Publications & Resources
A Simple Approach to Inference in Covariance Structure Modeling With Missing Data: Bayesian Analysis
In this report, CRESST/UCLA researcher investigates an improved approach for educational analyses where there are significant amounts of missing data. The author found that a Bayesian approach developed by himself and Gerhard Arminger, offers a promising technique for missing data covariance structure modeling. The technique should soon be available in covariance structure software.
Muthén, B. (1996). A simple approach to inference in covariance structure modeling with missing data: Bayesian analysis (CSE Report 411). Los Angeles: University of California, Los Angeles, National Center for Research on Evaluation, Standards, and Student Testing (CRESST).