算法厌恶:理论框架及研究展望
Algorithm Aversion:Theoretical framework and research prospects
张雯 1肖增瑞 2吕明杰3
作者信息
- 1. 浙江树人学院管理学院
- 2. 浙江理工大学国际教育学院
- 3. 之江实验室智能社会治理实验室
- 折叠
摘要
研究表明,尽管算法通常比人类表现更好,在算法与人类之间,人们还是选择人类提供服务,这种现象称为算法厌恶,得到了管理学、心理学等研究领域的高度关注.然而,关于算法厌恶的实证研究对于算法厌恶产生的前因和内在机制尚无定论,部分研究发现了算法欣赏的相反结果.通过对算法厌恶的实证研究进行文献回顾,本研究归纳与总结了影响算法厌恶的算法特征、个体特征和任务特征,算法厌恶产生的认知机制和情感机制,以及深层的理论解释.在此基础上,本研究从算法厌恶的边界条件扩充、内在机制挖掘、干预策略研究和引发后果探索四个方面为未来的研究提供了建议.
Abstract
Research shows that algorithms usually outperform human.Yet when people are deciding whether to use a human agent or an algorithm,they often choose the human agent.This phenomenon,which we call algorithm aversion,has drawn intense attention from management,psychology and other research fields.However,empirical studies on the antecedents and mechanisms of algorithm aversion are inconclusive and papers find the opposite effect of algorithm appreciation.Through a literature review of the empirical studies,this study shows the effects of algorithm characteristics,individual characteristics and task characteristics on algorithm aversion,and the cognitive and affective mechanisms as well as the theoretical explanation underlying algorithm aversion.On this basis,we provide suggestions for future research from four aspects:the boundary conditions,the internal mechanism,the intervention strategies and the consequences of algorithm aversion.
关键词
算法/算法厌恶/算法欣赏/人工智能Key words
algorithm/algorithm aversion/algorithm appreciation/artificial intelligence引用本文复制引用
基金项目
国家自然科学基金项目(72101233)
深圳市科创委基础研究项目(RKX20220705152810024)
浙江省自然科学基金项目(LQ22G020006)
出版年
2024
NETL
NSTL
万方数据