An Evaluation of Categorical Data
Analy~is Methodology for County Estimates in North Carolina
Nancy J. Carter;
California State University, Chico
Douglas C. Bond
Statistical Reporting Service;
u.S. Department of Agriculture;
Washington, D.C. 20250;
Agricultural estimates at the county level are very important.
report details an evaluation of three different Categorical Data Analysis (CDA)
estimators which were used. to derive county-level estimates of agriculture
in North Carolina.
These three estimators Case 1 (Full Association Structure),
Case 2 (Partial Association Structure), and Case 3 (Iterative Proportional
Fitting) were each evaluated using the 1978 North Carolina Census of Agriculture
data and the 1981, 1982, and 1983 A & P survey data.
The Case 1 estimator
spemed better than the other two estimators when evaluated using % AbRolute
This report describes the analyses done and makes
recommendations for further research in this area.
Association Structure, Allocation Structure, % Absolute Relative
Difference, Iterative Proportional Fitting.
This report would not have been possible ~ithout the assistance of
Kay Schenk of the CSU. Chico Computer Center provided the expertise in
She worked long hours on writing, debugging. and
running the many programs needed for this project.
Her assistance and
patience were both great.
Andrea Bowman of the Chico State Mathematics Department typed this report
in a short time period.
She ~orked very quickly and, as usual. very
BarryFord. Doug Kleweno, Bob Tortora, and Fred Vogel of USDA all contributed
to the success of this project.
They were instrumental in originating the
research agreement and in continued assistance as the project proceded.
Of course, without the major financial aupport of the USDA/SRS, this
project would not have been possible.