Yemi posted on Saturday, November 13, 2010 - 8:43 pm
I have three quick questions:
I have an id variable that goes from 6-9 digits. However for anything over 8 digits, the id number is rounded, and my id is no longer unique. Can I increase the format of my id number in the save variables command and how do I do that?
Secondly: I am running a latent class model with covariates. The number of individuals in each class changes with the number of covariates I add, as does the slope of the classes when I look at the graphs. I am just wondering the intuition behind this.
Thirdly, is it possible to control for individuals that belong to the same household (clustering) in the latent class model?
Older versions of the program had a limit on the number of digits for the ID variable.
When the addition of covariates changes the classes, this may point to a need for direct effects from the covariates to the latent class indicators. See the following paper which is available on the website:
Muthén, B. (2004). Latent variable analysis: Growth mixture modeling and related techniques for longitudinal data. In D. Kaplan (ed.), Handbook of quantitative methodology for the social sciences (pp. 345-368). Newbury Park, CA: Sage Publications.
Related to multilevel latent class analysis see the following paper which is available on the website:
Henry, K. & Muthén, B. (2010). Multilevel latent class analysis: An application of adolescent smoking typologies with individual and contextual predictors. Structural Equation Modeling, 17, 193-215.