Third–Party Responses to Unfair Decisions by Artificial Intelligence versus Humans

There are two datasets within the zip file.The experimental design of study1 was a 2(group: human vs AI)*6 (unfairness: 100:0, 90:10, 80:20, 70:30, 60:40, 50:50) design, with group as the between-subjects variable, unfairness as the within-subjects variable, and third-party punishment (TP) as the dependent variable. For easier understanding of the results, we combined the six levels of unfairness into three levels: 100-0, 90-10 allocation defined as high unfairness situations (corresponding to TP_high), and 80-20, 70-30 allocation combinations as median unfairness situations (corresponding to TP_median), and the 60-40, 50-50 allocation were defined as low unfairness situations (corresponding to TP_low). In addition gender, age, education level were collected as control variables.The experimental design of study2 was identical to study1, with the addition of TPoutrage_low, TPoutrage_median, and TPoutrage_high as the scores of the mediating variable moral outrage in the three unfairness situations. Responsibility attribution (TPFAE) scores were also measured, calculated from TPFAE_in-TPFAE_ex.