Saving factor scores for all observat... PreviousNext
Mplus Discussion > Confirmatory Factor Analysis >
Message/Author
 Hin Kwan Wong posted on Thursday, November 27, 2014 - 3:08 am
I have a set of data which I divided into a training subset and a validation subset. There is a column called TRAIN with a value of 1 indicating an entry is in the training set.

I wish to use only the training data to fit the model. but I wish Mplus to generate factor score for the whole set of data, what is the easiest way of doing so. Currently I have:

VARIABLE:
USEOBSERVATION = (TRAIN EQ 1)

but

SAVEDATA:
FILE is res.dat ;
SAVE = FSCORES;

only generate factor scores for the training set. Do I have to setup a new model and manually putting in the factor loadings from this result in a new zero dof CFA to generate factor scores for the validation data? Is there an easier way?
 Linda K. Muthen posted on Thursday, November 27, 2014 - 10:53 am
Yes, you would need to do that. Use the SVALUES option of the OUTPUT command to obtain input with the results as starting values. Change * to @ and use that input and the validation data to generate factor scores for the validation data.
 Hin Kwan Wong posted on Friday, November 28, 2014 - 2:39 am
Thank you very much for the promt reply, that's a real time saver!
Back to top
Add Your Message Here
Post:
Username: Posting Information:
This is a private posting area. Only registered users and moderators may post messages here.
Password:
Options: Enable HTML code in message
Automatically activate URLs in message
Action: