This research tested a central assumption of attitudinal ambivalence research: ambivalent attitude objects simultaneously trigger positive and negative evaluations. It further specifies at which stage this activation is likely to produce an evaluative conflict. Experiments 1 to 3 involved two evaluative priming paradigms, in which ambivalent stimuli served either as primes or as targets. The Ambivalent Primes Paradigm tested the degree to which the concurrent and unintentional activation of positivity and negativity influences responding to univalent targets. The Ambivalent Targets Paradigm tested the degree to which ambivalent targets entail an evaluative response conflict irrespective of prime valence. Experiments 1 and 2 revealed slower responses for ambivalent trials compared to congruent trials in both paradigms. Employing a longer stimulus onset asynchrony, Experiment 3 attested to the short-lived nature of the joint activation of opposite evaluations triggered by ambivalent primes. In contrast, the deliberate categorization of ambivalent targets was not affected by this procedural variation. Finally, by relying on a valent/neutral categorization task, Experiments 4 and 5 indicate that conflicts triggered by ambivalent stimuli occur at the response selection rather than the exposure stage. Our findings lend original empirical support to the assumption that positivity and negativity are activated simultaneously and unintentionally in ambivalent attitude objects. Moreover, the present research suggests that ambivalence generates a conflict only if the task at hand requires a univalent categorization. We discuss the extent to which the activation of ambivalent attitudes may be automatic and the implications of our findings for dual-process models of attitudes.

This work was supported by an Emmy Noether grant from the German Research Foundation awarded to Mandy Hütter (HU 1978/4-1)