ABSTRACT
Three experiments address the assumptions, derived from a dual-force model, that positive mood supports assimilative (knowledge-driven) processes whereas negative mood supports accommodative (stimulus-driven) functions, and that mood-selective recall (mood congruency) is mainly a matter of assimilation. The generation-effect paradigm was borrowed from memory research to test these assumptions. In Experiment 1, the theoretical variable, degree of assimilation, was operationalised by the ease with which stimulus meaning could be generated from word fragments. In Experiment 2, self-generated stimuli (assimilation) were compared to experimenter-provided stimuli (accommodation). As predicted, positive mood supported assimilation which in turn enhanced mood-congruent recall. In Experiment 3, retrieval mood rather than encoding mood was manipulated. In this situation, positive mood facilitated the recall of all self-generated information, whether congruent or not. The empirical results are generally consistent with the predictions derived from the dual-force framework.
