Hello, is anyone aware of work comparing/contrasting mixture modeling and taxometric methods? I have a copy of Cleland, Rothschild, & Haslam's (2000) paper, "Monte carlo comparison of taxometric, mixture model, and clustering procedures", but this paper was written prior to many of the new developments in mixture modeling. I'd appreciate anything either theoretical or empirical. Thank you.
I am not aware of any direct comparison, but that is a good topic for research. To read discussions of taxometrics from a modern point of view, see McDonald's 2003 JEBS review of the Waller-Meehl book.
The taxometric approach of Meehl involves graphical displays, resulting in a useful descriptive and exploratory device. The approach is, however, limited because in line with LCA, it assumes that there is conditional independence among the items within each class and furthermore it is applicable only to situations with two latent classes. For discussions of a more general approach, see factor mixture papers under Recent Papers on the Mplus web site.