Dehumanization and aggression: A meta-analysis

Dehumanization refers to perceiving people as less than fully human and is theoretically argued to contribute to aggression. Till now, a large body of empirical studies have been conducted to test this link directly or indirectly. The current research is to conduct a systematic meta-analysis to obtain a precise estimation of this relationship. More importantly, we aim to examine whether this relationship is further subject to moderators capturing two prominent theories, including the form (i.e., blatant versus subtle, the blatant dehumanization theory) and the type (i.e., animalistic versus mechanical, the dual model) of dehumanization. Based on a total of 184 effect sizes (32 published and non-published papers). The meta-analysis found a moderate overall effect for the association between dehumanization and aggression (r = 0.34, 95 % CI = [0.28, 0.41]). In addition, the association was larger for blatant, compared to subtle (r = 0.39, 95 % CI = [0.30, 0.49] versus r = 0.31, 95 % CI = [0.23, 0.39]) dehumanization, and more robust for dehumanization in both types (animalistic & mechanic, r = 0.38, 95 % CI = [0.24, 0.52]), compared to a single type (animalistic, r = 0.26, 95 % CI = [0.14, 0.37] or mechanic, r = 0.27, 95 % CI = [0.15, 0.40]). Implications have been discussed.