Correlation Matrix for EFA PreviousNext
Mplus Discussion > Exploratory Factor Analysis >
Message/Author
 Daniel Lee posted on Wednesday, June 13, 2012 - 5:45 pm
Hello,
I ran an EFA through CEFA and Mplus. It was interesting to note the discrepancies in factor loading matrix. I'm wondering if the differences result from the different type of rotation employed in CEFA and MPlus. In CEFA, I ran a Crawford-Ferguson Quartimax rotation, while in Mplus I ran a Promax rotation (both oblique). Do you think this might be the cause?

Additionally, in MPlus, I ran an EFA using a .dat file that contains the raw research data (response for each participant on each variable). For CEFA, I ran an EFA using correlation between measured variables. Can this cause discrepancy in factor loading matrix, as well?

Thank you so much!!
 Linda K. Muthen posted on Wednesday, June 13, 2012 - 9:05 pm
Different rotations will give different solutions. Try the CF-QUARTIMAX rotation in Mplus also.

In both cases a correlation matrix is analyzed. Be sure the sample sizes are the same.
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: