How should implicit memory phenomena be modeled?

In a reply to R. Ratcliff and G. McKoon's (1995a) article on bias in the object decision task, D. L. Schacter and L. A. Cooper (1995) critiqued their theoretical arguments and presented an updated view of priming in the object decision task. In the present article, the updated view is examined in detail, and it is questioned whether Schacter and Cooper's explanation of the data is sufficiently articulated to be falsifiable. It is also argued in the present article that evidence from other research domains is not directly supportive of the memory systems hypothesis and that the statistical power available in data from object decision experiments is not great enough to test some relevant hypotheses. Finally, the bias hypothesis (Ratcliff & McKoon, 1995a) is elaborated to show that it requires a particular pattern of experimental results and that it serves as a target phenomenon for modeling.