I can’t believe I’m typing these words, but I’m here today to convince you to give your students more tests.
I write this as someone who has frequently included just a midterm and final, and who has only planned those exams because of department requirements. When I teach literature or writing, it’s usually my opinion that a package of essays, assignments, presentations, and a participation grade is a more appropriate means of assessment than an exam, which can seem unimaginative, a blunt instrument. I find that students dread taking exams; they think that tests don’t accurately measure what we do in a college classroom. I suspect that many other teachers and students feel the same way, even those in disciplines less “subjective” than mine.
But what if, instead of giving students a handful of big exams—instead of struggling to shoehorn in a term’s worth of material and asking students to pull all-nighters to cram—you give them frequent, low-stakes tests throughout the semester?
There are a couple of immediately appealing features of such a scheme. For one, regular tests ensure that students are studying regularly throughout the term, not waiting until just before the big exam to relearn everything at once. Secondly, frequency breeds comfort: If tests are a regular, perhaps weekly, part of your class’s schedule, students’ performance anxiety should lessen considerably. No longer will there be a sense that everything (or a significant portion of everything) is riding on how they do on one or two tests.
Even beyond these features, there are compelling reasons to believe that frequent testing can have a significant effect on student learning. Writing recently for The Atlantic, Jessica Lahey details research by Henry L. Roediger III, a professor of psychology at Washington University, in St. Louis. Roediger looked into the so-called “testing effect”—the strange fact that the act of taking a test seems to help students retain class material. Lahey situates Roediger’s work within the world of K-12 education, in which the subject of testing is currently a hotly debated topic. But there’s no reason why Roediger’s surprising conclusions can’t be applied within the college classroom as well.
In a 2006 paper written with his then-graduate student, Jeffrey D. Karpicke, Roediger reviews nearly a century’s worth of studies into the testing effect and concludes that it is very real indeed. The paper suggests that the mental processes involved in taking a test are powerful ones, able to help students retain memories of class material better even than studying that material. There is something in the act of taking a test, of being forced to try to recall what we know of a subject, that works very effectively on our brains.
Remarkably, Roediger shows that the testing effect is independent of other seemingly important factors: It persists even when students do poorly on the test, and works even if teachers don’t give feedback on the missed material.
Roediger and Karpicke conducted their own experiments with college students, asking them to read texts that covered general scientific topics. After all of the students read their passage, half of the students were told to restudy that passage, while the other half were given a test on it. The results showed that studying the material provided a modest short-term benefit—the students who studied knew the passage better in the minutes after studying. But those who were tested without studying remembered significantly more about the passage both two days and one week later. Repeated variations of this experiment provided consistent results: Taking a test was a better means to long-term retention than studying.
What should this mean for our classrooms? Roediger’s research has convinced me that we should not see tests merely as instruments of assessment. They can be valuable teaching tools in their own right, helping students to better assimilate all the material we are constantly feeding them.
Near the end of their paper, Roediger and Karpicke mention their colleague Kathleen McDermott, also a psychology professor at Washington University, who has put this principle into practice. She devotes the last 10 minutes of every class to a test on the assigned readings and the lecture material. The students know they must come to class prepared, and the quizzes help students retain important information from each class.
I like the idea, and think I may do something like it next semester. Would anyone like to do my grading for me?