Interleaving¶
Definition¶
Practicing several distinct but related skills, problem types, or categories in mixed, unpredictable sequence — rather than drilling one type repeatedly before moving to the next (blocked or "massed" practice) — builds a different and more valuable skill: recognizing which solution or approach a given situation calls for. Interleaved practice performs worse during the practice session itself but produces better performance later, on new, unlabeled problems.
In the Book¶
Chapter 1 opens the idea with a study of eight-year-olds tossing beanbags into buckets: children who practiced on buckets two and four feet away (never three feet) outperformed children who drilled only on the three-foot bucket when both groups were later tested on the three-foot target. Chapter 3 develops the mechanism with a study of college students learning to compute the volumes of four obscure geometric solids (wedge, spheroid, spherical cone, half cone). Students who practiced problems blocked by type scored 89 percent correct during practice versus 60 percent for students given the same problems in mixed order — but on a test a week later, the interleaved group outperformed the blocked group substantially. The book also cites interleaved practice at identifying bird species or painters' styles as building the capacity to extract what distinguishes categories, not just memorize examples of each.
Why It Matters¶
Blocked practice optimizes for the feeling of fluency during the exercise — each repetition is easier than the last because context doesn't change — but that fluency is an artifact of predictability, not of understanding. Real-world problems rarely arrive pre-sorted by type; the skill that transfers is recognizing which of several tools a novel, unlabeled situation calls for. Wherever the true task is later discrimination among options rather than execution of one known routine, training in mixed, unpredictable order builds the capability that blocked training only appears to.