Academic Articles 2: Interleaved Practice

So, it turns out that a lot of debate instructors, myself included, have been running drills the wrong way. I came to this startling and bothersome realization while reading two books, both of which I highly recommend. The first book is Make It Stick: The Science of Successful Learning by Peter Brown, Henry Roediger III, and Mark McDaniel; Brown is a novelist and the other two are accomplished cognitive scientists. The second book is Small Teaching: Everyday lessons from the Science of Learning by James Lang; Lang is the Director of the Center for Teaching Excellence at Assumption College.

Both books dedicate their third chapters to the idea of ‘interleaving.’ Interleaving, as explained by Lang has two key aspects “(a) spacing out learning sessions over time; and (b) mixing up your practice of skills you are seeking to develop” (65). For both aspects, I will briefly review some of the literature that supports them and then mention some takeaways for debate camp. I will then end this piece by mentioning two important qualifications for when interleaving may not be the best strategy.

Spacing

Bloom, Kristine C., and Thomas J. Shuell. “Effects of Massed and Distributed Practice on the Learning and Retention of Second-Language Vocabulary.” The Journal of Educational Research 74.4 (1981): 245-248.

Moulton, Carol-Anne E., et al. “Teaching Surgical Skills: What Kind of Practice Makes Perfect?: a Randomized, Controlled Trial.” Annals of Surgery 244.3 (2006): 400-409.

Cepeda, Nicholas J., et al. “Distributed Practice in Verbal Recall Tasks: A review and Quantitative Synthesis.” Psychological Bulletin 132.3 (2006): 354.

1.1: Brief Literature Review

The consensus of educational experts is that spacing out learning, especially in a way that allows sleep in between lessons, significantly improves long-term retention of information. The first two articles listed above are individual studies. One looks at memorizing French vocabulary in high school and the second looks at mastering surgical procedures in medical school. Both studies found that, in the short-term, students who space out practice do no better than those who study in one long session; and in fact, they sometimes do worse. However, when you assess the students later, those who spaced out their practice significantly outperform those who did not.

For example, in the first study student were given, either 30 minutes on one day or 10 minutes on three consecutive days to study 20 French vocabulary words. At initial testing (at the end of study period) performance “was virtually equivalent for the two groups” (246). However, when retested four days later the students who had distributed practice did about 35% better.

The third article is a meta-analysis of 839 discreet assessments which concludes that “distributing learning across different days (instead of grouping learning episodes within a single day) greatly improves the amount of material retained for sizable periods of time; the literature clearly suggests that distributing practice in this way is likely to markedly improve students’ retention of course material” (371).

When students focus on material within a single day they end up, primarily, calling up information from their short-term memory. However, if they break up instruction, and especially if they sleep in-between instructional periods, then they call up information from long-term memory. When you use your long-term memory you solidify neural retrieval of that information in a way that will be helpful in the long run.

Brown et al. note that:

Psychologists have uncovered a curious inverse relationship between the ease of retrieval practice and the power of that practice to entrench learning: the easier knowledge of a skill is for you to retrieve, the less your retrieval practice will benefit your retention of it. Conversely, the more effort you have to expend to retrieve knowledge or skill, the more the practice of retrieval will entrench it. . . .

Effortful recall of learning, as happens in spaced practice, requires that you “reload” or reconstruct the components of the skill or material anew from long-term memory rather than mindlessly repeating them from short-term memory. During this focused, effortful recall, the learning is made pliable again: the most salient aspects of it become clearer, and the consequent reconsolidation helps to reinforce meaning, strengthen connection to prior knowledge, bolster the cues and retrieval routes for recalling it later, and weaken competing routes.  Spaced practice, which allows some forgetting to occur between sessions, strengthens both the learning and the cues and routes for fast retrieval when that learning is needed again . . .. (79-82)

1.2: Implications for Debate Camp

There are lots of ways that instructors can adopt these insights at camp. For example, it is not uncommon for labs to dedicate whole days to specific instruction. They might spend one day where they focus on Ks, then one day where they focus on theory, then one day where they focus on policy style arguments. However, if spacing improves retention, then it would be better to spend half of the day on theory and half the day on Ks and then repeat that the next day. While that spacing will cause students to feel like they have not developed as deep an understanding (because some review inevitably will need to happen on the second day) students will experience superior retention in the long-run.

Similarly, when I am running drills I often will do the same drill with a student for an extended period. I might spend an hour one night working with a student on generating answers to confusing frameworks. By the end of the night, they will be a lot better at generating answers. However, it turns out that five days later a lot of that skill development will be lost. It would make more sense for me to instead spend 20 minutes each night, for three nights running that drill. Rather than practicing with three different frameworks in one night, instead, I should practice with one framework per night for three nights.

A second distinct application deals with modules. VBI does not just have one off modules. We also offer seminars and tracks which build on the same content over multiple days. These build the benefits of spacing into the curriculum. And these studies seem to indicate that VBI should be placing a bigger emphasis on module tracks than we have in the past.

Interleaving

Kerr, Robert, and Bernard Booth. “Specific and Varied Practice of Motor Skill.” Perceptual and Motor Skills 46.2 (1978): 395-401.

Rohrer, Doug, and Kelli Taylor. “The Shuffling of Mathematics Problems Improves Learning.” Instructional Science 35.6 (2007): 481-498.

Goode, Michael K., Lisa Geraci, and Henry L. Roediger. “Superiority of Variable to Repeated Practice in Transfer on Anagram Solution.” Psychonomic Bulletin & Review 15.3 (2008): 662-666.

2.1: Brief Literature Review

Not only is it helpful to break up instruction, it is also useful to intertwine different things being learnt within each spaced out session. This principle can be a little hard to explain, so I will just jump into describing the studies.

In the first study, children, aged 8 and 12, were divided into two groups who practiced throwing beanbags into a target that were obscured from sight, though they can see the target before and after the throwing. One group had to practice tossing a beanbag into a target 3 feet away, the second group had to rotate between throwing a beanbag into a target 2 and 4 feet away.

At the end of the 12-week program, all the children were tested to see how well they could throw a beanbag at a target 3 feet away. The children who practiced at 2 and 4 feet significantly outperformed the children who practice only at 3 feet. Interleaving two different tasks meant the students performed significantly better at a new task then the students who had exclusively practiced that one task.

In the second study, students had to practice (at multiple spaced sessions) solving math problems that involve calculating the area of four different types of solids. Half the students were taught how to solve one type of problem, and then would practice it before moving on to the next type of solid. The other half were taught how to solve all four types of problems at the beginning and then practiced solving the same problems in a random order.

On the practice problems the first group, averaging an 89%, significantly outperformed the second group, averaging a 60%. However, when the two groups were retested a week later the first group had dropped to a 20% average while the second group improved to 63% average. Originally the first group did 150% better, but on the second test, the second group did over 300% better!

The idea of interleaving says that by mixing in multiple different skills at once into your practice regiment you significantly improve your long-term performance on all of them.

2.2: Implications for Debate Camp

The fact that students learn better from interleaved practice should not be surprising to debate instructors. Most instructors would agree that practice rounds are one of the most effective ways to learn, and part of what distinguishes practice rounds from lots of drills is that they interleave lots of different skills (C.X., case refutation, strategy, crystallization, weighing etc.).

However, there are other ways that instructors can incorporate interleaving, besides just running practice rounds.  First, interleaving shows why it is so valuable to have debaters practice different types of debate skills. If a debater must practice phil debate, then K debate, then theory debate interleaved, that will likely help them get better at policy debate than just practicing policy debate would have (and the same is true for any type of debate skill). Interleaving different skills will improve particular debate performance.

This is one of the reasons why our ‘focused labs’ are not like old focus weeks, where you just learn a certain type of debate. The idea of a K-focused lab is not to spend the whole time just learning K debate, but to learn all types of debate (just like a normal lab) but with a slight focus.

Another way to apply this insight is to keep rotating drills. Many instructors love to run the same drill repeatedly until the debater can perform it perfectly. Students and instructors love this kind of drill because it gives debaters the illusion that they are mastering a skill. However, it is illusory. Those benefits will be a lot less lasting than if they had switched between drills, coming back to that speech only after allowing the skills to lag.

Important Qualifications

Kornell, Nate, and Robert A. Bjork. “Learning Concepts and Categories: Is Spacing the “Enemy of Induction”?.” Psychological Science 19.6 (2008): 585-592.

Carpenter, Shana K., and Frank E. Mueller. “The Effects of Interleaving Versus Blocking on Foreign Language Pronunciation Learning.” Memory & Cognition 41.5 (2013): 671-682.

3.1: The Need for Buy-In

I mentioned that instructors and students often prefer to repeat one drill over and over again till they feel like they have perfected it. It turns out this is a general problem.

In this first study, students would learn about artists’ styles. Sometimes they would study a single artist’s work in a batch and sometimes they would interleave different artist’s work.  As you should expect, interleaving proved the much more effective strategy for being able to identify a painter by looking at a painting.

The troubling thing is “the participants’ metacognitive judgments were strikingly at odds with their actual prior performance. Of the 72 participants who did not say that learning in the massed condition and learning in the spaced condition were ‘about the same,’ 64 thought massing had been more effective than spacing” (590). Even having taken the tests which demonstrated they learnt better by interleaving the participants still felt like they learnt more from the massed instruction.

As Lang puts it:

Learning through interleaving can seem frustrating to learners, at least initially. In experiments in which learners have the opportunity to learn through blocked or interleaved practice, they overwhelmingly choose blocked practice because it gives them a feeling of mastery over the material. Pausing before you have fully mastered something can feel frustrating, as can be the demand to recall material or practice skills you thought you had mastered but then realize you don’t know as well as you had imagined.  (82)

Thus, students are not going to like it when you switch over to interleaved instruction. However, it is better for the students. It is important that you explain to your students why you are teaching as you are, and get them to buy into the procedure. Lang says, “make sure that you speak to your students about the benefits of interleaving, about the nature of your assessments, and about the differences between short- and long-term learning” (82-3). If you don’t get buy-in your students will just be frustrated with that they perceive to be slow development.

3.2: Some Exceptions

One final point, it turns out that there are a few exceptions where massed instruction is preferable. This is the conclusion of the second study, which found that when learning French pronunciation, extended blocks of time was preferable. In general, massed practice turns out to work better when you are first introducing a subject. Thus, if you are teaching novice lab and are going over cases, you don’t want to interleave between research, contention arguments and V.C. arguments. Doing so will just be confusing. Until you have a certain minimal grasp, interleaving can be counter-productive.  Lang acknowledges that “blocked study or practice, it seems to me, is an appropriate first step for any learning activity” (73).

Conclusion

I hope you find this research helpful, both as students who are planning what drills to work on and as instructors who are planning out activities for students. Obviously, there is a lot more to be said on this subject, so feel free to ask any questions you have in the comments and I will do my best to answer them (or more likely direct you to places in the literature which answer them). Also, let us know if there is a topic you would like us to look into for future posts in this miniseries!