PICTURE OR WORD SUPERIORITY EFFECTS IN IMPLICIT MEMORY - LEVELS OF PROCESSING, ATTENTION AND RETRIEVAL CONSTRAINTS

Previous research with implicit memory tests has shown that the typica l picture superiority effect with explicit memory instructions can be reversed on verbal perceptual tests (e.g., word-stem completion task). Two experiments explored whether the picture superiority effect can b e restored with a verbal conceptual test of implicit memory (i.e., cat egory instance generation task). In the first experiment, participants received either a perceptual test or a conceptual test of implicit me mory. Furthermore, in the study phase level of processing was varied. With a superficial orienting task, a word superiority effect was found on the perceptual test and a picture superiority effect was observed on the conceptual test. Only with a semantic orienting task, priming w as found for pictures on the perceptual rest and for words on the conc eptual test. In the second experiment, this pattern of findings was re plicated with a superficial orienting task in the study phase. Once ag ain, priming for pictures on the perceptual test and priming for words on the conceptual test was found to be at chance level. Dividing atte ntion in the study phase did not influence these results. It is conclu ded that pictures transmit primarily semantic information whereas word s are processed primarily at a lexical level. Both processes have a st rong automatic component. Prior semantic processing improves performan ce on conceptual tests of implicit memory whereas lexical processing i s an important precondition for priming in perceptual rests of implici t memory.