An Empirical Study of Social Media Exchanges about a Controversial Topic: Confirmation Bias and Participant Characteristics

Abstract

There has been a significant amount of research into social media commentary influences on human behaviors, ranging from its role in affecting political elections to predicting corporate revenues; however, to this point, the factors and influences of social media have not been completely explained and it is not entirely clear whether social media influences or simply confirms preconceptions. Moreover, with sentiment analysis, much of the research has relied on human expert interpretation of the sentiments and semantics written in various social media. It has also tended to be interpretive rather than predictive in nature. In our study, we wanted to know if social media conversations were reflective or influencers of human behavior. Using a social media mining technology we were able to determine sentiments, sentiment intensity, and the characteristics of participants. We found strong evidence of confirmation bias, but that bias was influenced by personal characteristics, and in some cases, whether the sentiments were strongly positive or strongly negative.