Replicating and extending Sengupta et al. (2023): Contact predicts no within-person longitudinal outgroup-bias change.

Abstract

Intergroup contact has long been touted as a premier means to reduce prejudice and forge positive bonds with outgroups. Given its origins in psychological research, it is perhaps of little surprise that contact is expected to induce change within people over time. Yet using random-intercepts crossed-lagged modeling that parses within-person from between-person effects, Sengupta et al. ( 2023) recently found no evidence of within-person change, only unexplained between-person effects, regarding contact's effects on outgroup solidarity in New Zealand. We conceptually replicated their study, focusing on modern racism and an affect thermometer as the outcomes, in a three-wave study of White British participants (N T1 = 946, N T2 = 667, N T3 = 591) and their attitudes toward foreigners. We replicated the general pattern described by Sengupta and colleagues, confirming between-person effects without within-person effects, suggestive of third-variable explanations. As a novel finding, we discover that differences in social dominance orientation (SDO) and right-wing authoritarianism (RWA) can account for the observed between-person effects. Problematically for contact theory, contact effects, at least those relying on self-reported accounts, increasingly appear to reflect differences between people (person factors) rather than being context-driven (situation factors)-such that those lower (vs. higher) in SDO and RWA are more favorable toward outgroups, rather than intergroup contact bringing about positive outcomes itself. Implications for theory development and intervention are discussed.

Does intergroup contact affect political attitudes and behaviours? - A longitudinal mediation analysis using the Chilean Longitudinal Social Survey (ELSOC)

“…participating in deliberative exchanges with political antagonists vs. a naturalistic measurement of a potentially stable level of intergroup contact over time) and many more differences in study characteristics. These differences point towards potential moderating effects of tertiary transfer effects, such as that tertiary transfer effects might be more likely to be to be observed in controlled Moreover, some features of our study might also explain the absence of statistically significant effects: For one, there is a growing list of publications on primary transfer effects of intergroup contact which indicate that random-intercept cross-lagged panel models rarely show within-person processes in line with the contact hypothesis (e.g., Bohrer et al, 2019;Friehs et al, 2024;Hodson & Meleady, 2024;Sengupta et al, 2023; but see also Górska & Tausch, 2023). Still, random-intercept cross-lagged panel models are highly sophisticated analytical models which allow for the separation of within-person and between-person variance in longitudinal data, and consequently, they are more informative in terms of the nature of the investigated process and allow for a stronger causal interpretation of cross-lagged effects compared to conventional crosslagged panel models (Hamaker et al, 2015;Lucas, 2023;Mulder & Hamaker, 2021).…”