Message/Author |
|
Fujiao Xie posted on Friday, November 27, 2015 - 9:58 am
|
|
|
Dear All, I have a question on how to interpret the factor loading of the latent variable indicated by both categorical and continuous variables in the measurement model. In the measurement model (CFA), the latent variable is indicated by 4 categorical variables and 2 continuous variables. The Mplus setup one categorical variable as fixed 1. After ran the model in Mplus, I got a factor loading for each variables of the latent variable. Now how can I interpret the factor loading of 4 categorical variables and 2 continuous variables? Please help! Thanks! |
|
|
If you are using WLSMV, the factor loadings for the categorical indicators are probit regression coefficients. The factor loadings for the continuous indicators are linear regression coefficients. |
|
Fujiao Xie posted on Friday, November 27, 2015 - 1:42 pm
|
|
|
Thank you, Linda! Further question is: In this measurement model, the latent variable is indicated by both categorical variables and continuous variables. what's the rule to setup one of the indicators fix to 1? Do I need choose one of the categorical variable fix to 1, or one of the continuous variables? |
|
|
There is no rule. Choosing a continuous variable may make the interpretation simpler. |
|
|
Hi there, I am fitting a one-factor CFA model in MPlus with 1 continuous indicator (values 0-10, and skewed distribution) and 10 categorical variables (4-option Likert). My sample is n=106 subjects. I have used ULSMV and all seems OK. However,I came across a comment about this estimation method and now wondering whether it is indeed OK to combine continuous and categorical indicators with ULSMV. thanks a lot k. |
|
|
Yes, it is. |
|
Back to top |