UC San Diego
Modeling Attitude Change and Cognitive Bias for Social Issues
- Author(s): Bardolph, Megan
- Advisor(s): Coulson, Seana
- et al.
Polarization is increasing in American politics and frequently involves disagreement over basic facts. Although this phenomenon is rare, the domain of social and political issues provides an environment where emotions may influence opinions and decisions in seemingly irrational ways. This research explores how individuals judge and respond to evidence about controversial socio-political issues, considering whether behavior is more appropriately modeled by accounts of motivated reasoning or Bayesian updating rules.
We find evidence of an attitude congruency bias, where people judge information to be of higher quality when it aligns with their existing attitudes on an issue. However, we find that this bias does not necessarily lead to polarizing. Instead, people’s change in attitudes is better described by Bayesian updating, where people are sensitive to the amount and quality of information they are presented with. This behavior does not seem to be driven by affect or knowledge in a domain.
Judgment of information, in the form of argument rating, was analyzed by creating separate measures of objective argument quality and individual rating bias. Both factors were found to model attitude change, indicating that participants are sensitive to both the objective quality of evidence and to the effects of their own biases. Accounts of Bayesian information processing and motivated reasoning both predict behavior when information is modeled in terms of these two factors.
Finally, this research explores the role of information choice and shows that a bias toward choosing attitude-congruent information may lead to motivated attitude change, where exposure to a biased set of evidence models attitude change in line with one’s existing views. People exhibit more sensitivity to information quality when they do not choose which information to view, indicating that choice may play a special role in allowing individual bias to outweigh information quality. These findings inspire questions about the role of curated information and people’s capacity for rational behavior in a tense political climate.