UC Davis

The uncanny valley phenomenon describes discomfort from imperfectly human stimuli. I conducted a meta-analysis of 70 empirical studies to synthesize knowledge, finding a significant combined effect (d = .93) and confirming the phenomenon's polythetic nature. Five hypothesized mechanisms (e.g., categorization ambiguity) significantly impacted four dependent variables (i.e., affect, cognition, behavior, and attitudes). Additionally, moderation analysis identified influence from participants’ age and sex— a first in uncanny valley review. However, hypothesized mechanisms were found underexplored. To address such, my subsequent study (N = 188) tested the doppelgänger hypothesis guided by self-reference theory, and explored a third media exposure variable to account for potential age and sex confounds— participants' science fiction and technology affinity (SF&T affinity). Contrary to expectations, self-face morphs elicited positive self-biases for self-oriented outcomes. Congruent with expectations, participants' SF&T affinity attenuated uncanny effects. Instead of the “uncanny valley” predicted by Mori (1970), a consistent growth relationship dubbed “the anthropomorphic escalator” implied reliable positive effects of stimuli’s anthropomorphism. My research provides theoretical and methodological expansion of the uncanny valley phenomenon beyond Mori (1970) in intellectual contribution to the Communication field.