Can social influence reduce bias in the interpretation of scientific information? Vital science communications are frequently misinterpreted as a result of motivated reasoning, where people misconstrue data on the basis of political and psychological biases. In the case of climate change, studies show that conservatives are much more likely than liberals to inaccurately judge climate data. Researchers attempt to mitigate motivated reasoning by facilitating communication across party lines, but evidence shows that these interventions can exacerbate bias. In this study, we provide a method for facilitating cross-party communication that eliminates biased interpretations of climate data among conservatives, while also improving the interpretations of liberals. In an online experiment, we placed conservatives and liberals in decentralized networks where they exchanged information while interpreting climate data. We find that networked peer influence drastically improves the accuracy of interpretations among both conservatives and liberals, compared to control subjects who interpreted the same data on their own. These results show how engineered network interventions are an effective tool for mitigating motivated reasoning in the science of science communication.
- 2018 Best Paper Award, International Conference on Computational Social Science