Accuracy of Parameter Estimates and Confidence Intervals in Moderated Mediation Models: A Comparison of Regression and Latent Moderated Structural Equations

被引:144
作者
Cheung, Gordon W. [1 ]
Lau, Rebecca S. [2 ]
机构
[1] Chinese Univ Hong Kong, Dept Management, Room 810,Cheng Yue Tung Bldg, Shatin, Hong Kong, Peoples R China
[2] Open Univ Hong Kong, Dept Management, Ho Man Tin, Hong Kong, Peoples R China
关键词
moderated mediation; latent moderated structural equations; confidence intervals; regression; MAXIMUM-LIKELIHOOD-ESTIMATION; MULTIPLE-REGRESSION; VARIABLE INTERACTION; MEASUREMENT ERROR; PARCELS; PRODUCT; LISREL; POWER;
D O I
10.1177/1094428115595869
中图分类号
B849 [应用心理学];
学科分类号
040203 ;
摘要
Currently, the most popular analytical method for testing moderated mediation is the regression approach, which is based on observed variables and assumes no measurement error. It is generally acknowledged that measurement errors result in biased estimates of regression coefficients. What has drawn relatively less attention is that the confidence intervals produced by regression are also biased when the variables are measured with errors. Therefore, we extend the latent moderated structural equations (LMS) methodwhich corrects for measurement errors when estimating latent interaction effectsto the study of the moderated mediation of latent variables. Simulations were conducted to compare the regression approach and the LMS approach. The results show that the LMS method produces accurate estimated effects and confidence intervals. By contrast, regression not only substantially underestimates the effects but also produces inaccurate confidence intervals. It is likely that the statistically significant moderated mediation effects that have been reported in previous studies using regression include biased estimated effects and confidence intervals that do not include the true values.
引用
收藏
页码:746 / 769
页数:24
相关论文
共 56 条
[1]   STATISTICAL POWER PROBLEMS WITH MODERATED MULTIPLE-REGRESSION IN MANAGEMENT RESEARCH [J].
AGUINIS, H .
JOURNAL OF MANAGEMENT, 1995, 21 (06) :1141-1158
[2]   Methodological artifacts in moderated multiple regression and their effects on statistical power [J].
Aguinis, H ;
StoneRomero, EF .
JOURNAL OF APPLIED PSYCHOLOGY, 1997, 82 (01) :192-206
[3]   A Note on Estimating the Joreskog-Yang Model for Latent Variable Interaction Using LISREL 8.3 [J].
Algina, James ;
Moulder, Bradley C. .
STRUCTURAL EQUATION MODELING-A MULTIDISCIPLINARY JOURNAL, 2001, 8 (01) :40-52
[4]  
[Anonymous], 1996, Advanced Structural Equation Modeling: Issues and Techniques, DOI DOI 10.1007/BF02291366
[5]  
[Anonymous], 2010, Mplus user's guide
[6]  
[Anonymous], 1987, Structural Equation Modeling with LISREL: Essentials and Advances
[7]   On making causal claims: A review and recommendations [J].
Antonakis, John ;
Bendahan, Samuel ;
Jacquart, Philippe ;
Lalive, Rafael .
LEADERSHIP QUARTERLY, 2010, 21 (06) :1086-1120
[8]   ORGANIZATIONAL THEORIES - SOME CRITERIA FOR EVALUATION [J].
BACHARACH, SB .
ACADEMY OF MANAGEMENT REVIEW, 1989, 14 (04) :496-515
[9]   The effects of item parceling on goodness-of-fit and parameter estimate bias in structural equation modeling [J].
Bandalos, DL .
STRUCTURAL EQUATION MODELING-A MULTIDISCIPLINARY JOURNAL, 2002, 9 (01) :78-102
[10]  
Bandalos D.L., 2001, New developments and techniques in structural equation modeling, P269, DOI 10.4324/9781410601858