Comparing two different models PreviousNext
Mplus Discussion > Exploratory Factor Analysis >
 Yadong Ji posted on Thursday, September 06, 2018 - 12:55 pm

In the same sample, I measured a same construct by two instruments that were both published before. These two instruments, 12-item and 15-item, were both reported reliable. I ran EFA for both and the results showed that both instruments were one-factor structure.

My interest, for this particular study, is to decide which instrument is better. Other than the model indices, one of the reviewer wanted to see whether these two measurements were different. S/he appealed to know "whether there is a mean difference of the two measures using latent scores". Is there a way to do such test, as this is a cross-sectional one group data?

Thank you!!!
 Linda K. Muthen posted on Thursday, September 06, 2018 - 2:51 pm
The factor mean for each instrument is zero in a cross-sectional study so I don't see how this could be done.
 Yadong Ji posted on Thursday, September 06, 2018 - 3:06 pm
Thank you Dr. Muthen for your reply. Any additional suggestions on how to compare two different measures of the same construct in general?
 Bengt O. Muthen posted on Thursday, September 06, 2018 - 3:20 pm
It can't really be done precisely unless some items in the two sets are the same.

All you can do is perhaps to analyze all 27 items together and see how high the factor correlation(s) is (are). If they are very high you can argue that they measure similar constructs. In analyses separately for the 2 items sets, you can then see which factor determinacy's is the highest.
 Yadong Ji posted on Thursday, September 06, 2018 - 4:35 pm
Thank you so much!
I really appreciate your reply!
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