I have a sample of 415 individuals and I want to make a latent profile analysis of the levels of organizational commitment using Mplus or SmartPLS.
Hi! If you are using smartPLS, free content is available on the developers web site. Book by Hairs et al (especially chapters four and five) is a recommended reading for step by step procedure.
The only problem with partial least squares is, it assumes error term to be centered around zero. If you have a strong theoretical support to model the data while assuming error term centered around zero, you can use it.
As mentioned by Singh, Mplus and smartPLS have quite different uses. Mplus is for analysing models where a concept is measured, which means that some of the variance in the indicators is irrelevant for the measurement and unique for the indicator (reflexive indicators), PLS is a method which can be used for prediction, where all variance is considered relevant (formative indicators).
If you want more ideas for your project, maybe you should describe it in more detail.
A good explanation of latent class analysis (and thus latent profile analysis) can be found as topic five of video presentations at the Mplus web site: www.statmodel.com
In addition, check the chapter of K. Masyn regarding mixture modeling. I found it really helpful:
https://www.statmodel.com/download/Masyn_2013.pdf
https://www.r-bloggers.com/quick-example-of-latent-profile-analysis-in-r/
This is very nice introduction and guide for LPA in Mplus.
Article Finding latent groups in observed data: A primer on latent p...