Recalibrating positive and negative weighting tendencies in attitude generalization

Abstract

Individual differences in the weighting of positive versus negative information when generalizing attitudes towards novel objects predict a variety of assessments that involve the integration of valence information (Pietri, Fazio, & Shook, 2013). The goal of the current research was to manipulate valence weighting in attitude generalization to demonstrate its causal impact on various judgments and behaviors. In four experiments, participants first played BeanFest—a game in which they approached/avoided novel stimuli (beans) varying in shape and speckles, in order to increase and not decrease their points (Fazio et al., 2004). Following the game, participants classified game beans, and novel ones that varied in resemblance to the game beans as either positive or negative. In the recalibration condition, participants were told whether each classification was or was not correct. Thus, they received feedback regarding the appropriate valence weighting of resemblance to a known positive versus a known negative. In Experiment 1, this recalibration influenced individuals' attitude generalizations regarding other (non-bean) novel objects. We then examined if recalibration would produce far-transferring effects by influencing interpretations of ambiguous situations (Experiment 2), risk assessments (Experiment 3), and finally risk-taking behavior (Experiment 4). Across the four experiments, the recalibration procedure led participants who were initially relatively cautious to be more positive when making these various judgments, whereas people who exhibited an initial risky bias became more negative as a function of recalibration.