Most Active Stories
13.7: Cosmos And Culture
Solving The Conundrum Of Multiple Choice Tests
Originally published on Mon December 9, 2013 3:53 pm
Multiple choice tests are:
A. Only effective for assessing superficial, rote memorization
B. Only effective for assessing deep, conceptual understanding
C. Best at promoting short-term retention of material (e.g., for an upcoming exam)
D. A good way to ensure long-term retention of material
According to an article just published in Current Directions in Psychological Science, the answer — perhaps surprisingly — can sometimes be choice D. But it depends on how multiple choice questions are designed and deployed.
Here's a hint: they shouldn't just appear on the final exam.
Researchers have known for a long time that repeatedly testing yourself on new material — by, for example, answering multiple choice or free-response questions — is more effective than repeatedly reviewing that material when it comes to remembering it one week later. They've also known that it's better to spread out studying over a long period of time rather than cramming right before an exam. This sounds like a recipe for many short tests throughout a course rather than the one or two midterms and final exams with which we're all familiar.
But how does this play out in actual classrooms to improve student learning? Are there benefits to incorporating multiple choice testing throughout the semester, perhaps with in-class or online quizzes or using newer "clicker" systems during lecture? And if there are benefits to repeated multiple choice testing, are they limited to memory for the specific content of the questions that were presented throughout the course, or do they generalize to related content?
The new paper, authored by Arnold L. Glass and Neha Sinha, reviews nine recent experimental studies that explore these questions in real classes, and the results are encouraging.
Students who received multiple choice questions throughout a course performed better than those who didn't when it came to final exams. And importantly, the benefit generalized to multiple choice questions that were related but not identical to those previously presented, and even to related short-answer questions. In some cases, the benefits of repeated testing yielded an advantage that amounted to a letter grade or more.
These findings have practical importance because they support a pedagogical strategy that's so easy to adopt. Repeated multiple choice testing is fast and inexpensive to implement and grade, even at large scales. It's also something that students (of any age and subject matter) can incorporate into their own study habits.
The findings are also theoretically important insofar as they shed light on the human mind and our remarkable ability to learn. Most likely, several factors converge to make repeated testing an effective strategy. For one thing, testing can be more active and engaging than passively reviewing notes or listening to lecture. It also forces you to recognize what you haven't yet mastered — a kind of awareness that simply reading or listening won't necessarily foster.
But my favorite hypothesis is that repeatedly encountering the need for particular information sends a powerful signal to your memory system: it indicates that the information is likely to be important across contexts and time periods. It's like telling your central library that a book should be stored in the main stacks, not in the auxiliary library — somewhere handy for quick access, because you might just need it again soon.
Sadly, our mental librarians aren't always perfect. So knowing how to design courses — and how to structure our own study habits — to make their jobs a little smoother is all the more important.
You can keep up with more of what Tania Lombrozo is thinking on Twitter: @TaniaLombrozo