Generating a latent variable for all subjects if some of the variables used in CFA have missing data

Polina_UH — Mon, 22 Sep 2025 11:43:57 GMT

Dear all,

I am running confirmatory factor analysis to generate a latent variable 'mental health'. Some subjects have missing data on on some dimensions of mental health. I use full information maximum likelihood option, and I would like to generate a latent variable 'mental health' for all subjects even if they have missing data on some of dimensions. My code produces this variable only for those who have full data on all variables.

proc calis method=fiml data=all outstat=ostat;
factor
mental_health ===>x1 x2 x3 x4 x5 x6 x7;
pvar mental_health=1.;
run;

proc score data=all score=ostat out=scores1;
var x1 x2 x3 x4 x5 x6 x7;
run;

Could you please help me?

Many thanks in advance!

topic Generating a latent variable for all subjects if some of the variables used in CFA have missing data in Statistical Procedures

Generating a latent variable for all subjects if some of the variables used in CFA have missing data