How to interpret Mplus output PreviousNext
Mplus Discussion > Confirmatory Factor Analysis >
Message/Author
 AMY TIAN-FOREMAN posted on Wednesday, December 09, 2009 - 2:37 pm
Estimate S.E. Est./S.E. P-Value
Thresholds
LT50$1 2.149 0.080 26.883 0.000
THREAT$1 1.499 0.113 13.304 0.000
INJURE$1 4.413 0.327 13.512 0.000
BLDG$1 4.935 0.397 12.426 0.000

Dear Dr. Muthen:
I am new to Mplus, and have several questions regarding factor loading would like to ask you.
1) How do I calculate AVE/Composite reliability (CR) using factor loadings and errors from Mplus output? I used the formula AVE= sum of squared factor loading/(sum of squared factor loading + sum of standard errors), and CR = Squared sum of facter loadings/(squared sum of factor loadings + sum of standard errors). However my AVE and CR is unbelievably high (over 0.98 for all my scales). Using the example that I post above (from Muthén, (2001) paper posted on Mplus website), the AVE is 0.982, and CR is 0.995. Aint they almost too high? Or is this not the right way to interpret Mplus outputs?

2) again using the output above. if I use the statistic for item LT50$1, then Est./S.E. is 2.149/0.080=26.863 instead of 26.883 as shown in the output. So, please please help me to understand this.

3) why are these standard estimate higher than 1.0??

Sorry for posting so many questions, I would be so grateful if you can help.

Best wishes
Amy
 Bengt O. Muthen posted on Wednesday, December 09, 2009 - 6:06 pm
You are looking at threshold estimates, not loading estimates.

1) I don't know what "AVE/Composite reliability" is.

2) The difference is due to your computations using only 3 decimals.

3) They are thresholds, not loadings.
 AMY TIAN-FOREMAN posted on Thursday, December 10, 2009 - 10:02 am
Dear Bengt:
Thank you very much for your reply.
Please tell me where to look for factor loadings..if my outputs are as following:
Estimate S.E. Est./S.E. P-Value

F2 BY
PS1 0.551 0.021 26.177 0.000
PS2 0.551 0.021 25.928 0.000
PS3 0.551 0.021 25.975 0.000
PS4 0.518 0.022 23.429 0.000

then factor loading for PS1 is 0.551, isnt it??
 Linda K. Muthen posted on Thursday, December 10, 2009 - 11:34 am
The second column are the parameter estimates, in this case factor loadings.
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: