Why does mixing up your practice beat doing one kind at a time?

Mixing up what you practise, called interleaving or varied practice, tends to leave you with better long-term learning than finishing one kind of problem before starting the next. The boost is real but moderate and depends on what you are learning: a 2019 meta-analysis of 59 studies found an average effect of g = 0.42, which its authors call moderate.

The contrast is simple. In blocked practice you focus on one aspect of the material until you have mastered it, then move on. In varied practice the schedule changes task frequently, so you keep meeting new versions of the thing you are learning. Interleaving is also a way of spacing study: the gap between repeats is filled by working on something else.

Think of a maths textbook. Two researchers, Rohrer and Taylor, noted in 2007 that most textbooks made each practice set almost entirely of problems from the lesson just taught, and only rarely shuffled in problems from many earlier lessons.

Blocked practice can look better at first, because the benefits of mixing appear only some time after practice. A test right after practice may therefore miss the advantage, and the mixed schedule shows it later.

The size of the benefit depends on the material. Interleaving was strongest for paintings (g = 0.67), smaller for maths (g = 0.34), and unclear for texts. For learning words, blocking actually won (g = -0.39). The meta-analysis authors concluded that interleaving can foster this kind of learning, but that the setting and the type of material must be considered.