Negativity effects in impression formation: A test in the political arena.

Abstract

Previous research had shown that negative information has a stronger influence on impressions of others than does positive information; a tendency known as the negativity effect. The hypothesis that this effect would characterize impressions of presidential candidates was tested using National Election Study surveys from 1984 and 1988. Analyses were conducted at both aggregate and idiographic levels. The aggregate level analysis revealed that personality characteristics that the nation, on average, judged to represent character weaknesses were more predictive of overall evaluations and voting than were characteristics judged as representing strengths. At the idiographic level, it was found that a trait was significantly more predictive when it fell below an individual's average trait rating for a candidate than when it was located above this mean. Thus, these results show that a negativity effect did characterize impressions of candidates; character weaknesses were more important than strengths in determining the public's evaluations of the candidate and the ultimate vote. Additional analyses were conducted to further investigate negativity in political impressions. It was found that a trait on which the candidates were judged to differ had a larger influence on evaluations than a trait on which the candidates were judged to be similar, but only for the candidate that received the lower rating for the distinctive trait. The expectancy-contrast and cost-orientation explanations for negativity were also examined. No support was found for expectancy-contrast theory, while mixed support was found for cost-orientation. In addition, negativity was found to be more descriptive of impressions of disliked candidates than liked candidates; favorable predispositions toward a candidate reduced negativity.