Design Effects and Generalized Variance Functions for the 1990-91 Schools and Staffing Survey (SASS).: Technical report

Sameena Salvucci

Jan 1995 · U.S. Department of Education, Office of Educational Research and Improvement, National Center for Education Statistics

About this ebook

This technical report provides the results of a study on the calculation and use of generalized variance functions (GVFs) and design effects for the 1990-91 Schools and Staffing Survey (SASS). The SASS is a periodic integrated system of sample surveys conducted by the National Center for Education Statistics (NCES) that produces sampling variances different from those produced by simple random sampling with fixed sample size. An introduction provides a conceptual and contextual discussion, addressing the reasoning for using general analytical techniques to calculate standard errors for complex survey data. The second section describes groups of survey statistics users might want to select. In Section 3, the procedure and computational formulas for calculating the design effects for the most common types of estimates (totals, means, and proportions) are presented. Section 4 then gives the technical details of GVF fitting. The design effect and GVF results obtained from the 1990-92 SASS are presented in Section 5. Section 6 discusses possible next steps in calculating standard errors for NCES surveys. (Contains 8 tables, 12 figures, and 25 references.) (SLD)