Data underlying the paper "Disentangling Fairness Perceptions in Algorithmic Decision-Making: the Effects of Explanations, Human Oversight, and Contestability"

This projects captures fairness perceptions towards algorithmic decision-making processes with varying levels of explanations, human oversight, and contestability. 267 people participated in a crowdsourced study via Prolific. They were shown a loan approval scenario where an individual would get their loan request rejected. Each participant was assigned to a scenario with or without explanation, with or without human oversight, and with or without the right to contest the decision. The dataset includes:<br>The materials used for designing the studyThe preregistration of the studyThe (anonymized) dataThe script used to analyze the data

本项目针对解释程度、人工监督水平及可申诉性各不相同的算法决策流程，采集了公众对其的公平性认知数据。研究依托Prolific平台开展众包实验，共招募267名参与者。实验设定了个体贷款申请遭拒的贷款审批场景，每名参与者被分配至包含/不包含决策解释、具备/不具备人工监督、拥有/不拥有决策申诉权利的实验情境中。本数据集包含：<br>实验设计所用材料、研究预注册文件、（匿名化）原始数据、数据分析所用脚本