603 – Frequentist and Bayesian Models in Government Studies
Estimating the Number of U.S. Farms with Adjustment for Misclassification
Denise A. Abreu
USDA/National Agricultural Statistics Service
Samuel C. Garber
National Agricultural Statistics Service
Andrea C. Lamas
National Agricultural Statistics Service
Kenneth K. Lopiano
University of Florida
Five years ago, the National Agricultural Statistics Service (NASS) began a research effort to address an undercount in the estimate of the U.S. number of farms derived from its annual June Area Survey (JAS). Misclassification of farm status was found to be a major cause of the undercount. NASS has evaluated a host of measures and methods to assess, quantify, and account for this misclassification. The approach derived from this process-employs record linkage techniques, logistic regression, and NASS's annual list sampling frame. The methods developed and the subsequent results are presented here.