JSM 2005 - Toronto

Abstract #303728

This is the preliminary program for the 2005 Joint Statistical Meetings in Minneapolis, Minnesota. Currently included in this program is the "technical" program, schedule of invited, topic contributed, regular contributed and poster sessions; Continuing Education courses (August 7-10, 2005); and Committee and Business Meetings. This on-line program will be updated frequently to reflect the most current revisions.

To View the Program:
You may choose to view all activities of the program or just parts of it at any one time. All activities are arranged by date and time.



The views expressed here are those of the individual authors
and not necessarily those of the ASA or its board, officers, or staff.

The Program has labeled the meeting rooms with "letters" preceding the name of the room, designating in which facility the room is located:

Minneapolis Convention Center = “MCC” Hilton Minneapolis Hotel = “H” Hyatt Regency Minneapolis = “HY”

Back to main JSM 2005 Program page
Legend: = Applied Session, = Theme Session, = Presenter
Activity Number: 348
Type: Topic Contributed
Date/Time: Wednesday, August 10, 2005 : 8:30 AM to 10:20 AM
Sponsor: Biometrics Section
Abstract - #303728
Title: Log-linear Models for Incomplete Categorical Data
Author(s): Paulo Soares*+ and Carlos D. Paulino
Companies: Instituto Superior Técnico, UTL and IST, Technical University of Lisbon
Address: Av Rovisco Pais 1, Lisboa, 1049 001, Portugal
Keywords: Incomplete categorical data ; Selection model ; Log-linear models ; Bayesian analysis
Abstract:

The observation of incomplete categorical data is a common situation in medical or biological studies. Often, the models used to analyze that type of dataset introduce strong assumptions about the missing data process. Alternative nonignorable models have been proposed that show this practice is unnecessary and can lead to poor results. In this work, we consider a nonignorable selection model for the incompleteness mechanism of multinomial data under a general missing-data pattern. This model is used with Dirichlet priors and provides a full Bayesian analysis with the support of current MCMC algorithms. The full multinomial model is general and allows the associations among categorical variables to be arbitrarily complex. Unless the number of variables is small, the observed data may not support such complexity. Loglinear models are a popular and flexible class of models that simplify the study of associations among variables, but when used in a Bayesian analysis, usually require prior election for several sets of parameters, each one associated with a candidate loglinear model.

  • The address information is for the authors that have a + after their name.
  • Authors who are presenting talks have a * after their name.

Back to the full JSM 2005 program

JSM 2005 For information, contact jsm@amstat.org or phone (888) 231-3473. If you have questions about the Continuing Education program, please contact the Education Department.
Revised March 2005