Microinfluence and Macrodynamics of Opinion Formation: Results From Two Field Experiments and a Natural Experiment

• In Event: 4377 - Computational Sociology

Abstract

Social media platforms, comment boards, and online market places have created unprecedented potential for social influence. Experts and scholars warned that this might reinforce dynamics of opinion polarization. Formal models of social-influence dynamics are powerful tools to describe polarization processes and explore the conditions under which social influence fosters polarization. However, these models’ assumptions about how individuals adjust their opinions as a result of social influence have hardly been tested empirically in online contexts. Here, we propose and calibrate an encompassing model that captures competing micro-level theories of social influence. Conducting a lab-in-the-field experiment with an online polling tool installed on news website, we observed that individual opinions shifted linearly towards the mean of others' opinions. There was no support for negative influence. From this finding on the micro-level, we formally predict the macro-level opinion dynamics resulting from social influence. Next, we tested our macro-predictions using data from another lab-in-the-field experiment. We found that opinion polarization actually decreased in the presence of social influence, in contrast to the notion that online social influence fosters opinion polarization. Finally, we corroborate our findings with large-scale field data from a natural experiment.

Authors

• Michael Maes, University of Groningen

• Bernhard Clemm von Hohenberg, European University Institute, Florence

• Bary Pradelski, ETH Zurich