|
Activity Number:
|
428
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, August 9, 2006 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Section on Bayesian Statistical Science
|
| Abstract - #307093 |
|
Title:
|
Analysis in Opinions about the Death Penalty in U.S. States
|
|
Author(s):
|
Shouhao Zhou*+ and Andrew Gelman
|
|
Companies:
|
Columbia University and Columbia University
|
|
Address:
|
1255 Amsterdam Ave., Room 1017, New York, NY, 10027,
|
|
Keywords:
|
time series ; hierarchical modeling ; Gibbs sampler ; AR(p) error
|
|
Abstract:
|
We use hierarchical logistic regression model to analyze opinions about the capital punishment from two datasets by different polls in U.S. states from 1953. Some sampling methodology for posterior conditional distribution of time correlated variables is introduced, as well as model evaluation by counting the degree of freedom for different factors. Interesting estimates are reported and discussed.
|
