Learning vs. Performance: A Distinction Every Educator Should Know
You’re teaching a concept in class, working through examples and explaining the steps. You have your students do some practice problems and they seem to be getting it! After a while, the bell rings, you dismiss your class, and you leave for home feeling satisfied with your students’ progress and your teaching methods. The next day, you assess how well your students retained the material. Alas, it’s as if you never taught the concept the day before!
Sound familiar? If you’re a teacher, I’m sure it does.
It’s a fascinating paradox in education: Students can be wildly successful on tasks in class but learn virtually nothing; conversely, students can do relatively poorly on those same tasks but learn quite a lot.
These scenarios illustrate one of the most important distinctions in all the literature on human learning and memory — namely, the difference between learning and performance. Learning refers to relatively permanent changes in knowledge or behavior. It is — or at least should be — the goal of education. Performance, on the other hand, refers to temporary fluctuations in knowledge or behavior that can be measured or observed during (or shortly after) instruction.
Simply put, performance is short-term, whereas learning is long-term. What this means is that teachers won’t know if their students have actually learned something until after a period of time in which the students didn’t use or think about the information.
Now, it makes intuitive sense to think that learning and performance are positively related. If your students seem to be understanding a concept in class and perform well on an immediate assessment, it stands to reason that your students have probably learned something. That is, performance should be, at the very least, a rough gauge of learning.
Decades of research suggest otherwise.
When I was a postdoctoral fellow at UCLA with Drs. Robert and Elizabeth Bjork, I spent about two years of my life sifting through the relevant journal articles, some dating back to the 1930s. I, along with Dr. Robert Bjork, published the first integrative review of the evidence — both in the verbal- and motor-learning domains — that bears on the critical distinction between learning and performance (see Soderstrom & Bjork, 2015).
Not only does the literature indicate that short-term performance is a lousy indicator of long-term learning, but it also shows that learning and performance can be inversely related. In other words, when one is up, the other can be down.
Enhancing Learning by Impairing Performance
There’s no doubt that performance gains can fail to support learning gains. As my initial example illustrated, progress during a lesson can be misleading in terms of what has been retained and can fool teachers and students alike. This is especially true if the performance gains are achieved by using teaching strategies that are designed to produce rapid progress. For example, doing the same type of problem repeatedly in class can lead to rapid, yet fleeting, gains in short-term performance.
But what about the flip side? If increased performance can lead to decreased learning, can decreased performance lead to increased learning? The answer is ‘yes.’ As counterintuitive as it sounds, long-term learning can be enhanced by intentionally impairing short-term performance.
The learning strategies that fall under this category are called desirable difficulties (Bjork, 1994). They are desirable because they lead to better long-term retention and transfer of knowledge, and they are difficult because they pose challenges that slow the rate of current progress and induce more mistakes during instruction or training.
In a nutshell, the concept of desirable difficulties embodies the adage: no pain, no gain. Just like how taking the stairs is better for our health than taking the escalator, making learning more challenging can lead to better retention.
As is always necessary when talking about desirable difficulties, it’s important to mention that the level of challenge matters. In order to reap the learning benefits of these strategies, students need to be equipped with the necessary background knowledge and skills to overcome the challenges. There’s a sweet spot with this stuff (think Goldilocks).
What are some examples of desirable difficulties? The strategies that have received the most empirical support include: (1) spacing repetitions out over time rather than cramming them; (2) mixing up different (but related) to-be-learned topics or skills rather than keeping them separate; and (3) testing material rather than presenting it (for more on these strategies, see my article, Want to Make Learning Stick? Make it Harder). These strategies impair short-term performance compared to their more common counterpart strategies, yet they enhance the very target of instruction — namely, long-term learning.
As a rule of thumb, if students aren’t struggling a bit — that is, if their performance isn’t somewhat hindered — they’re probably not engaged with the material in ways that will lead to meaningful, long-term comprehension and understanding.
Our goal as educators is to foster an environment in which equipping our students with knowledge and skills that are both durable and flexible has the best chance to succeed. We want our students to step into our classrooms today knowing more than they did yesterday. To achieve this goal, we need to be aware that short-term performance is a misleading indicator of long-term learning and that making the learning process a bit harder — not easier — is a viable path to longer retention.
As Dr. Robert Bjork put it, a major key to learning is to be “suspicious of the sense of ease and undeterred by the sense of difficulty.”
Bjork, R. A. (1994). Memory and metamemory considerations in the training of human beings. In J. Metcalfe & A. Shimamura (Eds.), Metacognition: Knowing about knowing (pp. 185– 205). Cambridge, MA: MIT Press.
Soderstrom, N. C., & Bjork, R. A. (2015). Learning versus performance: An integrative review. Perspectives on Psychological Science, 10, 176–199.