|
Activity Number:
|
12
|
|
Type:
|
Topic Contributed
|
|
Date/Time:
|
Sunday, August 6, 2006 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Social Statistics Section
|
| Abstract - #305782 |
|
Title:
|
Limited Information Goodness-of-Fit Testing in Multidimensional Contingency Tables
|
|
Author(s):
|
Harry Joe*+
|
|
Companies:
|
The University of British Columbia
|
|
Address:
|
Department of Statistics, Vancouver, BC, V6T 1Z2, Canada
|
|
Keywords:
|
multivariate discrete data ; item response theory ; composite likelihood
|
|
Abstract:
|
We introduce a family of goodness-of-fit statistics for testing composite null hypotheses in multidimensional contingency tables of arbitrary dimensions. These statistics are quadratic forms in marginal residuals up to order r. They are asymptotically chi-square under the null hypothesis when parameters are estimated using any consistent and asymptotically normal estimator. For an item response model, in nonsparse situations when the null distribution of X2 is approximately chi-square, we show empirically that the proposed statistics are also more powerful than X2. The proposed statistics, applied to subtables, also can be used for a piecewise goodness-of-fit assessment to determine the source of misfit in poorly fitting models. This research is joint with Albert Maydeu-Olivares.
|
