Survival analysis with latent indepen... PreviousNext
Mplus Discussion > Growth Modeling of Longitudinal Data >
 Anonymous posted on Tuesday, January 11, 2005 - 12:49 am
I am doing survival analysis similar to 'Muthén, B. & Masyn, K. (2001). Discrete-time survival mixture analysis. (#92)'. This works if I try to explain survival with observed independent variables. However, if I include a latent independent variable I get an error message:

The following MODEL statements are ignored:
* Statements in the OVERALL class:
* Statements in Class 1:

F is survival and E is a latent independent variable to explain survival chance.

Is it possible to use latent independent variables in survival analysis (in Mplus 2.12)?
 Linda K. Muthen posted on Wednesday, January 12, 2005 - 4:48 pm
You need Version 3 for this.
 Pablo Mora posted on Wednesday, April 30, 2008 - 3:41 pm
I am conducting survival analysis using latent intercept and slope as predictors of mortality. Everything looks fine except that when I look at the unstandardized values one of the vars' significance is .08 and when I look at the standardized value the significance is < .01.
Do you know why this is happening and what results I should trust?
 Linda K. Muthen posted on Thursday, May 01, 2008 - 9:28 am
Please send your input, data, output, and license number to
 Nan Zhang posted on Monday, June 21, 2010 - 8:23 am
Dear Linda
I have a question about the latent factor score in the Cox model with continuous a latent factor--
What is the estimated latent factor score based on? Is it related to
the correlations or Eigen values?

Thanks in advance!

 Linda K. Muthen posted on Monday, June 21, 2010 - 11:11 am
See Technical Appendix 11 for information on how Mplus estimates factor scores. That being said, I would not use factor scores as covariates in a Cox regression. I would define the factors in the MODEL command using the BY option and use these factors as covariates.
 Lena Herich posted on Sunday, August 21, 2011 - 11:12 am
Hi !
I am estimating a Cox regression model with some covariates. Is it possible to save the estimated survival (or hazard) function in Mplus Version 5/6 ?

Thanks for your help!
 Linda K. Muthen posted on Monday, August 22, 2011 - 3:14 pm
You can use the RESULTS option of SAVEDATA to save model results. The gph option created by the PLOT command contains all values that are plotted.
 Lena Herich posted on Friday, September 30, 2011 - 2:23 am

I have a question concerning cox regression analysis in Mplus. Can you tell me how the different standardized results are calculated and how they can be interpreted?

Thanks in advance.
 Linda K. Muthen posted on Friday, September 30, 2011 - 6:30 am
The are described generally under the STANDARDIZED option in the user's guide. See also Technical Appendix 3 on the website.
 Lena Herich posted on Tuesday, October 04, 2011 - 7:12 am
I found the description and see how the parameters are calculated.
Now I can interpretate the parameter estimates as the change in y in sd for a sd change in x. For this I am using the variance of the survival time, which is given in the Tech4 output.

I was wondering if you can give me a reference on how this variance is calculated ?

Thanks again!
 Linda K. Muthen posted on Tuesday, October 04, 2011 - 9:26 am
You can standardize a Cox regression coefficient with respect to x only.
Back to top
Add Your Message Here
Username: Posting Information:
This is a private posting area. Only registered users and moderators may post messages here.
Options: Enable HTML code in message
Automatically activate URLs in message