Pre-Analysis Plan for Preferences on the use and regulation of algorithms in the context of self-driving cars and ADM

As part of a bigger research project on the social embeddedness of autonomous systems, this study explores attitudes and preferences of individuals regarding the use and regulation of algorithms and algorithmic systems, using original data from representative online surveys administered in Germany, Japan and the United States. The first part of the study uses a conjoint experiment to provide evidence for the effect of multidimensional regulatory designs on mass support for the regulation of self-driving cars, and test the relative importance of different regulatory features. In a second step, we conduct a framing-experiment to understand the effect of different policy-frames on regulatory acceptance for self-driving cars. To explore the circumstances under which individuals accept algorithmic decision-making over human decision-making, third, we will analyze the societal acceptance of ADM in the context of hiring decisions, using a factorial vignette-experiment.本研究作为自主系统社会嵌入性（social embeddedness）大型研究项目的子课题，依托在德国、日本及美国开展的代表性在线调研所获原始数据，探究个体对算法及算法系统的使用与监管所持的态度与偏好。研究第一部分采用联合实验（conjoint experiment），验证多维度监管设计对大众支持自动驾驶汽车监管的影响，并检验不同监管特征的相对重要性。第二部分，研究开展框架实验，探究不同政策框架对自动驾驶汽车监管接受度的影响。为探究个体相较于人类决策更倾向于接受算法决策的情境条件，本研究第三部分将采用因子情景实验（factorial vignette-experiment），分析招聘场景下算法决策（ADM, Algorithmic Decision-Making）的社会接受度。