Mark Wade posted on Friday, April 25, 2014 - 6:26 pm
Hi Dr. Muthen,
I am estimating a CFA model with 4 indicators. There are variable amounts of missing data on the 4 indicators. If I want to save factor scores, how are these computed using the available data? Is this done on only on those individuals with full data for all 4 indicators, those with at least two measures, etc.? I am estimating the model using FIML and an MLR estimator. Thank you.
Factors scores have two components, the model estimated parameters and an individuals factor indicators. The model estimated parameters use all available information. When an individual has a missing value on a factor indicator, a value is imputed using other factor indicator information.