, & Angello, 2012) does not support such criticism. This study maintains that if constructed properly, multiple-choice tests can and do engage effectively the productive retrieval process.
Experimental Method and Results
To find out whether the multiple-choice tests can engage both recognition processes and productive retrieval processes researchers conducted two experiments.
The goal of the first experiment was to assess the effects of an initial test cued-recall test and an initial multiple-choice test on the participants’ performance on a final test. The final test contained questions identical to those used in the initial tests plus other content-related questions. What were the results?
- Results of initial tests indicated that (a) the number of correct answers for the multiple-choice test was larger than those given for the cued-recall test, and (b) the initial multiple-choice test improved later recall of the tested information more than the initial cued-recall test did.
- Results on final tests indicated that when participants took the initial cued-recall test, final-test performance on related items was worse than that on their corresponding control items. But, when participants took an initial multiple-choice test, final-test performance on related items were slightly higher than that on their corresponding control items.
The second experiment, a replication of the first, included an added condition: Half of the participants received feedback indicating the correct answer to each question on the initial test. The goal was to assess how the feedback on an initial multiple-choice test affects the recall of previously tested information and retention of other content-related information. The results indicated that:
- When feedback was provided, final-test recall of previously tested information improved. However, the improvement was nearly the same regardless of whether the initial test was a multiple-choice test or a cued-recall test.
- Remarkably, the advantage of multiple-choice testing for the recall of information related to initially incorrect alternatives was sustained, even when participants were given feedback during the initial test.
What Do the Experimental Results Mean?
The results listed above indicate that the multiple-choice tests are not inferior to cued-recall tests. Not only do they promote learning of previously tested information, but they can also facilitate learning of content related to incorrect alternatives (by forcing learners to think why the alternatives are incorrect answers). The latter part is especially important as it shows that multiple-choice testing has a superior advantage compared to cued-recall testing, which forces learners to figure out the correct answer only.
This study is an attempt to reassess the stigma of inferiority attached to multiple-choice tests. The experiments’ results suggest that multiple-choice tests are unfairly judged
and do not deserve the charges. At least when used as practice tests they can be particularly useful, as they help foster learning of both previously tested and related content.
Words of Caution
According to this study, the benefits above may not materialize unless one important condition is satisfactorily met: The multiple-choice tests have to be properly constructed, ensuring the alternatives are plausible or competitive. Only when alternatives are plausible and competitive (not trivial, obvious, and easy to guess), can the tests trigger the retrieval processes that will force learners to recall why the alternatives are incorrect, thus consequently fostering test-induced learning and deterring test-induced forgetting. When tests are not competitive and properly constructed, the opposite effect might occur: The later recall of information pertaining to those alternatives is not enhanced.
Last but not least, the poorly constructed multiple-choice tests do indeed deserve the “bad tests” label. They are easy to create and do not take much time to develop. Well-constructed multiple-choice testing, however, is time consuming, requiring hard work. The reward is that multiple-choice tests can enhance learning and can engage effectively the productive retrieval process, while being a good measure of learning acquisition.
Little, J. L., Bjork, E. L., Bjork, R. A., & Angello, G. (2012). Multiple-choice tests exonerated, at least of some charges: Fostering test-induced learning and avoiding test-induced forgetting. Psychological Science, 23(11), 1337-1344.
About the Author
Atena Bishka, MBA, MSc, is a consultant, speaker, and author who specializes in learning design and development, assessment of learning initiatives, and evaluation of program impact on business goals and work performance. She supports stakeholders’ decision making by providing instructive and evidenced-based information shedding light on program value and contribution to the organizational, business, and performance results.
A recipient of the 2010 Silver Canadian Award for Training Excellence by the Canadian Society for Training & Development (CSTD), she has presented at a variety of conferences, including CSTD, ISPI, ASTD, and the Conference Board of Canada. She has published articles in the Journal of Education for Business, Performance Improvement, and the Canadian Learning Journal.