95th ESA Annual Meeting (August 1 -- 6, 2010)

WK 9 - Structural Equation Modeling with Bayesian Methods

Sunday, August 1, 2010: 8:30 AM-4:30 PM
309, David L Lawrence Convention Center
James B. Grace
George Arhonditsis
In this workshop we will cover the fundamentals of structural equation modeling (SEM) using Bayesian methods. Both SEM and Bayesian statistics have received considerable attention in recent years, with little attention to their compatibility and integration. SEM represents a methodological framework for developing and evaluating multivariate hypotheses about systems, while Bayesian estimation techniques using Markov chain Monte Carlo procedures permit an expanded flexibility for estimation and extrapolation. Because of the complexity of the subject matter, the course will focus on a limited set of SE modeling examples to develop attendees' expertise. Illustrations of both data analysis and forecasting will be included. Modeling in R and winBUGs will be presented, along with modeling using classic SEM software. This workshop should appeal to (1) those interested in or experienced with Bayesian methodology who want to learn about SEM, as well as (2) those wishing to learn advanced SEM techniques. Those attending will benefit from a reasonably solid background in quantitative methods. We will present several illustrative case studies where classical and Bayesian SEM are used to address a variety of ecological problems. Students will be provided with a series of brief tutorials for use before and after the course for developing expertise.

Registration Fee: $0

See more of: Workshop