Factor score transformation to Rasch ... PreviousNext
Mplus Discussion > Confirmatory Factor Analysis >
 Kevin McGrew posted on Wednesday, February 16, 2011 - 8:30 pm
I have successfully saved factor scores for a three indicator single factor model, as well as other factor models The factor scores always have the excepted mean of zero, and an SD that varies by factor. For example, factor 1 scores have mean =0 and SD=6.6.

I now want to transform these to be comparable to other IRT based scores, specifically a W-score metric used in many educational tests (e.g., Woodcock-Johnson; Stanford Binet 5). If I run an item-level Rasch using Winsteps, the IRT log ability scores are transformed with the WINSTEPS commands: UMEAN=500; USCALE = 9.1024.

Can anyone tell me how to transform the saved factor scores to this new latent ability scale that is comparable to W-scores generated with Winsteps with the above parameters.
 Bengt O. Muthen posted on Thursday, February 17, 2011 - 12:14 am
Perhaps you would use

f2 = UMEAN + USCALE*(f1-M)/SD

where f1 is the factor score you get from Mplus, M is the mean of f1, SD is the standard deviation of f1. This assumes that USCALE has to do with the SD of f2, not f2's variance. Others?
Back to top
Add Your Message Here
Username: Posting Information:
This is a private posting area. Only registered users and moderators may post messages here.
Options: Enable HTML code in message
Automatically activate URLs in message