Implicit and explicit attitude tests are often weakly correlated, leading some theorists to conclude that implicit and explicit cognition are independent. Popular implicit and explicit tests, however, differ in many ways beyond implicit and explicit cognition. The authors examined in 4 studies whether correlations between implicit and explicit tests were influenced by the similarity in task demands (i.e., structural fit) and, hence, the processes engaged by each test. Using an affect misattribution procedure, they systematically varied the structural fit of implicit and explicit tests of racial attitudes. As test formats became more similar, the implicit-explicit correlation increased until it became higher than in most previous research. When tests differ in structure, they may underestimate the relationship between implicit and explicit cognition. The authors propose a solution that uses procedures to maximize structural fit.
Copyright 2008 APA, all rights reserved.