The Surprising Power Of Associative Learning


​Most psychology 101 classes cover associated learning in detail. Think Pavlov, think dogs, think salivating. A dog smells food and salivates. But if the dog associates food with the ringing of a bell, the dog will salivate when the bell rings, even if there is no food. The dog has “learned” to salivate at the sound of a bell. 

Think B.F. Skinner, think pigeons, think rewards and punishments. When the pigeon pecks the target, it gets a food reward. So the pigeon tends to hit the target more often. Which gives another reward. Soon, the pigeon has “learned” to peck the target to get food.

In both cases, there’s an association between one thing (a stimulus or a behavior) and another (a response or a reward). Associative learning comes up a lot when explaining why animals do what they do. The basic idea also comes up when trying to form (or break) habits. But what does it have to do with learning? How can people use associated learning to improve teaching students in a classroom? Or developing better study or learning habits?

Key to associative learning is emotions.

Associative Learning in Action


Skinner, a behaviorist, argued that our thoughts, emotions, and actions are more the products of the environment than of independent thought. Scientists don’t really believe that anymore, but Skinner did get this one thing right: people prefer rewards to punishments, and they will often behave in ways that maximize their rewards and minimize their punishments, even if they’re doing it subconsciously. 

Grades are the most obvious classroom reward. And, at first glance, they seem like a classic case of associative learning in action. Teachers reward good performance with good grades. Students value good grades, so they perform. Easy peasy.

Well, not exactly. What if students don’t value grades sufficiently? Or what if grades actually push students to do the wrong thing? Grades can make students strive for performance. But they can also incentivize negative associations, too.

Suppose a student works hard on a project, or studies long hours for a test, but still receives a poor grade? When the grade is linked to performance (and not effort), the student is incentivized to simply stop trying. Or consider a student who doesn’t study very much but performs well anyway. These students should be pushing themselves to try harder or to learn more, but why should they when there’s no incentive to?

The problem is that there are important differences between what we want Skinner’s pigeon to do and what we want students to do. We want the pigeon to peck a target. They either hit the target or they didn’t. There’s no distinction between what we want the pigeon to do and how we’re measuring what the pigeon is doing.

But we don’t just want students to solve this problem or that problem or any particular problem. We want students to use what they’ve learned to solve classes of problems; to have the tools to reason through situations we haven’t even thought of yet. Grades, even at their best, are only an imperfect measure of the durable, flexible knowledge most of us want to impart to students. This is important because the performance that we’re rewarding is not quite what we want students to be doing.​

Most psychology 101 classes cover associated learning in detail.

Associative Learning is, Well, Associative


Another crucial difference between the pigeon and the low-performing student (who I keep comparing to a pigeon) is that we’re asking the pigeon to do something pretty simple. Students face a much more complex set of decisions as they “do” school. 

For instance, “How should I study?” “How much should I study?” “Do I know the material?” “What material is most important?” And eventually, “Which classes should I take?” In other words, it’s not always obvious to them how they can best achieve “good performance.” This is important because students can spend a long time in the maze of decisions they make about learning as they continue to be “punished” with bad grades.

Certainly, we have all had this experience. For me, it was economics 101. I spent hours studying the textbook, doing a practice problem after practice problem. But then came the test and I landed bad grade after bad grade. If I remember correctly, I ended up with a C in the class, and I felt very lucky about it. ​

​Grades also distort how students approach learning. When researchers ask students how they study, students describe one approach if they were studying for a test–a lot of memorization and “cramming.” They describe a different approach if they were actually trying to deeply learn the material–understanding concepts and their practical applications. There is often, then, a clear difference between the type of learning that yields a good course grade and the type of learning that actually leads to an understanding of the underlying content.

Given these complexities, can associative learning still help? I think so. But before we explore the idea, we should associate more on associative learning to best understand how it works.

Key to associative learning is emotions. As Edward Wasserstein, a Professor of Experimental Psychology at the University of Iowa points out, emotions and associative learning are often used by politicians.

“Catchy nicknames like ‘crooked Hillary’ and ‘Don the con’ associatively link political rivals with negative feelings,” he writes. “Doing so can thereby hijack the learning process and prevent voters from more carefully considering all of the strengths and weaknesses of the candidates’ personal qualities and policy positions. So too can connecting candidates with reviled and revered personalities in the familiar political practices of “guilt by association” and “celebrity endorsement.”

How can people use associated learning to improve teaching students in a classroom?

Associative Learning to Improve Teaching and Learning


One solution is to change how we use grades. Instead of awarding grades based on outputs–test performance, term papers, etc.–use them to incentivize behaviors that we know improve learning outcomes. Reward the student for revising and resubmitting their work. Reward the student for deeply participating in the class. Many schools already do this to some extent–awarding so-called “participation points” or allowing students to retake exams they have flunked–but more should take the step necessary to fully incorporate these ideas into course requirements and classroom standards. 

Structure the class to allow for teachers to employ “Spaced Learning” techniques, and reward students for adhering to this structure. Teachers should reward students for making insightful visualizations or organizing principles of the material. In short, schools should step up efforts to reward teachers for taking advantage of learning science.

People can also use associated learning to study better. So associate good things with learning. For some, it might be a good cup of tea while reading the required essays. For others, it might be having candy after doing some hard coding for a computer science class.

More broadly, we can also associate learning to improve our memory. Mnemonics work like this.

A common example is remembering the names of new people you meet at a party or meeting. Suppose you met three new people named Keenan, Sophia, and Mark. A good way of remembering these names is to “hang” them on knowledge you already have. I can remember the names more easily if I think of classic Star Trek, where Kirk, Spock, and McCoy are the lead characters.

Another example is “Never Eat Shredded Wheat”. Many children learn this phrase as a way to not only remember the names of the four cardinal directions, but also to remember their order (north, east, south, west; moving in a clockwise direction). “Never eat shredded wheat” may be a silly command, but it’s a sentence and it’s often easier for kids to remember than four words with seemingly little connection.

Association requires, of course, changing how we think about learning, from incentives to memory. But it’s a way of improving learning outcomes simply by taking advantage of associative learning.

-Ulrich Boser

Leave a Comment

Your email address will not be published. Required fields are marked *