Transfer-appropriate processing
Learning helps most when the mental work at study matches the work at test.
What it means
Retention is best not when encoding is simply 'deep' but when the cognitive operations performed during learning match those required at retrieval. The principle refines levels-of-processing by showing that even shallow, rhyme-based encoding can outperform semantic encoding if the eventual test is itself rhyme-based. It reframes memory performance as an interaction between encoding and retrieval conditions rather than a property of encoding alone, aligning closely with encoding specificity. Practically, it implies that study should rehearse the form of the future task — recalling for a recall test, applying for an application test. The idea reshaped how researchers interpret apparent contradictions in the depth-of-processing literature.
Examples
Students who study by self-testing outperform re-readers on a recall exam, but the advantage narrows when the test merely asks for recognition.
A language learner who studies by reading translation lists aces the written quiz but freezes ordering coffee abroad; the practice rehearsed recognising words, not producing them under time pressure.
A pianist who only ever practises slowly in a silent room stumbles at the recital, because nothing in the rehearsal matched playing at tempo with an audience watching.
First described in Morris, Bransford & Franks (1977).