Climate change belief systems across political groups in the United States

have a greater impact on them, as there is more opportunity to insert new belief elements and/ or strengthen existing connections between elements. Network density analysis can then help communicators optimally allocate communication resources. Providing structured informa- tion about a topic to those with less organized belief systems can increase their belief system density, while individuals with high-density systems may prove less receptive. The density of belief systems has been examined in political contexts. Scholars have found that politically informed individuals tend to have denser political belief systems than do politi- cally apathetic individuals [8, 14, 15]. This could potentially explain why political elites consis- tently demonstrate relatively well-defined and coherent opinions on a range of political issues [8, 9]. A similar pattern can be observed with climate change belief systems, especially in the U.S., where climate change opinions have become deeply intertwined with political ideologies [16, 17]. As a result, individuals with strong political affiliations are likely to have denser cli- mate change belief systems than those who are politically unaffiliated. The structure of a network shapes the flow of information within it. Depending on this structure, a shift in a single belief could either rapidly or only gradually propagate, influenc- ing either the entire system or just a portion of it. Comparing structures within and between political groups can also illuminate the organization of people’s beliefs, showing how alike

or different they are. Although the concept of density offers a perspective on the structure of belief systems, it falls short of encapsulating the intricate architecture of a network, as it merely reflects the overall degree of connection. In this study, we examine the heterogeneity of belief systems within and between political groups using a network metric (i.e., graph dif- fusion distance) to quantify the heterogeneity between two networks. It is noteworthy that structural variations pertain to the difference in belief organization rather than the differ- ence in the level of belief. Therefore, although Republicans generally have lower climate change beliefs than Democrats, the way their beliefs are organized could nonetheless be similar. In sum, this study has three primary objectives: a) to examine the centrality of different cli- mate beliefs, b) to analyze how the network density of climate change belief systems changes across political groups over time, and c) to compare the network structures of belief systems both within and between political groups. For data, we use fifteen waves of nationally represen- tative cross-sectional survey data collected from 2010 to 2021. As these survey datasets are nationally representative, combining these cross-sectional surveys allows us to observe any sys- tematic structural changes in climate change belief systems over time in the U.S. Methods Survey design and samples This study utilized multiple waves of nationally representative cross-sectional surveys collected

by the Yale Program on Climate Change Communication (YPCCC) and the George Mason University Center for Climate Change Communication (Mason 4C). The original survey data- sets were collected under an exemption granted by the Institutional Review Board (IRB) of Yale University (IRB Protocol ID: 2000031972). Subsets of de-identified datasets were accessed for the purpose of this research on September 20, 2022. After launching an initial survey in 2008, YPCCC and Mason 4C have conducted a repre- sentative survey on climate change opinions twice every year since 2010. The researchers obtained a distinct sample for each survey from Ipsos KnowledgePanel of U.S. adults aged 18 and above, recruited using probability sampling and representative of the country’s popula- tion. The panel includes individuals recruited through different methods, such as random digit dialing and address-based sampling, covering nearly all U.S. residential phone numbers PLOS ONE Climate change belief system PLOS ONE | https://doi.org/10.1371/journal.pone.0300048 March 20, 2024 3 / 16