算法规避行为的系统性回顾与展望
Systematic Review and Prospect of Algorithm Aversion
杨勇 1李悦 1连星1
作者信息
- 1. 东北大学秦皇岛分校管理学院,秦皇岛 066004
- 折叠
摘要
算法规避行为是指即使算法提供更为准确的预测建议,人们仍然不愿意依赖算法进行决策的行为.这一概念自2015年提出以来,受到多个领域研究者的重点关注.然而,并没有研究者系统梳理与总结该主题的国内外研究现状.本文运用系统性文献回顾方法,对检索到的81篇国内外文献进行系统总结与分析,阐述算法规避行为的概念内涵和测量方式,从个体特征、算法特征和个体与算法互动过程三个视角梳理其前因,并揭示其间的中介机制和边界条件.针对现有研究的不足,本研究提出算法规避行为未来在量表开发、组织宏观层次影响因素、理论视角和研究情境等方面有价值的研究方向.
Abstract
Algorithm aversion is the behavior of people who are reluctant to rely on algorithms for decision making even when they provide more accurate predictive advice.This concept has received focused attention from researchers in several fields since its introduction in 2015.However,no researcher has systematically combed and summarized the current state of domestic and international research on this topic.In this paper,we apply a systematic literature review method to systematically summarize and analyze 81 retrieved domestic and international literatures to elaborate the conceptual connotation and measurement of algorithm aversion,sort out its antecedents from three perspectives:individual characteristics,algorithmic characteristics,and the process of interaction between the individual and the algorithms,and reveal the mediating mechanisms and boundary conditions between them.Aiming at the shortcomings of existing studies,this study proposes valuable research directions for algorithm aversion in terms of scale development,organizational macro-level influencing factors,theoretical perspectives and research contexts.
关键词
算法管理/算法规避行为/算法欣赏行为/算法决策/系统性文献综述Key words
Algorithm Management/Algorithm Aversion/Algorithm Appreciation/Algorithmic Decision-making/Systematic Literature Review引用本文复制引用
基金项目
国家自然科学基金面上项目(72372021)
国家自然科学基金面上项目(71972033)
出版年
2024
NETL
NSTL
万方数据