2026-05-21T07:59:17Zhttps://keep.lib.asu.edu/oai/request

oai:keep.lib.asu.edu:node-1995662025-02-19T20:58:47Zoai_pmh:alloai_pmh:repo_items

199566 https://hdl.handle.net/2286/R.2.N.199566 http://rightsstatements.org/vocab/InC/1.0/ All Rights Reserved 2024 61 pages Masters Thesis Academic theses en Betashour, Matthew Powell, Derek Carstensen, Alexandra Duran, Nicholas Arizona State University Partial requirement for: M.S., Arizona State University, 2024 Field of study: Psychology This paper examines how people infer the preferences of others, testing a variety of models including those based in egocentric (self-based) and allocentric (other-based) reasonings. In Study 1, 300 participants rated 60 paintings each to confirm a varied stimulus set to use in Study 2 and create an effective recommender system that could be compared to human performance at the task. In Study 2, 110 participants were asked to predict a target profile’s preferences for 18 paintings after viewing 6 reference painting-rating examples from the target profile. Notably, the top 20% of participants outperformed our machine learning models, despite having access to significantly less information (6 vs 50 reference painting-rating examples). A variety of models were then evaluated to find the best fit for each participant. The findings show that most participants incorporated a mix of ideas of the general population’s preferences, egocentric reasoning, profile-specific allocentric reasoning into their strategies to complete this task. These results provide insight into how people reason through making inferences about other’s preferences. The study also has implications for improving recommender systems and better understanding how humans make social judgments. Cognitive Psychology Statistics cognitive modeling Cue Integration Inference Preference recommender systems Guessing Your Likes: Modeling How Preference Inferences Are Made