Liu et al. (2026), Study 1 — LHC26

Authors consider online applications with different strengths (taken to be the average of Z scores for platform-provided metrics, which are: completion rate, number of job completed, work hours, and past earnings). The choice of the participant is whether to rely on an AI or human evaluation for the job application.

Usage

LHC26_S1

Format

A data frame with 207 rows and 6 variables:

choice: [factor] response of participant, either artificial intelligence (AI) or human
strength: [double] average Z score of scales capturing the job application strength
completionrate: [double] job completion rate
jobcompleted: [double] number of job completed
hours: [double] total work hours
earnings: [double] total earnings

References

Liu, Q., Häubl, G. and N. Castelo (2026). Consumers with Weaker Applications Are Less Receptive to Algorithmic Evaluation, Journal of Consumer Research, doi:10.1093/jcr/ucag033