The views expressed here are those of the individual authors and not necessarily those of the JSM sponsors, their officers, or their staff.
Online Program Home
Abstract Details
|
Activity Number:
|
578
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, August 1, 2012 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Section on Bayesian Statistical Science
|
| Abstract - #306444 |
|
Title:
|
How Far off Is Implicit Modeling in Multiple Imputation?
|
|
Author(s):
|
Florian Meinfelder*+
|
|
Companies:
|
Universitaet Bamberg
|
|
Address:
|
Lehrstuhl Fuer Statistik, Bamberg 96052, _, , Germany
|
|
Keywords:
|
Multiple Imputation ;
missing-data ;
Predictive Mean Matching ;
Bayesian Bootstrap
|
|
Abstract:
|
Multiple Imputation (MI) has established itself as a general-purpose solution for statistical analysis of incomplete data. With increasing acceptance of the method the demand for flexible and robust models has grown as well, and non-parametric elements have been implemented into MI algorithms. A Monte Carlo study is used to investigate the effects of e.g. replacing draws from the posterior distribution by (Approximate) Bayesian Bootstrapping, or using Predictive Mean Matching as a substitute for drawing from the conditional predictive distribution of the missing data. These apporaches fall into the category of 'implicit modeling', and results from the simulation study are compared to results based on 'explicit modeling' (i.e. fully-parametric Bayesian MI models).
|
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 2012 program
|
2012 JSM Online Program Home
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.