Worldviews about change: Their structure and their implications for understanding responses to sustainability, technology, and political change

Abstract

People hold different perspectives about how they think the world is changing or should change. We examined five of these “worldviews” about change: Progress, Golden Age, Endless Cycle, Maintenance, and Balance. In Studies 1–4 (total N = 2733) we established reliable measures of each change worldview, and showed how these help explain when people will support or oppose social change in contexts spanning sustainability, technological innovations, and political elections. In mapping out these relationships we identify how the importance of different change worldviews varies across contexts, with Balance most critical for understanding support for sustainability, Progress/Golden Age important for understanding responses to innovations, and Golden Age uniquely important for preferring Trump/Republicans in the 2016 US election. These relationships were independent of prominent individual differences (e.g., values, political orientation for elections) or context‐specific factors (e.g., self‐reported innovativeness for responses to innovations). Study 5 (N = 2140) examined generalizability in 10 countries/regions spanning five continents, establishing that these worldviews exhibited metric invariance, but with country/region differences in how change worldviews were related to support for sustainability. These findings show that change worldviews can act as a general “lens” people use to help determine whether to support or oppose social change.

The Fill-Mask Association Test (FMAT): Measuring propositions in natural language.

“…Nonetheless, these findings should be interpreted carefully. Unlike previous research testing actual change using historical texts for each year or decade, the present studies used BERT models trained on contemporary texts to test perceived change in a retrospective way (see also Bain et al, 2023, for worldviews about change). Notably, in the present studies, the perceived changes align with those actual changes documented in previous literature; but in some cases, they may not (Mastroianni & Dana, 2022).…”

“…In addition to theoretical implications for the study of attitudes and social cognition and for the understanding of semantic associations in natural language, the current research also contributes to historical psychology in three ways. First, the FMAT detects perceptions of change in a retrospective approach, which complements the cross-temporal approach used to test actual change (Varnum & Grossmann, 2017) by allowing for testing perceived change—an equally important theoretical question in historical psychology (e.g., Bain et al, 2023; Mastroianni & Dana, 2022). Second, the language-modeling approach that the FMAT adopts can be incorporated as an integrative framework to study historical psychology (Atari & Henrich, 2023) for both prevalence change (complementing the word-counting approach) and relationship change (complementing the word-embedding approach).…”

The Fill-Mask Association Test (FMAT): Measuring Propositions in Natural Language

“…Nonetheless, these findings should be interpreted carefully. Unlike previous research testing actual change using historical texts for each year or decade, the current studies used BERT models trained on contemporary texts to test perceived change in a retrospective way (see also Bain et al, 2023 for worldviews about change). Notably, in the current studies, the perceived changes align with those actual changes documented in previous literature; but in some cases, they may not (Mastroianni & Dana, 2022).…”

“…In addition to theoretical implications for the study of attitudes and social cognition and for the understanding of semantic associations in natural language, the current research also contributes to historical psychology in two ways. First, the FMAT detects perceptions of change in a retrospective approach, which complements the cross-temporal approach used to test actual change by allowing for testing perceived changean equally important theoretical question in historical psychology (e.g., Bain et al, 2023;Mastroianni & Dana, 2022). Second, the language-modeling approach that the FMAT adopts can be incorporated as an integrative framework to study historical psychology (Atari & Henrich, 2023) for both prevalence change (complementing the word-counting approach) and relationship change (complementing the word-embedding approach).…”