Key points are not available for this paper at this time.
Participants responded to pairs of words under different processing goals. In the first four experiments, they identified different attributes of the second word of each pair. Priming effects of irrelevant evaluations occurred if and only if the goal was to identify the valence of the second word. Equivalent goal-dependent priming effects were found for irrelevant nonevaluative attributes of the word pairs. In the second set of four experiments, participants made same-different comparisons on the word pairs. Irrespective of which (nonevaluative) attributes were task-relevant, effects of irrelevant evaluations occurred. Equivalent goal-independent effects of irrelevant, nonevaluative attributes were not found. A ninth study found crucial effects of both series simultaneously. The results bear on the issues of goal dependence of affective priming and automatic attitude activation, of whether affective processing differs from cognitive processing, and of the mechanisms underlying the effects of irrelevant evaluations.
Klauerxy et al. (Sat,) studied this question.