ISSN 1671-3710
CN 11-4766/R
主办:中国科学院心理研究所
出版:科学出版社

心理科学进展 ›› 2020, Vol. 28 ›› Issue (1): 178-190.doi: 10.3724/SP.J.1042.2020.00178

• 研究方法 • 上一篇    下一篇

变量间的网络分析模型及其应用

蔡玉清1(), 董书阳2, 袁帅3, 胡传鹏4,5()   

  1. 1 清华大学人文学院, 北京 100084
    2 Department of Developmental Psychology, Utrecht University, Utrecht,3584CS, Netherland
    3 Department of Methodology and Statistics, Tilburg University, Tilburg, 5037AB, Netherland
    4 German Resilience Center, Mainz, 55131, Germany
    5 Neuroimaging Center (NIC), University Medical Centre of the Johannes Gutenberg University, Mainz, 55131, Germany
  • 收稿日期:2019-05-27 出版日期:2020-01-15 发布日期:2019-11-21
  • 通讯作者: 蔡玉清,胡传鹏 E-mail:cyq_2016@outlook.com;hcp4715@hotmail.com
  • 基金资助:
    清华大学学术推进计划资助

Network analysis and its applications in psychology

CAI Yuqing1(), DONG Shuyang2, YUAN Shuai3, HU Chuan-Peng4,5()   

  1. 1 Department of Humanity, Tsinghua University, Beijing 100084, China
    2 Department of Developmental Psychology, Utrecht University, Utrecht,3584CS, Netherland
    3 Department of Methodology and Statistics, Tilburg University, Tilburg, 5037AB, Netherland
    4 German Resilience Center, Mainz, 55131, Germany
    5 Neuroimaging Center (NIC), University Medical Centre of the Johannes Gutenberg University, Mainz, 55131, Germany
  • Received:2019-05-27 Online:2020-01-15 Published:2019-11-21
  • Contact: CAI Yuqing,HU Chuan-Peng E-mail:cyq_2016@outlook.com;hcp4715@hotmail.com

摘要:

变量间的网络分析模型近年来被广泛应用于心理学研究。不同于将潜变量作为观测变量的共同先导因素的潜变量模型, 网络分析模型将观测变量作为初级指标, 采用图论的方法建立观测变量之间的关系网络, 其中变量为网络的节点, 而变量间的关系是节点之间的连线。因此网络分析可以突显观测变量之间的联系以及观测变量相互影响而形成的系统。通过变量网络中基于各个节点特征的指标(如中心性)以及基于整体结构特征的指标(如小世界性), 网络分析为研究各种心理现象提供了新的可视化的描述方式和理解视角。近10年来, 网络分析的方法已在人格心理学、社会心理学和临床心理学等领域得到一定的应用。未来研究应继续发展和完善网络分析模型的理论和方法, 使之运用到更多的数据类型和更广的研究领域中。

关键词: 网络分析, 潜变量模型, 心理测量, 临床心理学, 人格特质

Abstract:

Network analysis models (or Network Psychometrics) have been widely used in psychology research in recent years. Unlike latent variable models which conceive observable variables as outcomes of unobservable latent factors, network analysis models apply the graph theory to construct a network to depict the associations among observable variables. The observable variables are treated as nodes and the associations between them are treated as edges. As such, network analysis models reveal the relationships among observable variables and the dynamic system resulted from the interactions between these observable variables. With indices reflecting individual nodes’ characteristics (such as centrality) and network structural characteristics (such as small-worldness), network analysis models provide a new perspective for visualization and for studying various psychological phenomena. In the past decade, network analysis models have been applied in the fields of personality, social, and clinical psychology as well as psychiatry. Future research should continue to develop and improve the methods of network analysis models, making them applicable to more types of data and broader research fields.

Key words: network analysis, latent variable, psychometrics, psychopathology, personality traits

中图分类号: