Trumpism, climate and COVID: Social bases of the new science rejection

5.2 Interaction effects Views on climate change and vaccination are both predicted by education×ideology interac- tions in which education just slightly affects the probability of science rejection among extreme conservatives but lowers this probability among all other groups (H4). Fig 2 visualizes these two interactions as marginal probability plots, adjusted for other covariates in the GSEM model of Table 2. Fig 2A shows that, other things being equal, among extremely conservative respondents the probability of rejecting ACC rises slightly with education (from 0.41 to 0.43) but declines with education among all other groups. For example, the conditional probability of ACC rejection declines from 0.35 among moderates with high school education to 0.19 among moderates with college education. Fig 2B similarly shows that the conditional probabil- ity of rejecting COVID vaccination decreases slightly (from 0.36 to 0.30) with education among extreme conservatives but declines more steeply with education among all other groups (for example, from 0.25 to 0.10 among moderates). Similar interactions have been seen in many studies of climate change and other environ- mental topics, so Fig 2A is not surprising. Such interactions involving COVID vaccine rejec- tion (Fig 2B) have not previously been documented, however. A recent study by Shoots- Reinhard et al. [91] reports a complementary result: ideological differences in COVID-19 risk perceptions were most pronounced among individuals with higher verbal (but not numerical)

ability. Individuals with greater verbal ability, like those with higher education, may be more responsive to partisan cueing. Both climate-change and vaccination views exhibit greater partisan divisions among people with mostly same-party friends (friends×party interaction; hypothesis H5). Trumpism also is higher among people with mostly same-party friends. Fig 3 visualizes these statistically Fig 3. Friends×party interaction effects on probability respondent (a) strongly approves of ex-President Trump, (b) rejects reality of anthropogenic climate change, or (c) does not intend to get vaccinated for COVID-19. Marginal probabilities adjusted for other covariates, calculated from the GSEM model in Table 2. https://doi.org/10.1371/journal.pone.0293059.g003 PLOS ONE Trumpism, climate and COVID PLOS ONE | https://doi.org/10.1371/journal.pone.0293059 January 10, 2024 10 / 20