Hi, I did a CFA with longnitudanal Data for three measurement points (3 latent, correlated variables, each representing one measurment point; type=general missing H1). With the save=factos command I get factor scores even for subjects for which are no values available for a measurement point. I thought MPlus does not impute Data? How are these factor scores then estimated?
Mplus does not impute values. Any observations with missing values on all variables is eliminated from the analysis. For those with some information, all available information is used in estimating the factor scores.