JSM 2015 Preliminary Program

Online Program Home
My Program

Abstract Details

Activity Number: 263
Type: Contributed
Date/Time: Monday, August 10, 2015 : 2:00 PM to 3:50 PM
Sponsor: Section on Bayesian Statistical Science
Abstract #315735
Title: Model Averaging of Regression Coefficients: Considerations and Practical Guidelines
Author(s): Katharine Banner*
Companies: Montana State University
Keywords: Model averaging ; bayesian statstics ; model uncertainty ; prediction ; inference
Abstract:

Methods accounting for model uncertainty in regression problems have received a lot of recent attention. Bayesian Model Averaging (BMA) is often used to average posterior distributions of regression coefficients over a set of plausible regression models. This method implicitly assumes the same parameters exist in multiple models. However, in the context of regression, coefficients of particular explanatory variables appearing in multiple models do not necessarily hold equivalent interpretations across those models. The meaning of a model averaged regression coefficient is then completely dependent on the context of the problem as well as the estimated posterior model probabilities, making explanatory inference difficult. Perpetuating this problem in practice is the accessibility to easily implementable software for model averaging without diagnostic tools or guidelines for assessing its appropriateness. This gap between methods and practice can leave well-intentioned researchers with unclear inferences. We propose a set of considerations including explicit notation and two graphical diagnostic tools to aid researchers in making informed decisions when considering BMA.


Authors who are presenting talks have a * after their name.

Back to the full JSM 2015 program





For program information, contact the JSM Registration Department or phone (888) 231-3473.

For Professional Development information, contact the Education Department.

The views expressed here are those of the individual authors and not necessarily those of the JSM sponsors, their officers, or their staff.

2015 JSM Online Program Home