Etude expérimentale de l'activité de recherche mnésique dans la reconnaissance à long terme

Summary The experiments reported investigate the influence of attention and short-term accessibility on context effects in long-term recognition memory. The results show that the detrimental effects of context alteration between the study and the test are a function of the amount of attention paid to context. Moreover, it is shown that using strongly associated contexts in the study and the test will improve performance in a recognition test when subfects pay a great deal of attention to the context. This interaction between the effects of attention and context is shown to be affected by the degree of short-term accessibility as measured by forward and backward recalls. However, the probability of recognition for forward recallable items was found to be the same as for/backward recallable items. This result differs from that reported by Rabinowitz, Mandler and Barsalou (1977). This divergence can probably be assigned to an order effect triggered by the sequence of retention tests. Consequently, backward retrieval cannot be considered to characterize recognition memory specifically. Results support the hypothesis that check retrieval in recognition starts from the context of the test, with subjects trying to retrieve the context likely to have been used with the target in the study. This hypothesis accounts for the effects of attention paid to context and those of associative strength between test context and study context.