Robert Bjork coined desirable difficulties: obstacles that make learning slower, harder, and more frustrating in the short term, but stronger and more flexible in the long term. Ease during study is often a warning sign, not a success signal.

Core techniques

Generation effect — Struggling to produce an answer (even a wrong one) before seeing the right one improves later learning. Wildly wrong confident guesses produce hypercorrection when corrected.

Spacing — Distributed practice beats massed practice. Eight years after a Spanish vocabulary study, the spaced group retained 250 percent more.

Testing — Retrieval practice, including self-testing and testing before studying, primes the brain. Hint-heavy practice (like Oberon and Macduff the monkeys) looked great during training and collapsed on test day.

Interleaving — Mixing problem types during practice hurts immediate performance but builds discrimination: which strategy fits this problem? Expert solvers in wicked domains evaluate problem type first; kind-domain experts jump to a familiar procedure (kind-vs-wicked-learning-environments).

Making-connections problems — Conceptual math questions that stay conceptual. US teachers often convert them into procedural multiple-choice via hints; Japanese bansho preserves collective struggle.

The feeling trap

Learners and students rate blocked practice and hint-rich teaching as more effective because progress feels fast. Air Force Academy cadets gave low evaluations to professors who hurt their Calculus I scores but helped them in later STEM courses — they were punishing deep learning. Interleaving studies: 80 percent of students preferred blocking; 80 percent performed better with interleaving.

Relation to other wiki concepts

  • spaced-repetition is one instance of desirable difficulty
  • deliberate-practice targets the frontier; desirable difficulties target retention and transfer
  • Early childhood "closed skill" head starts fade out; open skills scaffold later learning

Sources