A Causal Role for Negative Affect: Misattribution in Biased Evaluations of Scientific Information

Abstract

The role of negative affect as a potential causal mechanism underlying the tendency for prior attitudes to bias evaluations of scientific information was investigated using a misattribution of affect manipulation. Participants read scientific studies that disconfirmed prior beliefs. Half the participants were given the opportunity to misattribute any negative affect to poor room conditions (Study 1) or caffeinated water (Study 2). All participants then evaluated the methodological quality of the scientific information. Those in the misattribution condition evaluated the information more positively than those in the control condition. In Study 2, no differences were found for participants reading confirming information. The role of affect in attitude resistance is discussed.

Keywords:

• Affect
• Attitude resistance
• Misattribution

Acknowledgments

We would like to thank Carrie Fried for supplying the misattribution materials used in Fried and Aronson (1995), which were adapted for the current research.

Notes

1. Participants with no attitudinal leanings (i.e., those circling the scale midpoint on the attitude assessment) were not recruited because the hypotheses apply only to people with pre-existing attitudes. Also, we had planned to recruit approximately half of the participants from those indicating unaccepting attitudes toward homosexuality. However, there was only a very small number of students with unaccepting attitudes.

2. All analyses relevant to the hypotheses were also conducted with initial attitudes and beliefs included as covariates. In terms of statistical significance, none of these covariate analyses yielded different results than the analyses reported that did not include covariates.

3. There were somewhat more pro-choice (N = 68) than pro-life (N = 41) participants in our final sample. However, when divided into the eight cells created by crossing the misattribution manipulation, the match variable, and the abortion attitude variable, there were at least 10 participants in every cell.