Research and Development Division
ASA and USDA-NASS Research Fellow and Associate Program
The American Statistical Association (ASA) in cooperation with the National Agricultural Statistics Service (NASS) of the U.S. Department of Agriculture conducts a Research Fellow and Associate Program. The purpose of this program is to supplement research in statistics with experience at NASS.
The program is designed to provide the selected Fellows and Associates with research opportunities or experience in the application of statistical theory in all phases of large scale agricultural survey operations, including design, collection, quality control, forecasting, estimation, and analysis. Participants in the program will engage in research which is of mutual interest to the participant and NASS. Fellows and Associates conduct research in residence at NASS, gaining unique opportunities for use of extensive NASS data and interaction with NASS staff.
Applicants for the fellowship program should have a Ph.D. and have established recognized research record and considerable expertise in their area of proposed research. Applicants for Associates are expected to have completed a Ph.D. in an appropriate field or to have made significant progress toward the degree (at least two years of graduate study). Selection of participants in the program is made by a panel of individuals from the ASA and NASS. The application procedure and program director information is provided on the last page of this brochure.
Applicants may request to work in a specific area or on a special problem, propose a broad area of learning, or seek an open program of research related to an area of expertise. All requests will be competitively evaluated based upon the content of the proposal, its applicability to NASS programs, and facilities and personnel available to assist in the applicant’s proposed research. A research report is expected of selected participants upon completion of the program.
To assist the participant in the conduct of research, NASS has a number of survey related facilities. These include actively maintained list and area sampling frames; survey design capability; computation and analysis capabilities using PC’s and a mainframe network; and a large scale data collection capability. The organizational structure of NASS features a large number of state offices capable of carrying out surveys utilizing a large corps of field and telephone enumerators. There are also a number of regularly scheduled surveys that provide vehicles for evaluating single or multiple frame methodologies. Some current research and application topics of interest to NASS are outlined on the following pages.
GENERAL STATISTICAL SURVEY METHODOLOGY TOPICS IN NASS
Proposals should be research projects applicable to the collection and analysis of agricultural statistics. Some areas of current interest are outlines for the following twelve topics concerning surveys in NASS. This list of topics is not exhaustive. Applicants are encouraged to propose other areas that may be of interest to NASS.
Statistical Survey Methodology
NASS personnel actively maintain a list sampling frame containing known operators of farms and agribusinesses. The list fame also has various measures of size capacity with the sampling units. Multiple frame methodology demands the use of an independent area frame to estimate for the list undercoverage. Therefore, a stratified and independent area frame based on land use is also maintained by the Agency. The use of a single frame or multiple frames involves sampling design and various kinds of nonsampling errors which need to be evaluated. Research in the efficient use of these frames is an area of ongoing evaluation.
Data collection within the operational Agency program involves telephone, personal, and self-administered interviews. The questionnaire design will vary by method of collection. Approaches for developing and testing questionnaires, including cognitive methods that may lead to improved efficiency or accuracy are of interest.
Computer Assisted Survey Methods
NASS has begun implementing Computer Assisted Interviewing into its operation program. Computer assisted telephone interview effects, and many other data quality issues. In addition, research is needed on the use of computer assisted personal interviewing, the use of computers in self-administered questionnaires, and associated interactive data entry and editing activities.
A diversity of agricultural surveys being conducted by NASS provides a wide variety of missing data situations. These situations range from missing information for the entire sampling unit to missing a portion of the desired information for the sampling unit. Development of consistent imputation procedures for multivariate and longitudinal survey situations is an area of research interest.
Survey Process Control
NASS continues to emphasize quality in its sample survey information collection efforts. More emphasis is being placed upon quality during the entire process frame preparation, sample selection, training, data collection, data handling and editing, and summarization plus analysis. Research into statistical methods which need to be applied across the entire survey process are part of this complex issue in survey sampling.
Nonsampling Error Research
The entire realm of nonsampling errors is potentially present in sample surveys. NASS collects information by telephone, personal, mail surveys and field and laboratory counts and measurements. Statistical procedures to identify, control, and measure nonsampling errors or measurement errors are of interest to the Agency program. An example would be the application and evaluation of alternative measurement error models.
Sample Design and Estimation Research
Sample design topics include list frame, area frame, multiple frame, objective yield designs for one time, cross sectional, spatial and longitudinal situations. Estimation topics include direct expansion, ratio, Bayesian, small area, regression, robust, time series, and longitudinal estimators.
Objective Yield Forecasting Methodology
Forecasting models are developed using historical survey relationships between early season plant and fruit characteristics and final plant and fruit yields. As the base survey design becomes more complex in the operational program of NASS, or the fruits or plants designated for the forecasting program change, estimation model parameters continually need research attention and evaluation. The estimation of mean square error components within this complex survey structure is a possible area of research. Recently, stochastic process modeling was introduced to the yield forecasting research effort in NASS.
Research in computer and other technology is part of the Agency research program. Areas of interest include the development of artificial intelligence software to evaluate multiple survey estimates and external sources of check data such as national soybean crushings or hog slaughter data, of expert systems to edit survey data, and of geographic information systems to access spatial environmental data.
NASS staff has participated in research in aerial photography and satellite imagery for several decades. The application of the earth resource and weather data satellite technology to agriculture estimation continues to intrigue the statisticians in this field of research. Efficient multivariate clustering and classification algorithms for very large data sets are of special interest and also the creation and evaluation of vegetative indices.
Economic and Environmental Statistics
NASS conducts an annual economics statistics survey as part of its Agriculture Resources and Management Study (ARMS). Environmental or farm chemical usage modules are also incorporated into the ARMS. A very detailed questionnaire is needed for each of the selected sample units. One area of research might be a survey design that uses both global and detailed questionnaires to increase efficiency and reduce respondent burden.
CONDITIONS OF APPOINTMENT AND BENEFITS
Fellows and Associates will conduct their research at NASS/Research and Development Division, 1400 Independence Avenue, SW, Washington, DC.. All participants are reimbursed by the American Statistical Association and are on a cooperative work arrangement with NASS.
The stipend for the Fellow will be commensurate with current faculty salary, plus an appropriate moving and travel allowance for the period of the fellowship. There is a possibility of an arrangement under the Inter-Government Personnel Act where the stipend supplements an existing sabbatical. The selection, however, will not depend on these arrangements. The stipend for the Associates will be set at the equivalent G.S. rating, depending on academic and other background, and include a moving and travel allowance.
Positions are typically an academic year for 9-12 months. They may be extended up to one additional year under special circumstances. Special appointments covering one or several shorter periods of 2-3 months will also be considered.
- A curriculum vitae.
- A general statement identifying and discussing the area of proposed research or interest. This should indicate research capabilities and knowledge of the area.
The fellow should have achieved a level of research stature in sample survey methods and theory, statistical methodology, or related area such as modeling or remote sensing. The individual should also indicate the value of the prepared research plan to the agency programs.
For the Associate
- A curriculum vitae.
- University transcript and graduate record.
- Two letters of recommendation.
- A statement of research or statistical interests.
Questions about the NASS program can be directed to:
Dr. Linda Young, Director, Research and Development Division, email: email@example.com.
Applications should be sent to:
ASA/USDA-NASS Fellows Program, American Statistical Association
732 Washington Street
Alexandria, Virginia 22314
The program year normally begins in January or September, but it is possible for an applicant to begin at another date. Applications for positions for the year beginning September should be received by March 1. For a starting date of January, the applications should arrive at the ASA office approximately six months in advance.
Equal Opportunity Employer