Indicator order for latent variable PreviousNext
Mplus Discussion > Confirmatory Factor Analysis >
Message/Author
 Jonathan Williams posted on Sunday, January 19, 2014 - 9:31 pm
Why does the order of indicators alter the regression of the latent variable on an independent variable?

Example.
1st arbitrary indicator order:
Estimate S.E. Est./S.E. Pval

METS BY
ZTG 1.000 0.000 999.000 999.000
ZHDL -0.623 0.126 -4.939 0.000
ZMAP 0.317 0.115 2.747 0.006
ZBMI 0.795 0.161 4.944 0.000
ZGLU 0.600 0.119 5.062 0.000

METS ON
APDTIME 0.078 0.037 2.115 0.034

Different arbitrary indicator order:
Estimate S.E. Est./S.E. Pval

METS BY
ZMAP 1.000 0.000 999.000 999.000
ZTG 3.153 1.148 2.747 0.006
ZBMI 2.505 0.900 2.783 0.005
ZHDL -1.965 0.758 -2.593 0.010
ZGLU 1.891 0.732 2.585 0.010

METS ON
APDTIME 0.025 0.014 1.714 0.087


Why do the estimates & SEs for APDTIME depend on the order of the indicators?

Data & fit indices for both models are identical.

The orders of the indicators in both models are arbitrary. So, which (if either) is correct?
 Linda K. Muthen posted on Monday, January 20, 2014 - 6:52 am
The order should not matter. Please send the two outputs and your license number to support@statmodel.com.
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: