Many teachers share a belief that we should make learning as easy as possible for students. We want students to "get it" and not struggle. Educators often favor techniques that show quick benefits, ones that make learning happen fast.
For example, when students tackle problems in math—say, learning to solve for the volume of a solid—they often practice many examples of the same type. Sure enough, they quickly get the hang of it. They repeatedly apply the same formula, learn quickly, and feel a sense of accomplishment—and so do their teachers.
This process seems great, but there's just one problem: This kind of instructional strategy has repeatedly been shown to create ephemeral learning and leads to rapid forgetting. It's good in the short term but bad in the long term. And that's hard for both teachers and students to realize.
Make It (Desirably) Difficult
Choose Mixed Practice over Blocked Practice
Blocked practice, in which problems of the same type are blocked together, is commonly used in education, industry, the military, and sports because it produces fast learning and students (and teachers) like it. After all, who wouldn't want to learn fast?
Yet the fact is, virtually every study comparing blocked practice to the alternative—mixing up examples of different types of problems after some initial instruction in each—shows that mixed practice is better on measures of long-term learning (Bjork, Dunlosky, & Kornell, 2013; Rohrer & Taylor, 2007). Why? Because certain difficulties encountered in learning—ones that can be overcome through effort—often produce better performance in the long term.
If students have to solve four different types of math problems, with the types given in a random sequence, the task feels hard and clunky to them because they're confused early on. With mixed practice of examples, students have to figure out what kind of problem they're facing, then remember (or look up) the formula for it, and then solve the problem. Blocked practice is much easier because it avoids the first two steps—discovering the type of problem and determining the formula needed. In blocked practice, students just repeatedly run through the same routine.
Students—and their teachers—may struggle in mixed practice of examples during learning because the learning is slower and more effortful. Students may also be more easily discouraged early on by their apparent lack of mastery. However, deliberately making the process more challenging by mixing up the examples can actually help student learning.
Psychologists have coined a term to capture this idea—desirable difficulties. This idea may seem counterintuitive because we usually want to avoid difficulties. Of course, not all types of difficulties are good for long-term learning, but mixed practice offers desirable ones. In mixed practice, the challenging process of discovering how to identify a problem is already built in (Rohrer, 2009). If students slow down, consider what type of problem is represented, and perhaps consult their teacher or textbook, they can often overcome the challenge. The learning that results is deeper, more durable, and more flexible—just what we want our students to take away from our instruction.
Comparing blocked and mixed practice may help explain a common problem that math teachers often face: Students show great competence on problems in class and on homework but then muff the exam. The exam, of course, covers many types of problems, not just one. Students often do poorly on the test because they've been taught with blocked practice, so they have little experience with the more complex skill of distinguishing among problems.
Choose Active Retrieval over Rereading
Suppose students read a passage they'll be tested on a week later. Just after reading it, one group of students is given a test—they're told to recall as much of the passage as they can. The other group is given a more passive strategy—to simply reread the entire passage (Roediger & Karpicke, 2006). Of course, the first group will not be able to recall everything, whereas the rereading group will be exposed to all the material. In terms of reexposure, the students in the recall group are at a disadvantage.
What happens later? If students are tested immediately after the manipulation, the students in the group that reread the material in its entirety recall the passage better than the students who read it only once and reviewed it using a test. This shows that cramming (repeated reading) works, at least for the short term. However, if the same experiment is conducted but the test is delayed a week, those who read the passage only once but tested themselves recalled more than those who read the material twice.
That outcome is a complete reversal from what happened when those students were tested immediately. This second outcome, the one found after a delay, is called the retrieval practice effect or the testing effect: If students review material by actively retrieving it, they'll remember it much better over the long term than if they just passively read the material (Roediger & Karpicke, 2006).
Quizzing students on material (as opposed to having them reread it) thus represents another desirable difficulty. Doing the hard work of reviewing material by actively retrieving it leads to much better retention later on, even when students can't retrieve everything on the first test. Testing leads to other benefits, too. Students find out what they know and what they need to study more (Roediger, Putman, & Smith, 2011).
In research my colleagues and I conducted at a middle and high school in Columbia, Illinois, we found that students who experienced active retrieval practice through quizzing with student response systems had better grades on their exams than students who didn't have the extra quizzing or who had the opportunity to reread the material that other students were quizzed on (McDaniel, Agarwal, Huelser, McDermott, & Roediger, 2011; McDermott, Agarwal, D'Antonio, Roediger, & McDaniel, 2014; Roediger, Agarwal, McDaniel, & McDermott, 2011). Active retrieval seems to help strengthen pathways that will aid later retrieval; it permits students to subsequently use information, which is what we hope they'll do outside the classroom.
Much of education seems to be about getting information into the mind—to store it. That's a necessary condition for us to use the information later, but it's not a sufficient one. We also need to be able to access information when we need it, to pull it out. Retrieval practice helps us keep information at our mental fingertips.
Old Habits Die Hard
Research shows that methods like blocked practice or rereading work fine if performance is only measured shortly after learning (Karpicke, 2012; Karpicke, Butler, & Roediger, 2009; Kornell & Bjork, 2007). That's why we're fooled so easily. Teachers can look at quiz performance soon after teaching with blocked practice examples, see good grades, and think, "My students are really learning!" Yes, for the short term, but will they know the material on the final exam?
Similarly, students can repeatedly read their textbook or notes and then ace a test the next day and think they've really learned the content. But they may wonder why their knowledge has melted away by final exam time or, worse yet, when they need to know the material outside of class.
Nevertheless, most students still prefer to use blocked practice, and surveys have shown that even college students prefer to study for tests by rereading rather than by using a more active strategy like self-testing (Karpicke, Butler, & Roediger, 2009). Moreover, even when students experienced the benefits of these approaches in experimental settings, when asked how they would like to learn subsequent material, most still picked blocked practice over mixed practice and rereading over retrieval practice (Karpicke, 2012; Kornell & Bjork, 2007).
And it's not just students who hold such illusions. So do their teachers. And so did I—until I found a better way.
More Than a Four-Letter Word
I used to teach the way most college professors do, thinking that my lectures were precious and that tests were mostly for assigning grades to students and making sure they did the reading. I never thought of giving tests or quizzes as a way to teach. It took my own research to show me how misguided this impulse is, how giving tests can be crucial to learning.
In light of all the controversies over standardized testing and teaching to the test, test is clearly a four-letter word to most educators. But there's a better way to think of tests, which my colleagues and I call retrieval-enhanced learning (Brown, Roediger, & McDaniel, 2014). Standardized tests have their place in education, but our use of testing is completely different. We advocate frequent low-stakes quizzing or testing to enhance the process of learning. Students' self-testing (for example, through flash cards) is also a good way to learn.
Retrieval practice by way of testing really works (Karpicke, 2012). When teachers introduce desirable difficulties into the learning process, students learn. Preparing for tests is hard work for students, but they learn both from preparing for them and from taking them (especially when they are given feedback on their answers).
The problem is that the benefits of testing are not immediately obvious because retrieval practice facilitates long-term use of knowledge. Further, retrieval practice doesn't enhance knowledge just of basic facts but also of procedures, visual representations, and narratives. Some evidence also shows that retrieval practice may enhance transfer of knowledge to new situations (Butler, 2009).
I've changed my approach because of my and others' research on testing. I now require some sort of written assignment in every class period for every class I teach. In some cases, students take a quiz. For higher-level courses, students write a one- or two-page essay on some aspect of their reading assignment. These assignments require students to grapple with the material daily, not just coast along until the big test. The assignments or quizzes also make them write and, with luck, think hard about the material (at least harder than they would have without the assignment).
In some courses, I've given students sets of essay questions that might appear on the test; if they use these as a study guide and then practice retrieving answers to those questions, they'll do well on the tests. I also suggest methods for students to study effectively: outlining material (putting it in their own words); relating the material to material they already know and to examples from outside the classroom; and composing questions as they read the material as well as using their own questions for retrieval practice before the test. In my classes, the daily quizzes or essays count for a small part of students' grades (hence the adjective low-stakes before quizzes). Because students know they'll have a quiz or other assignment, I never give them surprise quizzes. They know to prepare every day.
Another problem we see in college is student attendance. Having a daily graded assignment in class makes attendance mandatory for a good grade. Further, if the teacher concentrates on important material on the quizzes in class, students will learn it better through retrieval practice.
I still give larger, hour-long tests in addition to the shorter quizzes as well as a cumulative final exam, and these count for the bulk of the grade. The quizzes are low-stakes events to keep students on target throughout the semester. Of course, I have more grading to do, but it's worth it for the improvement in student learning.
Embrace the Difficulty
When I describe research on the benefits of introducing difficulties, such as frequent testing, into the learning process, teachers often tell me "my students will rebel" or "my student ratings will suffer." But my experience, and that of my colleagues, says otherwise.
Because we teach at the college level, our students can often vote with their feet; they can take an easier section of the course or a different course entirely. However, they typically haven't done so. Once we explain our reasons for frequent quizzing, they accept them well. Many students say that when courses are taught this way, they're always caught up with their work and preparing for big tests or the final exam is easier.
Further, in our research using frequent quizzing in middle school classes, the majority of students report that daily quizzing makes them less anxious about taking tests. (We had originally received criticism that constant testing would make students more anxious.) The quizzes in our research are "zero stakes," in that they have no effect on students' grades, so students often look at them as a kind of game. And the students always get immediate feedback in the form of the correct answer.
The middle school students also point out that taking quizzes every day helps break up the class. The quiz takes only five minutes or so, but that's enough for a change of pace. Further, they're practicing what they'll need to do when they take the "real" test later. Thus, their reported level of anxiety about tests is often lower than it used to be (Agarwal, D'Antonio, Roediger, McDermott, & McDaniel, in press). Once they understand the reasons for the practice, students don't seem to mind daily quizzing or other techniques of active retrieval, such as responding in writing to questions based on their reading.
Will students find you harder as a teacher if you adopt practices that embrace difficulties and foster long-lasting learning? Possibly. However, I often ask my undergraduate students to tell me about their favorite teachers in middle school and high school, the ones who set them on their path to college. They're happy to do so, and everyone has a story. Somewhere in the conversation, I ask where their favorite teachers fall on the ubiquitous "easy to hard" continuum that students use for teachers. I typically find that students describe their favorite teachers as "demanding," "rigorous," or "tough but fair." Other comments included, "she made me think hard and discover what I could do" or "I never thought I could understand physics, but he led me through it and showed me the way."
So don't worry about embracing desirable difficulties in your classes, even if your classes are perceived as somewhat harder than they used to be. Your students will profit, and they'll thank you in later years when they still know the material you taught them.
