Context and repetition effects in lexical decision and recognition memory

Abstract

The issue addressed in these experiments is the hypothesis that there are different consequences of experience with words encountered in lexical decision and recognition memory tasks. In Experiment 1, in which a two-word lexical decision task was employed in separate training and test phases, and in Experiment 2, in which the two-word lexical decision task was employed in training and a two-word recognition test employed at test, the following variables were isolated as having different effects on test performance in the two tasks: (1) relatedness of the words in training affected subsequent recognition memory but not lexical decisions; (2) pairs status of the items (repetition of the test items in a changed or identical combination) affected recognition memory for both related and unrelated pairs, but did not affect lexical decisions for unrelated items; (3) the semantic relatedness of the word pairs at test provided an additional advantage for subsequent lexical decisions over and above the mere repetition of two old words independently, but this advantage did not occur for recognition memory. The underadditive nature of these two effects of relatedness at test and item repetition for lexical decisions was demonstrated. Results are interpreted as supporting and extending the contention of L. L. Jacoby and M. Dallas ( Journal of Experimental Psychology: General , 1981, 10 , 306–340) that there are two forms of recognition memory, one that is sensitive only to prior occurrence (relative perceptual fluency) and one that is sensitive to contextual and episodic memory.