Memory researchers have long known that it’s better for trainees to retrieve an answer than to recognize it.
So if you’re using tests to reinforce learning and not just to assess it, the good old multiple-choice quiz would seem to be a poor choice, since it relies on recognition.
A better choice
Not so fast, says new research. It depends on how you create the multiple-choice test.
In a recent study, researchers showed that if you use the same answer – once as an incorrect alternative and later as the correct one – you will generate the kind of productive retrieval that leads to lasting memories.
For example, if the right answer to question #1 is “green,” you’d make one of the wrong answers “blue.”
Then you’d use “blue” as the right answer to a later question.
Why it works
Here’s why this technique works: It requires the trainee to think about the information twice – once in a context where it’s incorrect and once in a context where it’s correct.
As a result, the learner has to think about why it’s correct in one place and incorrect in another. That effort involves more than simple recall – it forces the brain to go back to the original lesson to get the right answer.
Caveat: The wrong answers need to seem plausible, so that learners can’t just guess. You want them to re-access the learning to come up with the correct response.
In other words, the wrong answer needs to be “almost right.”
Past, present and future
The same technique works across quizzes as well as within them. For example, you can use the right answer from an earlier quiz as the wrong answer on the current one.
Or vice versa: Embed a correct answer from an upcoming quiz as a wrong answer in today’s quiz. While it won’t benefit trainees right away, it plants a seed for future retrieval.
Say, for example, you’re training managers on sexual harassment laws.
Your first quiz question might be:
Q. What’s an example of quid pro quo harassment?
A. A boss implies that a worker will be promoted if she agrees to date him.
B. A boss uses crude sexual language that a female worker finds offensive.
C. A boss denies a worker a promotion because she’s a woman.
D. A boss fires a worker because she complains about harassment.
The correct answer is A. B also seems plausible – but it’s really an example of hostile workplace harassment, not quid pro quo harassment.
A later question could ask:
Q. What’s an example of hostile workplace harassment?
Then you’d offer the same four choices, prompting learners to consider what’s different about the two types of harassment.
Source: Little, J.L., et al. (2012). Multiple-choice tests exonerated, at least of some charges. Psychol Sci 23, pp,. 1337-44.
Subscribe to Rapid Learning Insights
Get the latest research on workplace learning with weekly posts delivered to your inbox