Abstract
In three experiments participants studied AB word pairs and completed two recognition tests. In the first recognition test, which was included in all three experiments, the B word had to be discriminated from two distractors that did not appear on the study list. In Experiment 1, in the second recognition test, an AB target was compared with distractors composed of words not on the study list. In Experiment 2, in the second recognition test, an AB target had to be discriminated from two other pairs that were created by randomly re-pairing A and B words that appeared on the study list. In Experiment 3, on the second recognition test, words from the study list were systematically re-paired to form distractors that contained either the same A term or the same B term as the target pair. Recognition of the B word on the first test was always at least partly independent of recognition of the AB pair on the second test. Even when recognition judgements were restricted to those for which the participants were most confident, all experiments demonstrated significant retrieval independence between the two tests.