Latent Variable Models Under Misspecification: Two-Stage Least Squares (2SLS) and Maximum Likelihood (ML) Estimators

By Kenneth A. Bollen, James B. Kirby, P. Curran, Pamela Paxton, Feinian Chen. 2007

161 citations recorded by the source corpus.

Subfields: Management Science and Operations Research, Statistics and Probability

View Paper: https://www.semanticscholar.org/paper/9a62f9301a25cad5d175950bcf869e6f1029085b

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Management Science and Operations Research

In Statistics and Probability

Citations in the corpus

On making causal claims: A review and recommendations (2010, methodology): “…new method performed better under conditions of model misspecification (which makes sense given that it is a limited-information estimator); however, it is unclear as to whether this estimation approach is better than other limited-information (e.g., 2SLS) estimators (e.g., Bollen et al., 2007).”
Testing Negative Error Variances (2012, background): “Other frameworks, such as the one used by Chen et al. (2001) and Bollen et al. (2007) with a milder and continuously varying degree of misspecification, can be used to further study some of the issues raised in the current article.”
The Causal Effect of Class Size on Academic Achievement (2011, background): “See Bollen et al. (2007) for a comparison between ML and 2SLS estimators under a misspecified model.”
A Polychoric Instrumental Variable (PIV) Estimator for Structural Equation Models with Categorical Variables (2007, background): “Similar conditions for robustness to structural misspecification are not known for the system-wide estimators such as polychoric ULS, WLS, or DWLS, but based on experience with continuous variables (Cragg, 1968; Bollen et al., 2007) these system-wide estimators are likely to be less robust than the ”
Generalized Structured Component Analysis with Latent Interactions (2010, background): “In the present study, absolute values of relative bias greater than ten percent are considered indicative of an unacceptable degree of bias ( Bollen, Kirby, Curran, Paxton, & Chen, 2007 ; L ei,2009).”
Guidelines for a graph-theoretic implementation of structural equation modeling (2012, background): “…local estimation are (1) it permits more complex specifications of responses and linkages than can be summarized by a covariance matrix and (2) a modularization of the estimation process that avoids propagating misspecification errors from one part of a network to other parts (Bollen et al. 2007).”