Dehumanization after all: Distinguishing intergroup evalutation from trait-based dehumanization

Over the last two decades, a plethora of research using a large variety of measures and paradigms has demonstrated how people often tend to dehumanize members of certain outgroups. When trait-based measures were used, ingroup members showed to attribute less human positive and negative characteristics to members of the outgroup. Enock, Flavell, Tipper, and Over (2021) recently criticized these findings stating that previous work on this topic has investigated the attribution of mostly socially desirable human traits making it impossible to determine that trait-based dehumanization is distinct from intergroup preferences. In a set of studies in which they balanced the human and non-human traits for valence, they did not find any evidence for outgroup dehumanization. Using the same experimental material, the current work conceptually replicated a subsample of these studies in two pre-registered experiments introducing a more parsimonious measure of dehumanization that is based on a large number of traits and takes both the variance in typicality and humanness judgments into account, at the same time allowing for a rigid control of the valence of the traits. Results clearly indicated the presence of a dehumanization effect in both studies over and above intergroup evaluations. As such, these results are in line with previous work on dehumanization and highlight the risks of gauging dehumanization through the attribution of a small number of fixed human traits that are not controlled for their desirability.