Methods of Selecting Samples in Multiple Surveys to Reduce Respondent Burden

The National Agricultural Statistics Service (NASS) surveys the United States population of farm operators numerous times each year. The list components of these surveys are conducted using independent designs, each stratified differently. By chance, NASS samples some farm operators in multiple surveys, producing a respondent burden concern. Two methods are proposed that reduce this type of respondent burden. The first method uses linear integer programming to minimize the expected respondent burden. The second method samples by any current sampling scheme, then, within classes of similar farm operations, it minimizes the number of times that NASS samples a farm operation for several surveys. The second method reduces the number of times that a respondent is contacted twice or more within a survey year by about 70 percent. The first method will reduce this type of burden even further.