Message/Author 

D. A. Work posted on Wednesday, October 26, 2011  4:34 am



Hello, my goal is it to estimate a Multilevel Model. But at first I want to try a basline model: A logistic regression model. Besides the dependend categorical variable I also habe some categorical independent variables with more than two categories. But in the variable statement I`m not able to define those independent variables as categorical variables. The output always says, that only dependent variables can be categorical. Now I don`t know how to understand the output. How do I have to interpret the findings, while the different categories of the independent (categorical) variables are not printed. Do I have to add more variables (dummy variables) to the model? Or is it possible to define a reference categorie like it is in SPSS? Thank you for your help! 


You need to create dummy variables for the categorical independent variables. You can do this using the DEFINE command. The number of dummy variables is the number of categories minus one. 

D. A. Work posted on Wednesday, October 26, 2011  9:39 am



Thank you very much!! I can`t find the intercept in my Output. My Output statements are stand tech1 tech4. Shall I nevertheless create "number of categories minus one" dummy variables? Or shall I create "number of categories = number of dummy variables"? Or do you have another advice? Again thanks a lot! 


Please send the output and your license number to support@statmodel.com. The number of dummy variables is the number of categories minus one. 

D. A. Work posted on Friday, October 28, 2011  5:35 am



Thank you very much for your answer! I already solved the problem respectively found the answer in the forum. 

Yvonne LEE posted on Thursday, July 10, 2014  5:12 pm



I am doing SEM with WLSMV estimation. The outcome variable is an ordinal data. One of the independent variables is also an ordinal data also. After running the model, this ordinal independent variable had no residual variance. Is this the expected outcome? Or, I did something wrong. 


Don't specify independent variables as categorical, if by independent you mean covariate (x in Mplus language). 

Back to top 