Norm-Breaking Responses to Sexist Abuse: A Cross-Cultural Human Robot Interaction Study
summary
A 3x3 between-subjects video experiment with 225 participants across the US, Sweden and Japan tests three robot retorts (argument-based, aggression-based, and a Siri-like default) using an English-dubbed stimulus and Japanese subtitles. The main claim is replicated: the argument-based response scores higher credibility than control across locations and genders, yet there is no pre/post change in gender bias or interest in robotics. Observers mostly pick the non-apologetic empathetic chosen by majority, with no Swedish or US women choosing apologetic replies while many Japanese women do.
AI-written summary · about