Materialien und Daten zum Research Paper "Artificial empathy in healthcare chatbots: Does it feel authentic?"

The dataset contains data and materials that have been created and collected as empirical basis for the first paper "Artificial empathy in healthcare chatbots: Does it feel authentic?" of the cumulative dissertation "Social Actor or Technology? Experimental Studies on the Perception of Chatbots Versus Humans and Their Implications for Anthropomorphic Chatbot Design". The researcher conducted two experimental online studies in which participants took the position of a sick person before interacting with a diagnostic chatbot programmed for the purpose of this paper. Participants were randomly assigned to one of four different diagnostic chatbots whose communication styles were designed in congruence with established empathy theories and related research (empathetic: feeling with; sympathetic: feeling for; behavioral-empathetic: empathetic helping; non-empathetic: control condition). Results from parallel mediation analyses revealed that the positive effect of empathy on trust and using intentions mediated by an increased perceived warmth is attenuated by a simultaneous loss in perceived authenticity for the chatbots utilizing experiential expressions of empathy. The effect occurred independently of whether the bot showed personifying elements (Study 2) or not (Study 1). A third study did not replicate the backfiring effect when participants watched a chat between a patient and a human physician, i.e., experiential expressions were only inauthentic in bots, not in humans.