The National Agricutural Statistics Service (NASS) and NISS will hold its first cooperative research conference. Workshop participants will explore statistical aspects of agricultural survey research. NASS will review the latest methods in editing surveys that other organizations are using and will highlight NASS' developing plans to improve its system.
NASS has typically used univariate approaches to both imputation and mean-squared-error estimation. This workshop will explore and develop schemes for multivariate imputation that optimally support both economic modeling and direct estimates, while providing for an estimable impace of imputation on mean squared error.
The workshop will cover dual-frame estimation and look at advances in a variety of statistical techniques including capture-recapture, multiple frame estimation methodologies and weighting.
A session on small area/county estimation will cover efficient, accurate and statistically defensible county estimates. These are very important to NASS and are used in administering the USDA Farm Program.
The workshop will explore statistical approaches to forecasting and estimation that will focus on utilizing data from multiple surveys; from adminstrative/auxiliary information, including weather and remotely sensed data; and from long, historical series of surveys and estimation results.
Opening Speakers:
Alan Karr, Director of NISS
Cynthia Z.F. Clark, Administrator of NASS