The Neural Correlates of Ethnic and Gender Similarity and Dissimilarity with Avatars in Online Shopping Environments

Virtual sales assistants with human-like interfaces (“avatars”) help consumers make better choices with less cognitive effort in online shopping environments. Following similarity-attraction theory, such avatars must be similar to consumers’ ethnicity and gender to be liked and used. In contrast, dissimilarity-repulsion theory posits that dissimilarity causes consumers to avoid avatars that mismatch their ethnicity and gender. This study tests these two competing theories by identifying the neural correlates of similarity/dissimilarity using functional Magnetic Resonance Imaging (fMRI) while consumers assess their similarity/dissimilarity with avatars who vary in their ethnicity and gender to match/mismatch the consumers’ ethnicity and gender. The fMRI results show activation in a brain area associated with utility (caudate nucleus) for similarity, only in men and mainly due to ethnic similarity. In contrast, dissimilarity spawns activity in brain areas linked to negative emotions (amygdala, insular cortex) and social prejudice (Broadmann Areas 9 and 32), only in women and mainly due to gender dissimilarity. While both men and women ultimately choose “matched” avatars, men do so because of utility derived from ethnic similarity, while women because they perceive emotionally-charged social anxiety from gender dissimilarity.