Climate change attitudes have been explained with a variety of factors, including cultural worldviews, environmental attitudes, political ideology, knowledge of climate change, severe weather exposure, and sociodemographic characteristics. These approaches, however, assume that an individual forms attitudes on the basis of preexisting values or beliefs and do not account for dynamic social interaction as a source of influence. This study introduces a network perspective that accounts for the social embeddedness of individuals, using key relational and structural network variables to predict climate attitudes.

This study explores the relationships between climate attitudes and the following social network variables: homophily, network strength, attitude diversity, centrality, network size, and network valence. Climate change attitudes were developed from an exploratory factor analysis, which identified climate change skepticism and belief strength as distinct attitudinal dimensions. Using egocentric data from a nationally representative Knowledge Networks survey collected for the Yale Project on Climate Change Communication in 2011, this study found that network variables were significant in predicting both climate attitude dimensions. Using hierarchical regression analyses that accounted for other known predictors of climate attitudes, this study determined that homophily, network strength, attitude diversity, and network valence were significant predictors of skepticism (R2change=4.8%), while centrality and network strength were predictive of belief strength (R 2change=8.9%). A multinomial logistic regression also determined that strong accepters have the most active communication styles, exhibiting the highest homophily, network strength, and centrality in their personal networks. Directions for future network research and practical applications are discussed.