Who is the Expert? Reconciling Algorithm Aversion and Algorithm Appreciation in AI-Supported Decision Making

The increased use of algorithms to support decision making raises questions about whether people prefer algorithmic or human input when decisions. Two streams research on algorithm aversion and appreciation have yielded contradicting results. Our work attempts reconcile these contradictory findings by focusing the framings humans as a mechanism. In three experiments, we created an result (Experiment 1) well 2) manipulating only description agent agent, demonstrated how different choices can lead inconsistent outcomes in previous studies 3). We also showed that results were mediated agent's perceived competence, i.e., expert power. provide insights into divergence literature. hope shift attention from two phenomena better design framing algorithms. call community theory power sources, it is systemic framework open up new possibilities for designing systems.