Desirable Difficulties and Inductive Learning

Last month, Jeff Bye’s post on desirable difficulties in the classroom inspired some good discussion. One question that came up was, Do desirable difficulties only apply to rote memory tasks, or can we use them for other types of learning as well? I’m going to expand on that idea using inductive learning as an example. As a refresher, a desirable difficulty is a learning condition that makes encoding more difficult, but also engages processes that support learning and improve long-term retention (Bjork, 1994). For example, when students study for an exam, they often engage in several hours of “cramming” (massed studying) instead of spacing out their study intervals throughout the week leading up to the exam. Besides being less convenient, spacing seems more difficult—every time you return to the material, you feel like you need to re-learn part of what you studied previously. However, as teachers will tell you, spacing leads to much better long-term retention of the material—thus, it is a desirable difficulty.

In an inductive reasoning task, people must make generalizations from a limited collection of specific examples. They can then apply these generalizations to the categorization of novel stimuli. For example, consider a child learning vocabulary for the first time. She may see her own family’s dog, a neighbor’s dog, and a third dog in the park. From these exemplars, she is able to categorize a new animal on the street as a dog even though she has never seen that particular dog before. This is an example of an inductive learning task.

Kornell & Bjork (2008) studied just this type of learning. They presented 6 different paintings by each of 12 relatively unknown artists. The paintings were presented either massed (6 paintings by the same artist in a row) or spaced (each painting by a given artist was interleaved with paintings by other artists). After the study phase, participants were shown a novel painting by one of the 12 artists and asked which artist had created the painting. They were also asked which study schedule—massing or spacing—they thought helped them learn the best. An overwhelming number of participants preferred massing, but people performed significantly better on the induction task when an artist’s paintings were spaced.

Bottom line—desirable difficulties can be successfully applied to inductive learning. Despite the apparent difficulty of inducing characteristics of a category across spaced presentations of exemplars, spacing has been shown to be more beneficial than massing not only for college students, but also for children (Vlach, Sandhofer, & Kornell, 2008) and older adults (Kornell, Castel, Eich, & Bjork, 2010). If you are interested in learning more about these topics, check out the papers below, Nate Kornell’s website at Williams College in Massachusetts, or the Bjork lab website at UCLA.


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.

Kornell, N. & Bjork, R.A. (2008). Learning concepts and categories: Is spacing the “enemy of induction”? Psychological Science, 19, 585-592.

Kornell, N., Castel, A.D., Eich, T.S., & Bjork, R.A. (2010). Spacing as the friend of both memory and induction in young and older adults. Psychology & Aging, 25(2), 498-503.

Vlach, H.A., Sandhofer, C. M., & Kornell, N. (2008). The spacing effect in children’s memory and category induction. Cognition, 109, 163-167.