Finite Mixture Modeling
A 2-Day Seminar Taught by Jeffrey Harring, Ph.D.
Finite mixture modeling (FMM), commonly referred to as model-based clustering methods, has been applied to a wide variety of cross-sectional and longitudinal data to account for heterogeneity in population characteristics. This two-day seminar will give you the theoretical background and practical skills to address interesting research questions using FMM.
The seminar starts with an introduction to mixture modeling in the context of univariate and multivariate distributions including general linear models, illustrating the use of Mplus. A brief introduction of how FMMs are estimated quickly gives way to identifying latent classes and a discussion of the challenges inherent in fitting such models. Next, structural equation models with measured and latent variables are briefly reviewed, providing a launching point to discuss latent variable mixture models for cross-sectional data and growth mixture models for longitudinal data. The integration of covariates and distal outcomes into an analysis will be discussed.
The seminar will combine lectures, software demonstrations, computer exercises, and discussion. There will be ample opportunities for participants to discuss how FMMs can be applied to their own research.
At the end of the seminar, participants will:
- understand how finite mixture models relate to the statistical methods and models (e.g., latent growth model) underlying the mixture models more broadly.
- conceptually understand the notation of finite mixture models.
- develop a principled modeling framework to step through any mixture analysis.
- understand the data analytic decision points in this modeling framework as they relate to the current methodological literature and conventional wisdom of fitting mixture models.
- understand the components necessary to specify mixture models using statistical software.
- be able to interpret the input/output from Mplus related to running a mixture analysis covered in the course.
Mplus will be used for all worked examples, but prior knowledge of Mplus is not essential. Participants are welcome and encouraged to bring their own laptop computer with the basic Mplus package installed.
The Mplus demo version (8.2) can be used for some examples used in the two-day seminar, but not all. For all examples, complete input and annotated output files are included in the packet of materials so participants can follow along, take notes, and run these on their own after the seminar. Datasets will be made available to all participants.
WHO SHOULD ATTEND
If your research includes analyses of cross-sectional or longitudinal data and there is reason to suspect that meaningful subgroups of individuals differ in characteristics of the analytic model used to analyze your data (i.e., factor means in a CFA or constant rate of change in a latent growth model), then this seminar is for you.
To benefit from this seminar, participants should have a good grasp of statistical methods up through multiple linear regression and exposure to structural equation modeling including topics such as confirmatory factor models, latent variable path models, model identification, estimation, and data-model fit assessment. Prior knowledge of Mplus is not required.
LOCATION, FORMAT AND MATERIALS
The class will meet from 9 am to 5 pm each day with a 1-hour lunch break at Temple University Center City, 1515 Market Street, Philadelphia, PA 19103.
Participants receive a bound manual containing detailed lecture notes (with equations and graphics), examples of computer printout, and many other useful features. This book frees participants from the distracting task of note taking.
Registration and lodging
The fee of $995.00 includes all seminar materials. The early registration fee of $895.00 is available until April 3.
If you cancel your registration at least two weeks before the course is scheduled to begin, you are entitled to a full refund (minus a processing fee of $50).
Lodging Reservation Instructions
A block of guest rooms has been reserved at the Club Quarters Hotel, 1628 Chestnut Street, Philadelphia, PA at a special rate of $165 per night. This location is about a 5-minute walk to the seminar location. In order to make reservations, call 203-905-2100 during business hours and identify yourself by using group code SH0502 or click here. For guaranteed rate and availability, you must reserve your room no later than Tuesday, April 2, 2019.
If you make reservations after the cut-off date ask for the Statistical Horizons room rate (do not use the code) and they will try to accommodate your request.
- Overview and preliminaries
- Univariate mixtures
- A brief, non-mathematical introduction to estimating FMMs
- Mixtures of general linear models
- Mixtures of multivariate normal distributions
- Class enumeration
- Challenges with fitting mixture models
- A brief introduction to structural equation modeling
- Direct and indirect mixture applications of structural equation models
- A brief introduction to latent growth models
- Growth mixture modeling
- Adding covariates and distal outcomes
- Discuss extensions to basic models as time allows