Are multiple-choice practice tests effective, or do you need to write full answers?

Multiple-choice practice tests are effective — especially strong in the meta-analyses (g ≈ 0.70) — provided you get feedback.

Adesope, Trevisan & Sundararajan (2017) found multiple-choice practice tests especially strong, g ≈ 0.70. The catch: a good multiple-choice question makes you consider and reject the wrong options, which itself is a form of retrieval, and feedback then corrects any error.

That said, production formats (short answer, free recall) generally build more durable memory because they demand more effortful reconstruction. The practical trade-off: multiple-choice is powerful, easy to automate, and easy to feed back instantly — which is why the app's duels use it.