Last month I attended the 2019 Conference on Teaching & Research in Economic Education (CTREE), sponsored by the American Economic Association, in St. Louis, Missouri. The sessions were very high quality and I learned a great deal. One thing that struck me was how little of what I learned was specific to economic education. Hence, the reason for this post.
Two important themes came out of the sessions I attended.
- How little we actually know about teaching and learning. Learning is complex, more than most realize.
- The critical importance of student effort, something which researchers often ignore because of the difficulty in measuring it.
The opening keynote speaker was Stephen Chew, 
a prominent cognitive psychologist, who readers of this blog will know I’ve followed for some time. Chew’s premise was that students and teachers each have a model of learning in their heads, which guides their actions. For some teachers, the model is implicit. A teacher’s mental model determines which teaching methods they select, how those methods are implemented and assessed, and how to make adjustments. The problem is that most of these mental models of both teachers and students are flawed, which makes teaching and learning less effective. In this post I will focus on how to teach more effectively. Chew has created a set of videos to help students learn, which can be found here: http://www.samford.edu/how-to-study/.
For me, the most interesting part of the keynote was Chew’s discussion of cognitive load. This may be because in my earlier studies, I didn’t fully grasp it. Anyway, let me give you my current understanding. In order to complete mental tasks, a person applies mental effort. Mental effort is always limited. Cognitive load is the total amount of mental effort a task requires to complete it. A person can do multiple tasks as long as the total cognitive load is no greater than available mental effort.
Memory is divided into working memory, where new ideas (e.g. data) are placed, and long-term memory, where knowledge and understanding reside. Learning is stored in long term memory in the form of frameworks or models. These are mental structures that give meaning to data. Cognitive psychologists call these “schemas.”
Things in working memory need to be processed into appropriate schemas in order to be moved into long term memory. But working memory is limited—Chew says it can hold only about five “chunks” of information (though experts can hold more). If you try to put more into working memory, the earliest contents get flushed before making it into long term memory.
When I first learned about working and long-term memory, I kind of discounted the theory because it sounded like memorization of facts. Learning is more than memorization of facts and I now see that that is implicit in this theory. We know that novices learn less efficiently than experts do. If teachers don’t recognize this, we may end up trying to do too much, with the results that students don’t learn what we’re teaching. This is related to the “curse of expertise,” which Chew explains as “the more one knows about a topic, the harder it becomes to remember not knowing a topic and the effort required to learn that topic.”
If you’ve taught for any length of time, you will have experienced those classes where at a certain point, a majority of your students are staring blankly at you. If you’re paying attention, you decide to end class early. I think this is where the cognitive load exceeds the mental effort available. If you continue the class session, you may “cover” more material but the students won’t get it.
Novices perceive new data as random facts—they have no schema to frame the data. Experts have existing schemas, so when they perceive new data, they file the data in the appropriate framework. That’s usable knowledge. Data needs to be linked to existing knowledge to make it usable (i.e. part of the schema).
There are three types of cognitive load:
- Intrinsic: How difficult is the subject?
- Germane: Load caused by pedagogy and activities to learn the subject.
- Extraneous: Additional load caused by factors unrelated to learning the subject (e.g. random animations on an instructor’s PowerPoint presentation).
If the cognitive load demanded exceeds the student’s available mental effort, then learning will not occur. There are two issues here. First, is understanding the subject being taught (i.e. the data). Second, is fitting the new data into a schema. The takeaway here is that teachers must monitor, manage and minimize cognitive load to allow schema development; they should also design activities to promote schema development.
As teachers, we need to eliminate extraneous load, and think about how to minimize germane load. Note that there appears to be a tradeoff between intrinsic and germane loads. The greater the intrinsic load, the less cognitive space is left for germane load, suggesting simpler pedagogy. This is something to think about.
The crux of learning is the processing of new data into an existing schema. This processing requires mental effort. It’s often said that students learn best when they can relate new information to something they already know. If they have no existing schema, this processing is harder, which explains why novices learn less efficiently than experts. Since many undergraduates have little or no experience with economics, this adds an additional challenge to an introductory course. Let’s call “study” the mechanism of processing new data. Learning scientists have determined that some study practices are more efficient than others. Reading and highlighting text, while popular among students, are known to be among the least effective ways of studying. Deeper learning requires building lots of links between new data and existing schemas. Reading provides only a weak link.
Herb Simon observed, “Learning results from what the student does and thinks and only from what the student does and thinks.” He continued, “The teacher can advance learning only by influencing what the student does to learn.” Koedinger et al (2016, p.28) found “[t]he learning effect of doing is about six times greater than that of reading.” Yes, that’s correct. Doing (i.e. working with the content) yields six times the learning of reading.
What does this mean for course design? What is doing? Doing means working with the content. Working with the content builds more and stronger links with one’s schema. Skimming a chapter is pretty much the opposite of this. Instead, as one reads, one needs to think about what one is reading. How does what one is reading about relate to what one already knows? Chew calls this Elaboration. Everyone knows (or thinks they know) something about economics. Chapter 2 can be related back to Chapter 1. How is the content different from other concepts? For example, how does accounting profit differ from economic profit (Distinctiveness)? How is this relevant to my life (Personal)? Can I think of any examples? Finally, how might I be expected to use this content to demonstrate my understanding (Appropriate to Retrieval and Application)?
Cognitive psychologists have identified the testing effect, also known as retrieval practice with prompt feedback. From what I understand, students who test themselves, even if they get the problems wrong, will be more likely to get them right on exams. The point is not to score well on your self-test, it’s to build connections between the concept and your schema. In short, assessment is integral to the learning process. This benefit goes away if you look up the answers. You need to force your brain to think about the question, even if you end up getting it wrong.
Two other things which promote deeper learning are spaced practice and interleaving. As I understand them, spaced practice means spreading out your study. Instead of (hopefully) reading the text and then reviewing the material just before an exam, it is preferable to spread the review out over multiple days or weeks. Interleaving means blending older concepts into the study of new material. This forces students to discriminate in retrieving their knowledge (e.g. how do I answer this kind of question?) and leads to deeper connections with the schema. One thing I plan to try in my intro courses this fall is to give students a short assignment each Friday, based on the content we learned the previous week. So after studying new material during the week, I’ll ask them about previous material. This seems like a low-cost activity that’s worth a try.
I am not done thinking about what I learned at the conference. Stay tuned.
