Learning Strategies Are Associated With Distinct Neural Signatures is the not-very-“sexy” title of a report in today’s Science Daily about a very important study on learning and the brain.
It appears — at least, to me — to provide more scientific evidence to the perspectives on motivation that Daniel Pink and others have written about (you can read more about it at My Best Posts On “Motivating” Students).
Before I discuss the study, I want to first review how I have described Pink’s perspective on motivation and how I’ve experienced it in the classroom:
He cites a lot of research debunking the effectiveness of extrinsic rewards on motivation. This isn’t news to the many of us whom have read Alfie Kohn’s excellent book Punished By Rewards. However, he seems to provide a slightly more nuanced critique.
Pink basically says that extrinsic rewards do work — for mechanical work that doesn’t require much higher-order thinking. But he says research says that it will not work for anything that requires higher-order thinking skills and creativity.
This analysis mirrors my own experience in the classroom. In Have You Ever Taught A Class That Got “Out Of Control”? I shared the challenges I faced last year in using extrinsic motivation to get students into a new pattern of behavior, and then moving them back toward intrinsic motivation. Using “points” was definitely effective in getting the class under control. They received them for being focused and doing their work.
However, I didn’t think students started doing their highest quality work until they were “weaned” off the point system and began to gain what Pink calls “autonomy, mastery, and purpose.” Pink says that those are the three essential elements in generating higher-order thinking
From what I have read from Pink and others, most — if not all — of this perspective is based only on the behavior of people in experiments (let me know if I’m wrong on this), which is certainly valid.
This new study, though, seems to validate these conclusions based on actually analyzing people’s brains during their behavior.
It contrasts “model-free learning,” which is based on rewards, with “model-based learning.”
Here is how one of the authors of the study describes the difference between the two:
“Think about a situation in which you always take the same route when driving home after work, but on a particular day the usual way is blocked due to construction work,” Gläscher says. “A model-free learning system would be helplessly lost; it is only concerned with taking actions that in the past were rewarding, so if those actions are no longer available it wouldn’t be able to decide where to go next. But a model-based system would be able to query its cognitive map and figure out an efficient detour using an alternative route.”
You can read more about their experiment in the report. It seems to me to pretty accurately reflect the differences between learning based on rewards and learning based on intrinsic motivation.
What do you think? Is my interpretation of this study correct?
I think you’re overstating the results of the study a bit. These researchers have found that the brain processes different types of learning in different areas of the brain. They state that both types of learning are necessary, and that actually the more basic level reward based learning is more efficient than the higher level process. Neither is inherently “better” than the other. Their goal was only to show that these processes occur in the brain differently, activating different neural pathways. I suppose you could argue that model-based learning “takes” a bit better because more of the brain is engaged, but I think that reward based learning works very well, too (it only takes grabbing a hot pan once to know not to do it again). In both approaches, the brain is trying out theories of what works or not, one is just more complex and involves more reasoning, prediction and so on.
Simple learning does not equal external reward, nor does complex learning equal internal reward. External reward (or lack thereof) reinforces behavior because of simple cause and effect, and it’s effective. A student learns to continue engaging in the behavior, or not, because he receives or avoids something. Intrinsic motivation as you describe it is still reward based…it’s just that the reward is pleasure at the finished product, or self-esteem, or autonomy, etc. Pride over a good result really isn’t that complex, is it?
How did you wean the students off the point system?