CINTERVAL Calculation PreviousNext
Mplus Discussion > Structural Equation Modeling >
Message/Author
 Stefanie Hinkle posted on Tuesday, September 06, 2011 - 7:11 pm
Hi.

I am trying to replicate the confidence interval calculation that Mplus produces with the CINTERVAL statement so that I can calculate others by hand using the variance covariance matrix when including X and X^2 terms in the model.

Can you please explain how the 95% CI is calculated for the parameter estimates?

I tried B +- 1.96*se, but the CI is slightly off. Am I missing something?

I am running an analysis with all continuous variables. I am running the model with a stratification, cluster and weight variable, but this should only effect the SE, not the calculation for the CI, correct?

Any help you can offer is greatly appreciated!

Thanks!
 Linda K. Muthen posted on Tuesday, September 06, 2011 - 7:39 pm
This is probably due to rounding error. You are using only three decimals where more are used by the program.
 Stefanie Hinkle posted on Tuesday, September 06, 2011 - 8:03 pm
Ok. Thank you!
 Stefanie Hinkle posted on Tuesday, September 06, 2011 - 8:27 pm
Dr. Muthen,

Can you offer any help on how to use to the variance covariance matrix from the tech3 output to calculate the standard error for an indirect effect?

Thank you very much!
 Linda K. Muthen posted on Tuesday, September 06, 2011 - 11:20 pm
I would Google the Delta method for the product of two coefficients.
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: