Teaching and Human Memory, Part 2

Brian Taylor for The Chronicle

December 14, 2011

Last month I wrote an overview about recent research on the workings of human memory, with the help of a journal article on the subject published in College Teaching.

Michelle Miller, a cognitive psychologist at Northern Arizona University, wrote the article and agreed to answer some questions I had, for this two-part series, on how memory research should affect our teaching.

As was pointed out in many of the comments to my first column, and as Miller herself acknowledged more than once in our written exchanges, the cognitive skills involved in memory storage and retrieval form only a part of what we want to help our students obtain or develop. But it's worth noting, even if you see yourself as teaching higher-order thinking skills, that our students—like all of us—typically need a strong foundation of knowledge and information in our memories before we can begin using those advanced skills.

As many commentators also pointed out, the Internet has made the storage and retrieval of information a much easier task for all of us, but that doesn't change the fact that we rely on memory all the time in the practice of our discipline and trades—not to mention in everyday life.

An emergency-room doctor rushing a patient to surgery, a lawyer brought up short by a surprising piece of testimony in a trial, a sales clerk responding to an unexpected question by a customer—in all of those moments, the professional in question has to draw quickly from a memorized store of previous experiences and information. No doubt the ability to apply the information from memory to a new situation, and respond accordingly, represents a different and more complex thinking skill—but people can't get to that more complex skill without access to their medical, legal, or professional knowledge.

Those of us in higher education surely know that from our own experiences. When a student in my survey course on British literature asks me about the Irish potato famine, she should expect me to expound upon the role that the British government played in exacerbating that "natural" disaster, but she should also expect me to know—without my having to stop class and Google it—that the first potato crop failed in 1845, and that crop failures continued for the next half-dozen years.

Likewise, if I expect my students to understand the complex historical and literary relationship between Ireland and England, I want them to know roughly where the famine falls in the historical timeline of that relationship. I might not care whether they know the exact years of the worst crop failures, but I do care that they know how the crisis related to the rise of industrialism and economic theory in the 19th century, how it influenced the writings of Marx and Engels, and how it precipitated an Irish diaspora. To understand those more complex historical issues, students must know the approximate dates of the famine.

So memory matters, even for those of us teaching the most complex cognitive skills we can imagine. Given its importance to our work in higher education, I sought help from Miller, first of all, in thinking about how her research might apply to the design and presentation of college courses.

"The mind isn't a sponge that absorbs whatever disjointed information we happen to pick up through our senses," she said. "Rather, we acquire information from the environment that we (a) understand, and (b) care about. It follows that when we design our courses, we should start by asking ourselves how we will capture and direct students' attention, and then plan how we will frame the information in a meaningful, interpretable way. This is different from the traditional approach of starting with the material to be covered and how we plan to spread it out over the course of the semester."

The traditional approach we use is to present information to students and then ask them to reflect upon it, respond to it, or relate it to their lives. Instead, Miller says, begin with exercises or framing questions that will engage students. Once you have their attention, then cover your material. Their retention of the material—once they have become engaged with the questions that framed it—should improve significantly.

Or, to borrow a wonderfully concise formulation from a recent book review in America Magazine, "Do not offer them answers before the question itself is intriguing."

Hand-in-hand with course design, as any good pedagogical theorist will tell you, comes thinking about course assignments and evaluation. Course planning should always begin with the end of the semester, as you envision what you want your students to be able to do or know by the time they leave your classroom. So I asked Miller whether her research suggested that certain types of assignments or exams were better than others.

Not really, Miller said. When it comes to assessment, "frequency is more important than format."

That easy-to-remember principle stems from what researchers have dubbed the "testing effect." Put simply, when you take a test or complete any type of assignment involving memory, you are drawing material from your long-term memory. In doing so, as I explained last month, you are practicing the cognitive skill that proves the greatest challenge for our memories—and in the act of practicing that retrieval skill, you are getting better at it. So it turns out that the process of taking a test, instead of just measuring learning, actually improves learning. The more testing, the more learning.

The "testing effect," Miller said, "turns out to be quite robust across different formats, types of questions, and so forth."

All of which suggests that we should be testing, quizzing, and assigning homework to our students as frequently as possible—or perhaps as frequently as we can handle the challenge of grading all of that work. A course with a dozen low-stakes exams or quizzes, and plenty of homework, will do a much better job of promoting retention of course material than a class with only two or three high-stakes exams.

The testing effect has an equally important implication for how our students should study, Miller said. I frequently see students sitting in the study lounge of the honors program I direct, poring over their notes and textbooks in preparation for exams. The next time I see a student doing that kind of review, I should share with them Miller's advice to her own students: "You do your best studying with the book closed."

"Reciting and self-testing," Miller elaborates, are study methods that "provide a great return on investment." Students who close their books and test their ability to recall information and put it to use in self-administered learning challenges are giving themselves the benefit of the testing effect. Students can improve their habits even further by following the longstanding study advice to avoid cramming: "Breaking study time into shorter sessions promotes retention—a phenomenon called the spacing effect."

To be even more helpful to my students, I should kick them out of the lounge every once in a while and tell them to go study someplace new.

That advice stems from the theory of cues, described in last month's column, which posits that information enters our long-term memory accompanied by a specific set of cues. We are more likely to retrieve information from our long-term memory when we encounter a cue—such as a sensory impression from our surroundings—that was present when we first learned that piece of information. But if I do all of my studying in one specific location, in one long burst or at the same time every day, I am giving myself a very limited set of cues associated with the information I'm trying to remember.

By contrast, Miller said, "varying the time and place of study actually promotes retention because it reduces one's dependency on a specific set of cues."

In other words, if I study an hour in the lounge, an hour in my room, and an hour in the library, I'm building up a much larger set of cues associated with the material I am studying, and I am much more likely to encounter one of those cues in the test situation than I would be if I had limited myself to one study time and location. Miller believes we should share that kind of information with our students to help them improve their study habits. She mentioned a colleague from another discipline who spends a few minutes at the start of class talking about those principles.

But the final and best piece of advice Miller offered was aimed at faculty members who want to incorporate recent research on human memory into their teaching. Her comments on that front will come as no surprise to those who have spent any time with the literature on teaching in higher education: If you want students to better retain course material, she said, use teaching strategies that "require students to respond, and respond frequently. That is, hands down, the most important application of this research."

Classroom-response activities, she said, should come in forms that will prepare students for the assignments and exams. Students who have to produce essays should be writing in class; students who have to take multiple-choice exams should be responding to questions with clickers; and students who will be tackling equations should be solving problem sets. Have students practice in class, as often as possible, the retrieval activities that you will be asking of them in their graded work.

In these two columns, I have only been able to scratch the surface of this research and its implications for our classrooms. If you are interested in learning more, and doing some additional reading on your own, you can visit my Web site,, for a short list of resources recommended by Michelle Miller.

James M. Lang is an associate professor of English at Assumption College and author of "On Course: A Week-by-Week Guide to Your First Semester of College Teaching" (Harvard University Press, 2008). He writes about teaching in higher education, and his Web site is He welcomes reader mail directed to his attention at