Adding Missing-Data-Relevant Variables to FIML-Based Structural Equation Models

By J. Graham. 2003

593 citations recorded by the source corpus.

View Paper: https://www.semanticscholar.org/paper/9e975c0d73ea6f986a99b03cd59787fbfa23c4b1

How researchers see this paper

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

In Statistics and Probability

Citations in the corpus

Testing Measurement Invariance in Longitudinal Data with Ordered-Categorical Measures (2016, unknown): “…for both the full information and the limited information estimators, auxiliary variables—variables that are not part of the measurement model but that have a substantial correlation with both missingness and the variables in the measurement model (Graham, 2003, 2009)—can be added to the model.”
Attrition in developmental psychology (2017, methodology): “To not influence the interpretation of model parameters, auxiliary variables are incorporated into the analysis model of FIML (i.e., saturated correlates approach; see Graham, 2003).”
Item Parceling in Structural Equation Modeling: A Primer (2008, background): “New techniques are continually proposed to address a range of challenges, such as missing data (e.g., Enders & Bandalos, 2001; Graham, 2003), statistical power (e.”
An introduction to modern missing data analyses. (2010, methodology): “See Graham (2003) or Enders (2010, chapter 5) for additional details.”
Missing data techniques for multilevel data: implications of model misspecification (2011, methodology): “Estimation bias was calculated as the average deviation, across R replications, of the parameter estimates from the population parameter value [8],”
Multiple Imputation by Chained Equations: What is it and how does it work? (2011, background): “Graham (2003) does describe more complex maximum likelihood analyses that can incorporate auxiliary variables but those methods are relatively rare.”
Principled missing data methods for researchers (2013, methodology): “Graham (2003) has shown that by constructing the saturated correlates model this way, one can include an auxiliary variable in the SEM model without affecting parameter estimate(s), or the model fit index for the model of substantive interest, which is Equation 18 in the current study.”
Multiple Imputation of Missing Data for Multilevel Models (2018, unknown): “When more information can be included from auxiliary variables, then missing values can be inferred from the observed data with greater accuracy (for a discussion about the use of auxiliary variables under FIML, see Enders, 2008; Graham, 2003).”
Exploring gains in reading and mathematics achievement among regular and exceptional students using (2013, methodology): “Under MAR, ML-based procedures (full information ML, the expectation-maximization [EM] algorithm) can effectively handle missing data in the respect of providing unbiased estimates (Enders, 2001; Graham, 2003; Shin, Davison, & Long, 2009).”
Missing Data in Educational Research: A Review of Reporting Practices and Suggestions for Improvement (2004, methodology): “be slightly larger (Collins, Schafer, & Kam, 2001; Schafer & Graham, 2002). One of the primary advantages of MI is the ease with which auxiliary variables can be incorporated into the imputation process. For example, one might wish to impute missing values using a superset of the variables included ”
On Obtaining Estimates of the Fraction of Missing Information From Full Information Maximum Likelihood (2012, background): “If the entire model is saturated (e.g., if the model is a regression model), O and O† are equal to the “saturated ML” mean and covariance matrix estimates typically printed in the output, also known as the EM means and covariances (Enders & Peugh, 2004; Graham, 2003).”
Best practices for missing data management in counseling psychology. (2010, methodology): “Because FIML is model-based, we used a saturated correlates modeling technique3 (see Graham, 2003) to include the covariate (attitude toward authority) as well as the predictor (condition) and the outcome variable in the model.”
Robust Two-Stage Approach Outperforms Robust Full Information Maximum Likelihood With Incomplete Nonnormal Data (2014, methodology): “EM means and the EM covariance matrix (Enders & Peugh, 2004; Graham, 2003) because the likelihood in Equation 1 can be easily maximized by the EM algorithm.”
Pathways to Occupational Attainment Among Poor Youth of Color (2009, methodology): “From a SEMperspective, the obtained fit indices and loading coefficients provide validity evidence for this operationalization of academic performance, occupational expectations, sociopolitical development, and occupational attainment (Bollen, 1989).”