Please wait a minute...
心理学报
  论文 本期目录 | 过刊浏览 | 高级检索 |
有调节的中介模型检验方法:竞争还是替补?
温忠麟;叶宝娟
(1华南师范大学心理应用研究中心/心理学院, 广州 510631) (2江西师范大学心理学院, 南昌 330022)
Different Methods for Testing Moderated Mediation Models: Competitors or Backups?
WEN Zhonglin;YE Baojuan
(1Center for Studies of Psychological Application/School of Psychology, South China Normal University, Guangzhou 510631, China) (2School of Psychology, Jiangxi Normal University, Nanchang 330022, China)
全文: PDF(534 KB)   评审附件 (1 KB) 
输出: BibTeX | EndNote (RIS)      
摘要 

在心理和其他社科研究领域, 经常遇到中介和调节变量。模型的变量多于3个时, 可能同时包含中介和调节变量, 一种常见的模型是有调节的中介模型。本文检视文献上各种检验有调节的中介模型的方法, 理清方法之间是竞争关系(分清优劣)还是替补关系(分清先后), 在此基础上总结出检验有调节的中介模型的步骤, 并用一个实例进行演示。文中也讨论了有调节的中介模型与有中介的调节模型的联系与区别。

服务
把本文推荐给朋友
加入引用管理器
E-mail Alert
RSS
作者相关文章
温忠麟
叶宝娟
关键词 中介调节有调节的中介有中介的调节    
Abstract

Mediation and moderation models are frequently used in the research of psychology and many other social science disciplines. Mediation indicates that the effect of an independent variable on a dependent variable is transmitted through a third variable, namely a mediator. For example, students’ gratitude promoted their academic achievement medially by increasing everyday academic resilience. Moderation occurs when the effect of an independent variable on a dependent variable varies according to the level of a third variable, which interacts with the independent variable. For instance, adolescents’ perceptions of their teachers’ authoritative teaching moderated the effect of antisocial disruption on peer acceptance. It is not uncommon for hypotheses about moderation and mediation relationships to occur in the same context where more than three variables are involved. When a mediation effect is moderated by a moderator, the effect is termed moderated mediation and the model is moderated mediation model. For example, everyday academic resilience acts as a mediator between gratitude and academic achievement, and this mediation process is moderated by stressful life events. There are several methods for testing moderated mediation models. The moderated mediation models being used for proposed testing methods are different. We are wondering whether different testing methods are competitors, or some of them are only backups. We discussed the different testing methods based on the most general type of model. The traditional testing method is the moderated causal steps approach, in which the regression coefficients are tested in sequence. Modern methods include the testing of the products of coefficients by using Bootstrap method or MCMC method, and testing of the difference between the maximum and minimum of the mediation effects. On the basis of the previous studies it can be summarized that the power of test with the moderated causal steps approach is the lowest among the three testing methods, whereas the testing of the difference between the maximum and minimum of the mediation effects has the highest power of test. After comparing significant results of the three testing methods by reviewing the simplification, implication, information, and explanation, we concluded that the moderated causal steps approach should be recommended first; testing of the products of coefficients by using Bootstrap method or MCMC method should be treated as a backup; testing of the difference between the maximum and minimum of the mediation effects should be the final choice. We proposed a hierarchical procedure for testing moderated mediation models as follows: Step 1. Adopt the moderated causal steps approach to test the model. If the significant result is obtained from the test, the mediating effect is moderated. Otherwise, go to Step 2. Step 2. Test the products of coefficients by using Bootstrap method or MCMC method. If any product is significantly different from zero, the mediating effect is moderated. Otherwise, go to Step 3. Step 3. Test the difference between the maximum and minimum of the mediation effects. If the difference is significant, the mediating effect is moderated. Otherwise, the mediating effect is not moderated. As an illustration, the procedure was applied to an empirical study in which everyday academic resilience played the role of a mediator between gratitude and academic achievement, and this mediation process moderated by stressful life events. The relationship and difference between moderated mediation models and mediated moderation models were also discussed.

Key wordsmediation    moderation    moderated mediation    mediated moderation
收稿日期: 2013-05-02      出版日期: 2014-05-24
基金资助:

国家自然科学基金(31271116)、教育部人文社会科学重点研究基地项目(11JJD190005)、中国博士后科学基金项目(2013M540535)、教育部人文社会科学研究青年基金项目(13YJC190029)和江西省博士后科研择优资助项目(2013KY08)资助。

通讯作者: 温忠麟   
引用本文:   
温忠麟;叶宝娟. 有调节的中介模型检验方法:竞争还是替补?[J]. 心理学报, 10.3724/SP.J.1041.2014.00714.
WEN Zhonglin;YE Baojuan. Different Methods for Testing Moderated Mediation Models: Competitors or Backups?. Acta Psychologica Sinica, 2014, 46(5): 714-726.
链接本文:  
http://journal.psych.ac.cn/xlxb/CN/10.3724/SP.J.1041.2014.00714      或      http://journal.psych.ac.cn/xlxb/CN/Y2014/V46/I5/714
[1] 王 燕, 林镇超, 侯博文, 孙时进.  生命史权衡的内在机制:动机控制策略的中介作用[J]. 心理学报, 2017, 49(6): 783-793.
[2] 王艳辉;李董平;孙文强;赵力燕;赖雪芬;周月月. 亲子依恋与初中生亲社会行为:有调节的中介效应[J]. 心理学报, 2017, 49(5): 663-679.
[3] 刘馨元;张志杰. 空间−时间关联的中介共同表征结构: 来自反转STEARC效应的证据[J]. 心理学报, 2017, 49(4): 427-438.
[4] 杨文琪;李强;郭名扬;范谦;何伊丽. 权力感对个体的影响:调节定向的视角[J]. 心理学报, 2017, 49(3): 404-415.
[5] 黄敏学;王艺婷; 廖俊云;刘茂红. 评论不一致性对消费者的双面影响:产品属性与调节定向的调节[J]. 心理学报, 2017, 49(3): 370-382.
[6] 张林;刘燊;徐强;吴晓燕;杨梦圆. 日常环境中的暴力暴露对攻击行为的长期影响:一个有调节的中介模型[J]. 心理学报, 2017, 49(1): 50-59.
[7] 周宵;伍新春;曾旻;田雨馨. 青少年的情绪调节策略对创伤后应激障碍和创伤后成长的影响:社会支持的调节作用[J]. 心理学报, 2016, 48(8): 969-980.
[8] 郑晓明; 刘鑫. 互动公平对员工幸福感的影响:心理授权的中介作用与权力距离的调节作用[J]. 心理学报, 2016, 48(6): 693-709.
[9] 温忠麟;范息涛;叶宝娟;陈宇帅. 从效应量应有的性质看中介效应量的合理性[J]. 心理学报, 2016, 48(4): 435-443.
[10] 崔楠;徐岚;谢雯婷. 做会后悔还是不做会后悔?——自我调节模式对不作为惯性的影响[J]. 心理学报, 2016, 48(4): 423-434.
[11] 温红博;梁凯丽;刘先伟. 家庭环境对中学生阅读能力的影响:阅读投入、阅读兴趣的中介作用[J]. 心理学报, 2016, 48(3): 248-257.
[12] 张淑华;刘兆延. 组织认同与离职意向关系的元分析[J]. 心理学报, 2016, 48(12): 1561-1573.
[13] 姜英杰; 王志伟; 郑明玲; 金雪莲. 基于价值的议程对学习时间分配影响的眼动研究[J]. 心理学报, 2016, 48(10): 1229-1238.
[14] 刘俊升;周颖;李丹;陈欣银. 儿童中期和青春期早期独处偏好与心理适应之关系:有调节的中介效应[J]. 心理学报, 2015, 47(8): 1004-1012.
[15] 张奇勇;卢家楣. 先入观念对情绪感染力的调节——以教学活动为例[J]. 心理学报, 2015, 47(6): 797-806.
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
版权所有 © 《心理学报》编辑部
本系统由北京玛格泰克科技发展有限公司设计开发  技术支持:support@magtech.com.cn