When I was getting certified in elementary education during college, I distinctly remember being terrified about having to develop and teach a two-week-long history unit on Christopher Columbus in a 5th grade classroom. I hated history. In my experience, history was simply the memorization of names, dates, and facts that I didn't much care for. As a student teacher, I literally cried in my undergraduate methods course. How could I teach a subject I despised?
In retrospect, I don't think I taught the unit very well. I remember presenting the original log from Christopher Columbus's journey, and having students compare it to the edited version in their social studies textbook. I gave a few lectures on the exploration and discovery of the "new world," I created worksheets (we all love those, right?), brought in a guest speaker who portrayed Queen Isabella in costume, and ended the unit with recaps and reviews of what students learned.
More than 15 years later, I wonder: Do my former students remember anything from that unit? Over the years, my career as a cognitive scientist has revealed the most critical teaching strategy I could have used in that 5th grade classroom: retrieval practice. When students practice their knowledge and bring information to mind, their long-term learning increases. With low-stakes mini-quizzes and frequent opportunities to retrieve, my students would have retained much more about Christopher Columbus.
Retrieval Practice: A Powerful, Evidenced-Based Tool
Since that history unit, I've taught thousands of K–12, college, and graduate students. I now teach psychological science at Berklee College of Music in Boston, in addition to conducting research and translating cognitive science for educators. The driving force behind all of my lessons is to ensure that I foster students' long-term learning. Retrieval practice is one of my essential strategies.
The cognitive science research my colleagues and I have conducted, as well as research going back more than 100 years, reveals retrieval practice to be one of the most powerful strategies to boost students' long-term learning. In our book, Powerful Teaching: Unleash the Science of Learning (Jossey-Bass, 2019), middle school teacher Patrice Bain and I share the research behind retrieval practice and how we use it in our classrooms.
In addition to retrieval practice, our book highlights three other "power tools" that increase student learning: spacing (learning over time rather than cramming content in all at once); interleaving ("mixing up" presentation of related information to foster the ability to differentiate between discrete concepts or content); and feedback-driven metacognition (helping students reflect on what they know and don't know). I'll concentrate here on retrieval, showing how any teacher might draw on its effectiveness.
Retrieval practice is the intuitive idea of encouraging students to "pull out" information they have learned from memory. For example, during the unit just described, I might've asked students, "What do you remember about Queen Isabella?" or "Do you remember the name of the Spanish king at that time?" Instead of starting class by reviewing information for my 5th graders, I could have asked them to retrieve what they learned the previous day.
Think of retrieval practice as mentally time traveling backwards and bringing information to mind. Research demonstrates that engaging students in this action frequently improves learning for a range of ages (from early childhood to older adults) and for a range of subject areas (Agarwal, Nunes, & Blunt, 2020). While all good teachers engage students in retrieval (such as through review activities or discussions), we sometimes rely more than we should on getting information "into" students heads rather than having them pull information out.
We can—and should—change that emphasis. Retrieval practice is not only supported by evidence, it's also a no-prep, no-grading strategy any teacher can use to substantially impact student learning. For example, in my classroom, I use a "brain dump." I simply ask students to write down everything they can remember about a specific class topic. Brain dumps (called free recall in the scientific literature) engage students in retrieval practice and improve the organization of the information they're learning (Zaromb & Roediger, 2010). In a simplified version, I ask students to recall "two things" about a topic. Any teacher could use brain dumps or two things as entry or exit tickets or during a pause in the middle of the lesson.
"Desirable Difficulty"
Why does retrieval practice boost learning? Because it challenges the learner. It presents what scientists call a "desirable difficulty," a term coined in 1994 by Robert Bjork. Retrieval practice is demanding—and that's a good thing for learning! Decades of research have shown that easy strategies like rereading and cramming lead to short-term learning, whereas more effortful strategies lead to long-term learning.
Notably, retrieval practice improves students' higher-order learning of complex concepts, not just memorization of facts (Agarwal, 2019). In one study I conducted (that included research in Patrice Bain's classroom at Columbia Middle School in Illinois), middle school students and college students engaged in retrieval practice through quizzes that either presented fact-based questions, higher-order questions, or a mix of question types, to examine the optimal type of retrieval practice for enhancing higher-order learning. In laboratory and K–12 settings, such retrieval practice consistently increased delayed test performance, compared with rereading the content or not doing either rereading or quizzes. Quizzes using higher-order questions and "mixed" quizzes improved test performance on higher-order questions, but fact-based quizzes did not.
So contrary to popular intuition about higher-order learning and Bloom's taxonomy, building a foundation of knowledge via fact-based retrieval practice may be less potent than engaging in higher-order retrieval practice, a key finding for future research and classroom application. The takeaway: It's best to provide a mix of fact-based questions and higher-order questions that students must retrieve answers for.
Retrieval practice also enhances transfer of learning. In a 2018 meta-analysis, researchers Pan and Rickard found that retrieval practice produced more robust transfer than many other research-based strategies, likely due to the desirable difficulty of retrieving information. When students retrieve, they learn information in a deeper manner than when they simply review it, and with this increased effort, students don't simply recall information, they have to use it in new and different ways.
How Cognitive Psychologists Conduct Research
Before I describe more about my research on retrieval strategies with Patrice, let me explain some things about research in this field. Learning is an incredibly complex behavior. The phrase "science of learning" is an umbrella term that spans many research fields, including psychology, computer science, and neuroscience. My research sits in the field of cognitive science, specifically cognitive psychology. In cognitive psychology, we typically examine mental operations, or "behind the scenes" behaviors occurring inside our heads, like perceiving, remembering, thinking, and decision making.
While research on learning can be based on observations, surveys, or correlations, the research my colleagues and I conduct includes true experiments; we determine causation by comparing whether one approach or strategy boosts learning more than another under controlled conditions.
Cognitive psychologists conduct experiments in laboratories on everyday learning and memory activities in typical human populations, examining everything from how students learn basic facts and vocabulary to how students apply their knowledge using complex higher-order materials.
Consider a landmark study by Henry L. ("Roddy") Roediger III and Jeffrey Karpicke (2006). College students at Washington University in St. Louis read brief passages about topics like sea otters and the sun. After reading the passages once, some students were instructed to reread them over and over, whereas other students retrieved information by writing down everything they could remember. Roddy and Jeff measured students' performance on an exam five minutes later and also one week later. After five minutes, student performance in the rereading condition was much higher (83 percent) compared with the retrieval practice condition (71 percent). But after one week, this pattern completely switched: Student performance in the retrieval condition was now higher (61 percent) than performance in the rereading condition (40 percent). In the span of one week, when students reread passages, they forgot more than half of what they once knew, but when students engaged in a simple, quick retrieval method, they forgot a little bit (which is expected), but forgot much less over the following week.
Investigating Retrieval Methods in Classrooms
Recently, researchers around the world have started conducting research on retrieval practice in real, functioning classrooms, with our team's work in Patrice Bain's classroom being one of the first. In 2006, Roddy Roediger, Mark McDaniel, Kathleen McDermott, and I (at the time all at Washington University in St. Louis) embarked on one of the first long-term studies exploring retrieval-based learning in a classroom setting.
We began with an experiment with Patrice Bain's 6th grade social studies class. While studying a chapter on ancient Egypt, the students were given three brief low-stakes quizzes. All information in the chapter was covered during Patrice's lesson, but half of the chapter material was presented on the quizzes and half the material wasn't quizzed. In other words, students engaged in retrieval practice for only part of the chapter material.
Our team found that student performance was significantly greater for the quizzed (in other words, retrieved) material than for non-quizzed material (94 percent versus 81 percent) on an end-of-chapter test (Roediger et al., 2011). Even at the end of the semester (one to two months after Patrice's lessons), student performance was still significantly greater for recalling quizzed material than non-quizzed material—by an entire letter grade (79 percent correct compared with 67 percent correct). So even with the various interruptions that happened in her classroom—fire alarms, snow days, and day-to-day chaos—using retrieval practices raised grades on assessment of this learning, the equivalent of going from a B- to an A.
Other Strategies for the Long-Term
Let's consider briefly one of the three other "power tool" strategies—spacing. Like retrieval practice, all these strategies—spacing, interleaving, and feedback-driven metacognition—are effective, and none requires any additional prep or grading. In fact, these power tools should remain low-stakes and ungraded when possible. Using them this way emphasizes that they are learning strategies, not assessment strategies. Each of these tools builds on retrieval practice. (Visit retrievalpractice.org for more information about each of these strategies.)
With spacing, which I use in my own classroom, I simply ask students to retrieve information they learned months ago, not just recently. It's similar to trying to use a foreign language that you learned quite a while ago in school when you're traveling in a country where that language is spoken. In such situations, students' memory for that information is strengthened. In other words, this type of forgetting is good for learning.
For example, my students complete weekly written "mini-quizzes" that are worth two percent of students' grades. In the sixth week of the semester, I might ask my students to retrieve information they learned in the second week. Initially, students find this jarring ("You want me to remember what?!"). But quickly, my students realize that spacing—or delayed retrieval—improves their learning. When we discuss study strategies in my class, my students generally say that cramming works in the short term—they get good grades. But when I ask them what happens after the exam, my students respond nearly in unison that they forget everything. The desirable difficulty of spacing out learning improves long-term learning.
Flash Forward
My teaching approach has evolved since I was a nervous student teacher in a 5th grade classroom. In reflecting on whether the students I teach now at Berklee College will, 10 years from today, remember anything from my classes, I'm confident that they will remember much more than the 5th graders I taught back when I didn't use strategies like retrieval practice. In fact, toward the end of each semester, I ask my Berklee students to respond to a "Flash Forward" prompt (which combines retrieval practice and spacing): If, 10 years from now, you could remember one thing about what you learned in this class, what would it be? This question makes students take ownership; they have to retrieve their previously learned information and evaluate which parts of it represent the most relevant concepts they want to remember.
As a scientist and a teacher, I practice what I preach. I take advantage of all opportunities to infuse retrieval practice throughout my lessons, whether with brain dumps, mini-quizzes, or other techniques. My 5th graders of years ago may or may not remember much about Christopher Columbus or Queen Isabella. But as I retrieve and reflect, I'll never forget how much I learned from them.
Copyright © 2020 Pooja K. Agarwal
