When Does Affect Influence Consumer Judgement? A Meta-analytic Integration, 1983-2020

This research addresses the debate of whether creating a positive affective reaction in a marketing environment will lead to more favorable consumer reactions. When will these effects be greatest? We report a meta-analysis of 44 papers from 1987 to 2020 that reconciles mixed findings in the literature on affect and develop a comprehensive model to provide a more nuanced understanding of the nature of affect effects. The majority of the literature highlights an affect-as-information effect (AAI) and our results suggest that under certain conditions positive affect can increase favorable evaluation by as much as 94% over a negative affect baseline. We examine when the AAI effect is enhanced as a function of: judgment factors (i.e., malleability and relevance), target factors (i.e., representativeness, affect source, task involvement, and cognitive complexity), situation factors (i.e., arousal, culture, and salience. We find that when affect is seen as more representative of the target, the source of affect is semi-integral to the target, the task is less cognitively complex and involvement in the task is higher the AAI effect is even stronger. Further, when the situation renders affect more salient there is a stronger AAI effect on evaluation. When the situation is characterized by lower arousal, there is a stronger AAI effect on behavior. Finally, in Western cultures a stronger AAI effect is observed for evaluation and in Eastern cultures a stronger AAI effect is observed for behavior.A classic result in the marketing literature—as well as some of the psychology literature—is that affective feelings, whether integral or incidental, often result in affect-congruent evaluations. In this paper, we report a meta-analysis of 90 experimental studies of the affect-congruent-evaluation (ACE) phenomenon, with a total of 212 independent effect sizes. In this analysis, we document the typical size of ACE effects on both attitudinal and behavioral measures of evaluation; we systematically examine the primary proposed moderators of these effects; and, importantly, we evaluate and compare the fit of four major theoretical explanations of these effects. The results show that, on average, ACE effects are midway between “small” and “medium” size, and they are comparable for attitudinal and behavioral measures. These effects depend on nine moderating factors that are person-related, target-related, feelings-related, or task-related. Overall, these effects are best accounted for by a combination of two explanations: affect-as-information and judgment simplification (use of heuristics). There is little to no evidence that the phenomenon is driven by thought-priming or by evaluative conditioning.