Fig. 3: Tertiary analysis.
From: Advancing the psychology of social class with large-scale replications in four countries
Interaction between each social class indicator and moderator in the pooled sample. Each panel of this figure displays coefficient estimates from fixed-effects regression models (with country fixed effects) testing interaction effects between a specific social class indicator and social class identification (left), system-justification beliefs (middle) and local income inequality (right) on a given outcome (excluding the original control variables for comparability). Counts of significant strengthening interactions, significant weakening interactions, and non-significant interactions appear on the right of each panel, with the largest count in bold. H12′–H14′, H25′, H16′, H27′–H28′ and H32′–H33′ were not tested because these hypotheses do not involve main effects of social class. Supplementary Table 12 provides the estimates for the 23 (hypotheses) × 9 (indicators) × 3 (moderators) = 621 interactions, along with the simple slopes. For the hypotheses marked with a superscript ‘a’, the statistical effect of social class observed in the secondary analysis was in the opposite direction to the original hypothesis for the majority of indicators; thus, the interactions refer to a strengthening or weakening effect of the observed social class effect in the secondary analysis rather than to the original hypothesized effect.
