Interpretation of the coefficients PreviousNext
Mplus Discussion > Structural Equation Modeling >
Message/Author
 Ujjwal Kumar Das posted on Sunday, January 21, 2018 - 9:38 am
Dear all,

I am new in Mplus and SEM. May be it is too naive to ask, but why there is only one coefficient for a categorical predictor variable which has five categories. In usual regression, dont we have n-1 (in this case four) coefficients comapared to a base category?

How do we interpret the sole coefficient in this situation? Please note, I have used WLSMV estimator and the categorical variable under discussion is used both as predictor and endogenous.
 Bengt O. Muthen posted on Sunday, January 21, 2018 - 5:27 pm
You have to create a set of binary dummy variables from your nominal predictor; this is not automatically done by Mplus. Also, the Categorical= and Nominal= lists are for DVs only, not for IVs( predictors).
 Ujjwal Kumar Das posted on Tuesday, January 23, 2018 - 2:20 am
Dear Prof Muthen,

Thank you for your reply. I have replaced original categorical variables each with n-1 dummies. But it results in two problems, a) standard errors and p value can't be determined, b) model doesnt converge unless forced. Just wanted to confirm if I didn't anything wrong.

1. Replaced original categorical variables with corresponding n-1 dummies
2. Declared categorical for the endogenous variables. For example, if X and Z are two endogenous categorical variables, I declared

Categorical=x2-x5 z2-z5;

Would you please respond?
 Bengt O. Muthen posted on Tuesday, January 23, 2018 - 3:50 pm
Please send your output to Support along with your license number.
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: