Eliciting an in-group/out-group facial recognition bias based upon falsified, trivial groupings

Facial recognition biases favouring one‚ÄövÑv¥s own group and impacting social interactions have been seen for race, gender, age and recently, experimentally created groups. This study tests if allocation into correctly constructed minimal groups elicits an own group bias and overcomes the own gender bias (OGB). 166 participants (age M = 31.50, SD = 12.35) identifying as female were randomly allocated into the control or red/green experimental group. Experimental groups were mis-informed allocation was based upon personal values. The learning phase presented 20 faces (10 male, 10 red). Following a filler task, 40 faces (20 old) were judged old or new. The minimal group manipulation did not significantly effect recognition between in and out-groups (p = 0.307). Nor did target gender produce an OGB in control participants (p = 0.177). A small male face advantage in the experimental group (d = 0.22, p = 0.026) and a significant interaction between target group, target gender and participant group (p <.001) were deemed uninterpretable. Results indicate allocation into experimentally constructed minimal groups does not elicit an own group facial recognition bias; replication with a manipulation check is recommended. The mixed OGB findings questions gender as an automatic basis for categorisation in modern contexts.