Categorical/observed variables PreviousNext
Mplus Discussion > Structural Equation Modeling >
Message/Author
 Liz Woodruff posted on Tuesday, June 07, 2011 - 3:59 pm
2 of the variables in my SEM are categorical observed, whereas the other 4 are latent (with indicators, of course). When specifying my SEM in M+, should I treat the observed variables (i.e., F5 & F6) in the same way as my latent (i.e., F1-F4):

E.g.,
F4 ON F3; (both continuous latent vars)
F3 (latent var) ON F5 (categorical var);
 Liz Woodruff posted on Tuesday, June 07, 2011 - 4:18 pm
Actually, I am just confused about using categorical/binary variables in my SEM. Basically, I have 2: sexual orientation and participation in sports. Both are yes/no variables. For example, either one participates in sports yes (1) or no (2). I am simply confused about how to represent these in the model in M+. If you could point me in a direction, I would appreciate it.
 Liz Woodruff posted on Tuesday, June 07, 2011 - 4:18 pm
Given that all of my other variables are continuous.
 Linda K. Muthen posted on Tuesday, June 07, 2011 - 5:31 pm
The scale of an independent variable is not an issue is regression. Your only categorical variable is an independent variable. You don't need to do anything.
 Liz Woodruff posted on Wednesday, June 08, 2011 - 9:59 am
Thank you
 Luo Wenshu posted on Saturday, December 23, 2017 - 5:39 am
Dr. Muthen, if I have both observed variables and latent variables as predictors for some latent outcome variables, do I need to specify correlations between the observed and latent predictors? It seems by default they are not set to be correlated in Mplus. Thank you!
 Bengt O. Muthen posted on Saturday, December 23, 2017 - 2:46 pm
Yes.
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: