|
Activity Number:
|
312
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Tuesday, August 5, 2008 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Social Statistics Section
|
| Abstract - #302526 |
|
Title:
|
Spatial Empirical Bayes: Borrowing Strength Through Spatial Dependence
|
|
Author(s):
|
Natalya Verbitsky*+ and Stephen W. Raudenbush
|
|
Companies:
|
Mathematica Policy Research, Inc. and The University of Chicago
|
|
Address:
|
600 Maryland Ave, SW, Washington, DC, 20024,
|
|
Keywords:
|
empirical Bayes ; small area estimation ; Chicago neighborhood measures ; collective efficacy ; spatial dependence ; ecometrics
|
|
Abstract:
|
Social scientists have long been interested in differences between neighborhoods in their social structure and informal organization. In this paper we show that precision and validity of ecometric measures can be improved by exploiting spatial dependence of neighborhood processes within the framework of empirical Bayes shrinkage. We compare three estimators of a neighborhood social process: ordinary least squares (OLS), empirical Bayes estimator based on the independence assumption (EBE), and empirical Bayes estimator that exploits spatial dependence (EBS). Under our model assumptions, EBS dominates EBE and OLS in estimating the true latent social process. A cross-validation study using 1995 and 2002 data from the Project on Human Development in Chicago Neighborhoods is conducted and shows that that the empirical benefits of EBS approximate those expected under our model assumptions.
|
