Is Algorithmic Accessibility Sufficient? The Pivotal Role of Accessibility and Accountability in Shaping Trust in Automated Decision‐Making
通过两个基于真实场景的调查实验，研究了算法可及性和问责制如何影响人们对自动化决策的信任，发现问责制比可及性更重要，且与决策风险共同影响信任。
ABSTRACT The widely accepted view that algorithmic accessibility and accountability enhance the perceived trustworthiness of automated decision‐making has not yet been fully validated. This study fills this research gap by conducting two survey experiments using “real‐world” scenarios. We examine participants' trust preferences when exposed to varying levels of algorithmic accessibility and accountability, under different degrees of decision risks. Our findings highlight that algorithmic accountability exerts a stronger influence on the perceived trustworthiness of both algorithms and bureaucrats compared to algorithmic accessibility. Additionally, our research uncovers that the interaction between algorithmic accountability and contextual factors (e.g., high vs. low‐risk scenarios) significantly affects trust in algorithm. This study offers novel empirical insights into the intricate dynamics among algorithmic accessibility, accountability, decision risk, and trust perceptions.