I originally published this piece on the Learning Scientists Blog.

Introduction: In education, there have been discussions about whether gimmicks -- tricks or devices intended to attract students’ attention -- are good practice in the classroom. On one hand, you one could argue that the first step to learning something is to pay attention, and if the topic to be learned is somewhat dry, adding a fun example will grab attention and therefore could increase learning. (On the basic level, even something that just includes bright colors and loud sounds will capture the attention of an otherwise distracted student, (1)). On the other hand, a gimmick could draw attention to the wrong thing.

When I was in graduate school, the first class I taught solo was a hybrid section of introduction to psychology. The students would watch videos of different professors in the department giving lectures on their expert topics, and then I would meet with the students once per week to go over the material and do activities. I had three sections of 70 undergraduates. Two of the sections were great. But one section --the class that met around 2pm on Wednesdays --was always half asleep.

It was my first time teaching all by myself, so I was nervous, and desperately wanted the class to go well. So when we got to the chapter on learning, I decided to bring in candy to show them how positive reinforcement could increase behavior (in this case, participation behavior). I explained reinforcement to the class and the difference between positive and negative reinforcement. I then asked if anyone could come up with examples. I stood at the front of the class silently and waited. After what felt like a long few seconds, a student raised her hand and volunteered an example. Thankfully, it was correct! I said “great job,” reached into my bag, pulled out a piece of candy, and walked it over to her. The students’ eyes widened a bit. A couple of them raised their hands, and I called on someone. They said “is the candy to get us to participate?” I said yes, and gave him a piece of candy. From there, the examples kept coming. I hardly had to speak the entire class! In later classes when they were tired I was able to make jokes about being a poor graduate student, and begging them to bring their own sugar fix from here forward. It was a turning point in the class, and the students could tell I was trying and became generally more cooperative.

At the end of the semester, when I received my course evaluations, many of the students said “I liked her class, she gave us candy.” Part of me was stunned. Why didn’t they realize that the candy was used to demonstrate a principle?!

Gimmicks in education

I suppose I shouldn’t have been surprised. My students remembered the gimmick, the surface details of the example, without remembering the underlying structure (i.e., the entire point of the activity). In education, there have been discussions about whether gimmicks -- tricks or devices intended to attract students’ attention -- are good practice in the classroom. On one hand, you one could argue that the first step to learning something is to pay attention, and if the topic to be learned is somewhat dry, adding a fun example will grab attention and therefore could increase learning. (On the basic level, even something that just includes bright colors and loud sounds will capture the attention of an otherwise distracted student, (1)). On the other hand, a gimmick could draw attention to the wrong thing. David Didau (@LearningSpy) made this point in a recent blog post. What’s wrong with gimmicks? He writes:

Research on what students remember and notice

Indeed, researchers have asked what students in higher education remember from lectures. Back in 1977, Walter Kintsch and Elizabeth Bates (2) examined students’ ability to recognize statements from their classroom lectures 2 days later and 5 days later. The researchers examined key topic statements, details, and extraneous remarks such as jokes or announcements. The students remembered all three types of statements days after the lecture; however, extraneous remarks were remembered best. In a study on humor and memory for speeches (3), students listened to speeches containing jokes and speeches without jokes. Students remembered the jokes better than the rest of the information, though crucially, memory for the information was not harmed by the presence of the joke. This was true regardless of whether the joke was placed at the beginning of the speech or the end. From this research, we might conclude that students are more likely to remember our extraneous or fun examples, and less so the actual content. However, the presence of the gimmick shouldn’t hurt memory for the real content -- it just may seem to do so by comparison.

But, should we blame the gimmick? In the face of this research, I’m not so sure avoiding the fun gimmicks is the answer (or at least, not always). There is a plethora of research showing that students often notice and remember the surface details of an example rather than the underlying structure. This is part of what makes transfer of information so difficult. (To read more about transfer check out this post and this post.) First, some research (4) has demonstrated that physics experts are able to extract underlying structure from problems to sort them into categories, while physics novices tended to sort problems by surface details.

Noticing and remembering the surface details seems somewhat similar to what my students were doing – remembering candy but not remembering why there was candy.

How do surface details affect transfer?

Students seem to have trouble ignoring the surface details and focusing on the underlying structure of examples. How does this relate to transfer? Gick and Holyoak (5) examined whether college students could use one problem to solve another analogous problem. First, the students read a story about a general trying to capture a fortress:

Then, after a few minutes, the students were given a couple of problems to solve, including a problem analogous to the fortress problem:

The problems have very different surface details. One includes a general, an army, a fortress, roads, and mines, while the other includes a patient, a doctor, a tumor, radiation, and healthy tissue. Yet they both have the same underlying structure, and the solution presented in the fortress problem – break up a large force into smaller forces to converge in the middle – can be used to solve the tumor problem.

However, across a number of experiments, few students were able to spontaneously transfer the solution from one problem to the next. In one of their experiments, only 20% of students spontaneously solved the tumor problem using the analogous general problem. This is surprising because the two problems were presented during the same experimental session! When students were given a hint – “In solving this problem you may find that one of the stories you read before will give you a hint for a solution of this problem” – many more solved the problem (92%), but this means with an explicit hint 8% of students still could not make the connection between the two examples. Certainly the hint scenario is not practical; how many students have a teacher following them throughout their lives giving them hints about when to apply various things they have learned?

Other work by Gick and Holyoak (6) also showed transfer to be difficult to find, though they did find a few things to help improve spontaneous transfer. Providing multiple examples seemed to help, and especially when the multiple examples had different surface details. Having students actively explain how two examples are similar, encouraging them to extract the underlying structure on their own, also helped improve transfer. Based on this research, it seems to me that whether examples are fun or gimmicky is not the issue. The gimmick just makes it more likely that students will remember the surface details, as opposed to forgetting the example entirely. Instead, it seems to me that the solution is to provide as many fun and engaging examples as possible, and then ask the students to compare and contrast the examples. (Possibly even using elaborative interrogation!)

If students often remember only surface details, does it matter if they're gimmicky?

So, it seems that students have trouble remembering the underlying structure of examples, regardless of whether or not they are fun. We started thinking about this issue after we posted our video on using Concrete Examples to improve learning. In this video, Yana dresses up like Pikachu and provides her own concrete example for the abstract concept scarcity. This is technically a gimmick -- a fun example intended to gain attention and make discussions about study tips more fun. But, did we harm learning by using Pikachu? This is an empirical question. Based on the research above, we would infer probably not. Memory for the point of the video should not be made worse because of the presence of Pikachu, even if Pikachu is remembered more than another example might have been remembered. And, our purpose in using Pikachu in the video wasn’t to teach students specifically about scarcity, but to help them remember to make their own fun examples. If all they remember is that we used a Pikachu example, then they get concrete examples! And, after all, we have to capture attention of students (other than our own) in order to get the message about study strategies across to begin with.

References:

(1) Johnston, W. A., & Dark, V. J. (1986). Selective attention. Annual review of psychology, 37, 43-75.

(2) Kintsch, W., & Bates, E. (1977). Recognition memory for statements from a classroom lecture. Journal of Experimental Psychology: Human learning and Memory, 3, 150-159.

(3) Baldassari, M. J., & Kelley, M. (2012). Make ‘em laugh? The mnemonic effect of humor in a speech. Psi Chi Journal of Psychological Research, 17, 2-9.

(4) Chi, M. T. H., Feltovich, P. J., & Glaser, R. (1981). Categorization and representation of physics problems by experts and novices. Cognitive Science, 5, 121-152.

(5) Gick, M. L., & Holyoak, K. J. (1980). Analogical problem solving. Cognitive Psychology, 12, 306-355.

(6) Gick, M. L., & Holyoak, K. J. (1983). Schema induction and analogical transfer. Cognitive Psychology, 15, 1-38.