Regularized Structural Equation Modeling to Detect Measurement Bias: Evaluation of Lasso, Adaptive Lasso, and Elastic Net

ABSTRACT Correct detection of measurement bias could help researchers revise models or refine psychological scales. Measurement bias detection can be viewed as a variable-selection problem, in which biased items are optimally selected from a set of items. This study investigated a number of regularization methods: ridge, lasso, elastic net (enet) and adaptive lasso (alasso), in comparison with maximum likelihood estimation (MLE) for detecting various forms of measurement bias in regard to a continuous violator using restricted factor analysis. Particularly, complex structural equation models with relatively small sample sizes were the study focus. Through a simulation study and an empirical example, results indicated that the enet outperformed other methods in small samples for identifying biased items. The alasso yielded low false positive rates for non-biased items outside of a high number of biased items. MLE performed well for the overall estimation of biased items.

[1]  Kevin J Grimm,et al.  Comparison of Frequentist and Bayesian Regularization in Structural Equation Modeling , 2018, Structural equation modeling : a multidisciplinary journal.

[2]  A. Boomsma,et al.  Robustness Studies in Covariance Structure Modeling , 1998 .

[3]  H. Zou,et al.  Regularization and variable selection via the elastic net , 2005 .

[4]  Sy-Miin Chow,et al.  A Comparison of Bayesian and Frequentist Model Selection Methods for Factor Analysis Models , 2017, Psychological methods.

[5]  Jerome H. Friedman,et al.  On Bias, Variance, 0/1—Loss, and the Curse-of-Dimensionality , 2004, Data Mining and Knowledge Discovery.

[6]  Xinya Liang,et al.  Evaluation of Structural Relationships in Autoregressive Cross-Lagged Models Under Longitudinal Approximate Invariance:A Bayesian Analysis , 2018 .

[7]  G. Schwarz.  Estimating the Dimension of a Model , 1978 .

[8]  Gideon J. Mellenbergh,et al.  Item bias and item response theory , 1989 .

[9]  Xinyuan Song,et al.  Bayesian Adaptive Lasso for Ordinal Regression With Latent Variables , 2017 .

[10]  James G. Scott,et al.  The horseshoe estimator for sparse signals , 2010 .

[11]  H. Akaike.  A new look at the statistical model identification , 1974 .

[12]  R. Kievit,et al.  A Practical Guide to Variable Selection in Structural Equation Modeling by Using Regularized Multiple-Indicators, Multiple-Causes Models , 2019, Advances in methods and practices in psychological science.

[13]  Suzanne Jak,et al.  A Test for Cluster Bias: Detecting Violations of Measurement Invariance Across Clusters in Multilevel Data , 2013 .

[14]  Bengt Muthén,et al.  Bayesian structural equation modeling: a more flexible representation of substantive theory. , 2012, Psychological methods.

[15]  Holger Brandt,et al.  An Adaptive Bayesian Lasso Approach with Spike-and-Slab Priors to Identify Multiple Linear and Nonlinear Effects in Structural Equation Models , 2018, Structural Equation Modeling: A Multidisciplinary Journal.

[16]  G. Casella,et al.  The Bayesian Lasso , 2008 .

[17]  Howard T. Everson,et al.  Methodology Review: Statistical Approaches for Assessing Measurement Bias , 1993 .

[18]  Yves Rosseel,et al.  lavaan: An R Package for Structural Equation Modeling , 2012 .

[19]  Albert Maydeu-Olivares,et al.  Understanding the Model Size Effect on SEM Fit Indices , 2018, Educational and psychological measurement.

[20]  R. Purc-Stephenson,et al.  Reporting practices in confirmatory factor analysis: an overview and some recommendations. , 2009, Psychological methods.

[21]  Gerhard Tutz,et al.  A Penalty Approach to Differential Item Functioning in Rasch Models , 2015, Psychometrika.

[22]  Sy-Miin Chow,et al.  Bayesian Factor Analysis as a Variable-Selection Problem: Alternative Priors and Consequences , 2016, Multivariate behavioral research.

[23]  Helfried Moosbrugger,et al.  Maximum likelihood estimation of latent interaction effects with the LMS method , 2000 .

[24]  R. Terry,et al.  Bayesian SEM for Specification Search Problems in Testing Factorial Invariance , 2017, Multivariate behavioral research.

[25]  Wynne W. Chin,et al.  A Partial Least Squares Latent Variable Modeling Approach for Measuring Interaction Effects: Results from a Monte Carlo Simulation Study and an Electronic - Mail Emotion/Adoption Study , 2003, Inf. Syst. Res..

[26]  R. Tibshirani.  Regression Shrinkage and Selection via the Lasso , 1996 .

[27]  W. Meredith.  Measurement invariance, factor analysis and factorial invariance , 1993 .

[28]  Po-Hsien Huang,et al.  A penalized likelihood method for multi-group structural equation modelling. , 2018, The British journal of mathematical and statistical psychology.

[29]  Miao Yang,et al.  Structural Equation Modeling With Many Variables: A Systematic Review of Issues and Developments , 2018, Front. Psychol..

[30]  Bengt Muthén,et al.  Latent variable modeling in heterogeneous populations , 1989 .

[31]  Fredrik A. Dahl,et al.  Model Selection with Lasso in Multi-group Structural Equation Models , 2019, Structural Equation Modeling: A Multidisciplinary Journal.

[32]  T. J. Mitchell,et al.  Bayesian Variable Selection in Linear Regression , 1988 .

[33]  Christopher D. Nye,et al.  How Big Are My Effects? Examining the Magnitude of Effect Sizes in Studies of Measurement Equivalence , 2019 .

[34]  Hung-Hsuan Chen,et al.  A Penalized Likelihood Method for Structural Equation Modeling , 2017, Psychometrika.

[35]  Xinya Liang,et al.  A Comprehensive Comparison of Model Selection Methods for Testing Factorial Invariance , 2020, Structural Equation Modeling: A Multidisciplinary Journal.

[36]  Michael W. Browne,et al.  Topics in Applied Multivariate Analysis: COVARIANCE STRUCTURES , 1982 .

[37]  B. French,et al.  Multigroup Confirmatory Factor Analysis: Locating the Invariant Referent Sets , 2008 .

[38]  Fritz Drasgow,et al.  Effect size indices for analyses of measurement equivalence: understanding the practical importance of differences between groups. , 2011, The Journal of applied psychology.

[39]  H. Zou.  The Adaptive Lasso and Its Oracle Properties , 2006 .

[40]  Sarfaraz Serang,et al.  Exploratory Mediation Analysis via Regularization , 2017, Structural equation modeling : a multidisciplinary journal.

[41]  Jianqing Fan,et al.  Nonconcave penalized likelihood with a diverging number of parameters , 2004, math/0406466.

[42]  Peter M. Bentler,et al.  Practical Issues in Structural Modeling , 1987 .

[43]  Morten Moshagen,et al.  The Model Size Effect in SEM: Inflated Goodness-of-Fit Statistics Are Due to the Size of the Covariance Matrix , 2012 .

[44]  R. Millsap,et al.  Evaluating the impact of partial factorial invariance on selection in two populations. , 2004, Psychological methods.

[45]  John J. McArdle,et al.  Regularized Structural Equation Modeling , 2015, Multivariate behavioral research.

[46]  Xin-Yuan Song,et al.  Bayesian Regularized Multivariate Generalized Latent Variable Models , 2017 .

[47]  Cun-Hui Zhang.  Nearly unbiased variable selection under minimax concave penalty , 2010, 1002.4734.

[48]  Nicholas D. Myers,et al.  A Comparison of Uniform DIF Effect Size Estimators Under the MIMIC and Rasch Models , 2013 .

[49]  Francis Tuerlinckx,et al.  Detection of Differential Item Functioning Using the Lasso Approach , 2015 .