Although research into implicit stereotypic knowledge conventionally employs priming paradigms to activate automatically such knowledge, other measurement techniques have been utilised (Hastie, 1984; von Hippel et al., 1997; Maass, Ceccarelli & Rudin, 1996). In a study by Hastie (1984) that assessed the tendency to ascribe causal attributions for behaviours, it was found that behaviours incongruent with stereotypic expectancies were more likely to be explained than those congruent with the stereotype. This tendency von Hippel et al. termed attributional implicit prejudice. On this basis, von Hippel et al. (1997) developed a list of incomplete sentences that described behaviours congruent or incongruent with gender stereotypes, and so named it the Sentence Completion Task. Respondents were required to complete the sentences, and their tendency to provide explanations for congruent behaviours as compared to incongruent behaviours was compared. In order to reduce participant awareness of the aims of the task, incomplete sentences unrelated to gender stereotypes were incorporated into the task. If respondents were unaware of the way their responses were interpreted, then according to Bargh's thesis, the task could feasibly be measuring a level of automatic processing.
In testing this measure Von Hippel et al. found that participants tended to explain stereotype-incongruent more than -congruent behaviours and thus possessed implicit attributional prejudice. He further found that this measure correlated with other measures of implicit stereotypic knowledge, and not with measures of explicit beliefs. However, it must be noted that, to date, this task has not been employed along side the customarily employed priming paradigms; an objective of this current study.