Sampling and data collection methods in the National Evaluation of School Nutrition Programs.

A nested sampling design is described in which school districts are selected, schools within those districts that cover grades one through 12 are selected, and finally students within each of the grades are selected. The article describes how nutritional status information is collected from each student, economic and food expenditure data are collected from the students' parents, and information on food services is collected from district and school administrators. Sample weighting and adjustments for nonresponse bias are discussed in the appendix.