A study of ‘imperfect robot strategies’ and ‘tolerance window models’ in child-robot interaction (CRI).

These data aim to investigate the impact of robotic errors (imperfect behaviour) on children’s emotional bonding, trust-building and psychological mechanisms. The study challenges the traditional notion that ‘only perfect performance constitutes a positive experience’, seeking instead to identify an ‘optimal window of imperfection’—that is, the frequency and manner in which robots should make mistakes in order to enhance their appeal without undermining children’s trust.