To determine the extent to which the increased race bias seen in the alcohol group reflects differences in the influence of automatic or controlled processes, we calculated estimates of these processes using the process dissociation procedure (PDP) approach first outlined by Jacoby (e.g., 1991), and applied to WIT performance by Payne (2001, 2005). As explained by Payne (2005), the PDP approach assumes that any given behavior is determined by both automatic and controlled processes. The relative influence of these processes can be estimated in tasks in which some trials allow these processes to act in concert while others place these processes in opposition. On black-gun trials in the WIT, both automatic stereotyping and the goal-driven “gun” response call for the same behavior (i.e., congruent trials). In contrast, on black-tool trials automatic stereotyping calls for the “gun” response, which opposes the goal-directed “tool” response (i.e., incongruent trials). The critical set of equations for deriving PDP estimates of controlled (C) and automatic (A) processing components are as follows: C=P(correct∣congruent trials)−P(stereotypic error∣incongruent trials)A=P(stereotypic error∣incongruent trials)/(1−C)
