Fit Structural Equation Models to Multiply Imputed Data
Calculate the "D2" statistic
Class for a lavaan Model Fitted to Multiple Imputations
Fit a lavaan Model to Multiple Imputed Data Sets
Covariance and Correlation Residuals
Likelihood Ratio Test for Multiple Imputations
Score Test for Multiple Imputations
Wald Test for Multiple Imputations
Modification Indices for Multiple Imputations
Pooled Parameter Estimates
Fit a Saturated lavaan Model to Multiple Imputed Data Sets
Standardized Pooled Parameter Estimates
The primary purpose of 'lavaan.mi' is to extend the functionality of the R package 'lavaan', which implements structural equation modeling (SEM). When incomplete data have been multiply imputed, the imputed data sets can be analyzed by 'lavaan' using complete-data estimation methods, but results must be pooled across imputations (Rubin, 1987, <doi:10.1002/9780470316696>). The 'lavaan.mi' package automates the pooling of point and standard-error estimates, as well as a variety of test statistics, using a familiar interface that allows users to fit an SEM to multiple imputations as they would to a single data set using the 'lavaan' package.
Useful links