Data for: How Much Weight Does an AI Score Carry in Hiring? Qualification Conflict and Evaluator Heterogeneity in a Choice-Based Conjoint Experiment

This dataset contains the anonymized participant-level data from a choice-based conjoint (CBC) experiment investigating how much weight an AI-generated employability score carries in hiring decisions, relative to a formally required qualification and biographical candidate attributes.In the study, 432 participants each completed eight forced-choice tasks, choosing among four hypothetical candidate profiles (or an explicit "none" option). The profiles varied on five attributes: an AI-generated employability score (7–91%), a formally required educational qualification, gender, age, and hobby. Each choice was followed by a second-stage hiring-commitment question (whether the participant would actually advance the chosen candidate). The data support analyses of attribute weighting, qualification–score conflict, behavioral override of the qualification, and evaluator heterogeneity via latent-class modeling.The Excel workbook contains four sheets: Respondents — demographic and self-report items for each participant;Individual preferences — individual-level hierarchical Bayes part-worth utilities;Model matrix — the long-format experimental design with the chosen alternative per task;Very raw responses — task-level records including the second-stage hiring-commitment response. Privacy and anonymization: This is an anonymized public version. All direct and indirect identifiers collected by the survey platform have been removed prior to deposit. Participant identifiers were replaced with sequential numeric pseudonyms; IP addresses, geolocation, date/time stamps, browser and device strings, session identifiers, process-tracking telemetry, free-text responses, and the student index number used for course-credit assignment have all been removed. The data contain no directly identifying information. Use is subject to the condition that no attempt be made to re-identify participants.This dataset accompanies the manuscript "How Much Weight Does an AI Score Carry in Hiring? Qualification Conflict and Evaluator Heterogeneity in a Choice-Based Conjoint Experiment" (Petrović, Kuzmanović, Anđelković Labrović, Kovačević, & Llorens). The R code reproducing all reported analyses is available from the corresponding author on reasonable request.