Effects of color on perceptual and conceptual tests of implicit memory

Two experiments are reported that explore why recent investigations of impl icit memory failed to find any effects of color information on test perform ance. In the first experiment, participants studied colored pictures as wel l as words printed in colored ink without any memory instructions. During t he test phase, a. verbal and a pictorial version of a color-choice task (a conceptual priming test) were compared to two perceptual tests (word-stem c ompletion and picture-fragment identification). Similar and significant amo unts of priming to color occurred in both color-choice tasks. The perceptua l tests were found to be sensitive to changes in the stimulus-presentation mode from study to test, but stimuli remaining the same color and those cha nged to black-and-white did not differ in the priming scores. In the second experiment, a mild division of attention was introduced in the study phase . Once again, priming to color was observed only in the verbal version of a color-choice test and not in the word-stem completion test. Dividing atten tion did not decrease performance on both implicit tests, whereas an explic it test of color recall for studied pictures suffered from dividing attenti on at encoding. It is concluded that a perceptual attribute such as color m ay be represented and coded by conceptual processing. Furthermore, automati c (or not attention-demanding) encoding processes may suffice for later con ceptual tests of implicit memory. Previous failures to find any effects of color information on implicit performance are attributed to the use of perc eptual priming tests.