Interpreting coefficients PreviousNext
Mplus Discussion > Categorical Data Modeling >
Message/Author
 anonymous posted on Friday, February 17, 2006 - 7:29 pm
Just a basic and quick question. I'm using categorical data (i.e., dependent variable) for a path analysis. I have gotten the following coefficients: y1 to y2 = .55 and y2 to y3 = .65. How would I interpret these? Sorry to be so basic.
 Linda K. Muthen posted on Saturday, February 18, 2006 - 2:45 pm
I am assuming that you are using the default WLSMV estimator and that y2 and y3 are categorical variables. In this case, the regresison coefficients are probit regression coefficients. You can interpret their sign and significance or turn them into probabilities.
 anonymous posted on Saturday, February 18, 2006 - 7:28 pm
I am using the WLSMV default estimator. For me, I would like to interpret the sign and significance and also provide information about the probabilities. Would the proper transformation be the "logit" transformations that are on pages 312 - 316 of the user's guide? If not I would greatly appreciate some assistance in where I should read or look in the user's guide. Thank you in advance
 Linda K. Muthen posted on Saturday, February 18, 2006 - 11:01 pm
There will be a section in the Version 4 user's guide that describes how to do this. It will be online in about a week. The informaiton in the user's guide you refer to is for logistic regression not probit regression. It is also described in Technical Appendix 1 which is on the website.
 Anonymous posted on Saturday, February 18, 2006 - 11:54 pm
Any idea when version 4 will be ready for market?
 bmuthen posted on Saturday, February 18, 2006 - 11:59 pm
Shipping the week of Feb 27. See also

http://www.statmodel.com/acatalog/shop.html
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: