LCA: Sample Size & Subgroup Compariso... PreviousNext
Mplus Discussion > Latent Variable Mixture Modeling >
 Jeffry W Thigpen posted on Sunday, March 02, 2008 - 9:56 am
Greetings Dr. Muthen! I have two general questions regarding exploratory latent class analysis. First, are there rules regarding sample size? Second, do formal tests of significance exists for testing differences between subgroups within a sample? More specifically, if separate analyses are run for subgroups within a sample are formal tests of significance available (and necessary) to compare and test for differences in results between subgroups? Finally, can you identify reference materials that speak to these issues. Thanks for your help. JT
 Linda K. Muthen posted on Monday, March 03, 2008 - 8:57 am
There are no rules regarding sample size. The best way to determine the sample size needed for your particular data is to do a Monte Carlo simulation study where you generate data according to the values in your data set. See below for more details:

Muthén, L.K. & Muthén, B.O. (2002). How to use a Monte Carlo study to decide on sample size and determine power. Structural Equation Modeling, 4, 599-620.

In latent class analysis, you can include the grouping variable as a covariate given that the only parameters that vary are thresholds or intercepts. You can also use the KNOWNCLASS option to define the groups and test equalities using difference testing or MODEL TEST. See also the reference below:

Clogg, C.C. & Goodman, L.A. (1985). Simultaneous latent structural analysis in several groups. In Tuma, N.B. (ed.), Sociological Methodology, 1985 (pp. 81-110). San Francisco: Jossey-Bass Publishers.
Back to top
Add Your Message Here
Username: Posting Information:
This is a private posting area. Only registered users and moderators may post messages here.
Options: Enable HTML code in message
Automatically activate URLs in message