Select Standardized Model Results for... PreviousNext
Mplus Discussion > Confirmatory Factor Analysis >
Message/Author
 Gwo-Bao Liou posted on Monday, June 03, 2013 - 3:47 pm
There are three types of standardized model results in Mplus:
1. STD Standardization;
2. STDYX Standardization; and
3. STDY Standardization.

Would you please let me know which standardized result is the correct one for CFA?

Besides, when and why we should report the standardized results instead of unstandardized results?

Thank you very much!!
 Linda K. Muthen posted on Monday, June 03, 2013 - 4:20 pm
See the STANDARDIZED option for a discussion of which coefficients to use in different situations.

Whether to report raw or standardized results would be your choice. You might check in journals in your area to see what they report.
 Gwo-Bao Liou posted on Monday, June 03, 2013 - 6:06 pm
Dear Dr. Muthen,

Many thanks for your prompt response!!
According to your advice, I knew whether to report raw or standardized results.

But after checking the discussions, I am still not clear how to select different standardized results in different situations.
For example, there are four situations as the following. Would you please suggest me using correct standardized results for these situations?
1. CFA with all continuous observed variables
2. CFA with some categorical and some continuous observed variables
3. factors on factors with all continuous observed variables
4. factors on factors with some categorical and some continuous observed variables

Thank you so much again!!
 Linda K. Muthen posted on Tuesday, June 04, 2013 - 9:37 am
You can use StdY for 1 and 2. You can use StdY or Std for 3 and 4.
 Gwo-Bao Liou posted on Tuesday, June 04, 2013 - 6:38 pm
Dear Dr. Muthen,

A lot of thanks for your valuable suggestions!!

I tried to understand the three types of the standardized results by reading the Mplus user guide.

After reading the page 721 and 722 of the user guide Version 7, I felt that STDYX perhaps is not suitable to be used when there are categorical (or binary) variables in the model. If I am wrong, please let me know.

I had read your suggestions and the user guide, but I still have the following questions. Would you mind let me know:

In condition one (CFA with all continuous observed variables), can I use StdXY and Std?
I found some researchers used StdXY in condition one.

In condition two (CFA with some categorical and some continuous observed variables), can I use Std?

In condition three (factors on factors with all continuous observed variables), can I use StdXY?

Thank you very much for your help!!
 Linda K. Muthen posted on Wednesday, June 05, 2013 - 12:22 pm
The standardization to use is not determined by the scale of the dependent variables.

For independent variables, use StdY with binary variables and StdYX with continuous variables.
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: