Latent Variables in Statistical Models: Mixed-Effects, Finite Mixture and more
Many statistical models used in practice utilize latent variables. Latent variables are simply variables that are unobserved. It may seem odd at first to consider fitting a statistical model when some of the data is unobserved (i.e., latent). However, these types of situations arise frequently in statistical practice. In this lecture, I will introduce the basic idea behind latent variables and illustrate the concept using longitudinal data analysis, finite mixture models and other examples as well. I will also highlight the use of latent variables in controversies on the issue of selecting an appropriate statistical model for analyzing your data.
Expectations: The participants should be familiar with basic statistical concepts and with regression in particular.
- Monday, February 10, 2020
- 10:00am - 12:00pm
- Lapidus Library Classroom G101