Liu et al. (2026), Study 3 — LHC26

Participants performed in a online game where they had to find fruits given a set of three boxes, and would get a score for correctly detecting the box containing the fruit. Unknown to them, the game was rigged and they were given, along with their final score, a measure of how other people performed, to inform them whether they performed better or worst than other participants. Half of the participants were supposed to receive a bonus based on strong performance, as judged by algorithmic and human judges: they had t choose which among the two. The mediator is a measure of relative preference for their choice of judge.

Usage

LHC26_S3

Format

A data frame with 996 rows and 5 variables:

choice: [factor] first choice for the judge, either ai or human
mediator: [double] strength of preference (from 10 to zero) for the given choice
strength: [factor] strength of the applicant, as given by the score relative to the reported average, either high or low
gender: [factor] gender of participant, either male or female
age: [integer] age of participant

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