中国安全科学学报 ›› 2021, Vol. 31 ›› Issue (6): 37-43.doi: 10.16265/j.cnki.issn 1003-3033.2021.06.005

• 安全社会科学与安全管理 • 上一篇    下一篇

死亡悖论视角下群体性主观概率偏差仿真研究

张羽 讲师, 谯丽, 岑康 教授   

  1. 西南石油大学 土木工程与测绘学院,四川 成都 610500
  • 收稿日期:2021-03-05 修回日期:2021-05-11 出版日期:2021-06-28 发布日期:2021-12-28
  • 作者简介:张羽 (1986—),男,江西新余人,博士,讲师,硕士生导师,主要从事安全知识与行为、安全监管博弈等方面的研究。E-mail: 307871884@qq.com。
  • 基金资助:
    教育部人文社会科学青年基金资助(14XJCZH004);教育部产学合作协同育人项目(201802339011)。

Simulation research on group subjective probability deviationfrom a perspective of death paradox

ZHANG Yu, QIAO Li, CEN Kang   

  1. School of Civil Engineering and Geomatics, Southwest Petroleum University, Chengdu Sichuan 610500, China
  • Received:2021-03-05 Revised:2021-05-11 Online:2021-06-28 Published:2021-12-28

摘要: 为探究安全生产过程中群体性主观概率偏差的形成与演化机制,构建个体主观概率元胞自动机(CA)模型,运用Netlogo仿真平台进行仿真试验;先对比分析仿真图形,再运用层次回归法分析试验所得数据。研究结果表明:群体主观概率演化过程分为激变、波动、稳定3个时期;个体死亡会导致群体的主观概率出现系统性低估,且会加剧群体内部的主观概率差异;客观事故概率的减少会增大群体内部主观概率的差异;群体人数增加有助于减少主观概率偏差和差异,促使群体主观概率估计更客观、更统一。

关键词: 死亡悖论, 主观概率, 元胞自动机(CA), 群体性偏差, 层次回归分析

Abstract: In order to study formation and evolution of group subjective probability deviation in the process of safe production, a CA model of individual subjective probability was constructed and simulation experiments were conducted by using Netlogo simulation platform. Then, simulated graphics were compared and analyzed, and hierarchical regression method was used to analyze experiment data. The results show that evolution process of subjective probability of a group is divided into three periods, namely sudden change, fluctuation, and stability. Individual death will lead to a systematic underestimation of subjective probability, and aggravate its difference within the group. Reduction of objective accident probability will also increase difference in subjective probability. However, an increase in the number of group members helps to reduce such deviation and difference, and promotes a more objective and unified estimation of subjective probability of a group.

Key words: death paradox, subjective probability, cellular automata (CA), group bias, hierarchical regression analysis

中图分类号: