ISSN 1671-3710
CN 11-4766/R

心理科学进展 ›› 2023, Vol. 31 ›› Issue (1): 60-77.doi: 10.3724/SP.J.1042.2023.00060

• 研究前沿 • 上一篇    下一篇


谢才凤1, 邬家骅1, 许丽颖2(), 喻丰1(), 张语嫣1, 谢莹莹3   

  1. 1武汉大学哲学学院心理学系, 武汉 430072
    2清华大学马克思主义学院, 北京 100084
    3灵山县那隆镇中心校, 广西 钦州 535414
  • 收稿日期:2022-02-19 出版日期:2023-01-15 发布日期:2022-10-13
  • 通讯作者: 许丽颖,喻丰;
  • 基金资助:

The process motivation model of algorithmic decision-making approach and avoidance

XIE Caifeng1, WU Jiahua1, XU Liying2(), YU Feng1(), ZHAND Yuyan1, XIE Yingying3   

  1. 1Department of Psychology, School of Philosophy, Wuhan University, Wuhan 430072, China
    2School of Social Marxism, Tsinghua University, Beijing 100084, China
    3Lingshan County, Nalong Town, Central School, Qinzhou 535414, China
  • Received:2022-02-19 Online:2023-01-15 Published:2022-10-13
  • Contact: XU Liying,YU Feng;


算法常用于决策, 但相较于由人类做出的决策, 即便内容相同, 算法决策更容易引起个体反应的分化, 此即算法决策趋避。趋近指个体认为算法的决策比人类的更加公平、含有更少的偏见和歧视、也更能信任和接受, 回避则与之相反。算法决策趋避的过程动机理论用以解释趋避现象, 归纳了人与算法交互所经历的原初行为互动、建立类社会关系和形成身份认同三个阶段, 阐述了各阶段中认知、关系和存在三种动机引发个体的趋避反应。未来研究可着眼于人性化知觉、群际感知对算法决策趋避的影响, 并以更社会性的视角来探究算法决策趋避的逆转过程和其他可能的心理动机。

关键词: 算法决策, 人类决策, 决策趋避, 心理动机, 人?算法交互


With the advantages of objectivity, accuracy, high speed and low cost, algorithmic decision-making has been widely used in human daily life, such as medical, judicial, recruitment and transportation situations. How will people react to the shift from traditional human decision-making to the newly emerged algorithmic decision-making? If people perceive algorithms as social actors, there would be no difference when faced with the same decision made by two different agents. However, researches show that algorithmic decision-making is more related to different responses in individuals than human decision-making on the same content. In other words, people will approach or avoid the same algorithmic decision-making, which is defined as the algorithmic decision-making approach and avoidance. Specifically, the algorithmic decision-making approach means that algorithmic decision-making is considered fairer, less biased, less discriminatory, more trustworthy, and more acceptable than human decision-making. But the algorithmic decision-making avoidance is the other way around. By analogy with the distinct ideologies when facing outgroup members, the process motivation model of algorithmic decision-making approach and avoidance simulates human psychological motivation when facing the same decisions made by algorithms and humans. Based on the premise that quasi-social interaction (relationship) and interpersonal interaction (relationship) develop parallel, the theory summarizes the three interaction stages between humans and algorithms. Namely, the interaction of initial behavior, the establishment of quasi-social relationships and the formation of identity. Furthermore, it elaborates how cognitional, relational, and existential motivation trigger individual approach and avoidance responses in each specific stage. More precisely, it occurs to meet the cognitive motivational needs to reduce uncertainty, complexity, and ambiguity in the interaction of the initial behavior stage, fulfill the relational motivational needs for establishing belonging and social identity in the establishment of the quasi-social relationship stage, and to satisfy the motivational needs for coping with threats and seeking security in the identity formation stage. In accordance with the three psychological motivations of cognition, relationship, and existence, the process motivational theory introduces six influencing factors, such as cognitive load, decision transparency, moral status, interpersonal interaction, reality threat and identity threat respectively. For future directions, we suggest that more researches are needed to explore how mind perception and intergroup perception influence algorithmic decision-making approach and avoidance. Meanwhile, what is the reversal process of the algorithmic decision-making approach and avoidance from a social perspective and what other possible motivations are associated with it are also worthy of consideration.

Key words: algorithmic decision-making, human decision-making, decision-making approach and avoidance, mental motivation, human-algorithm interaction