An Analysis of Total Nonresponse in the 1993-94 Schools and Staffing Survey (SASS)

Book excerpt: This technical report presents results of an analysis of unit response rates for the components of the 1993-94 Schools and Staffing Surveys (SASS). The study was motivated by the general need to evaluate and improve the quality of SASS data and to identify potential sources of nonsampling error associated with nonresponse in SASS. As background, the report describes the survey design and nonresponse adjustment procedures for each of the components of the SASS. A primary focus of the analysis is to compare the response rates for known characteristics of schools, administrators, teachers, school districts, libraries, librarians, and student records, and to assess the extent and pattern of these differences. Where possible findings from the 1993-94 analyses are compared to results from an exploratory analysis of response rates from the 1990-91 SASS. In addition, the hierarchical nature of response rates is examined, and a multivariate model of unit response is developed for one of the SASS components (public schools) to explain the relationship between these factors and the level of unit response. The following chapters are included: (1) "Overview"; (2) "SASS Core Components"; (3) "New SASS Components"; (4) "Summary of Significance Tests"; (5) "Hierarchical and Cross-Classified Testing"; (6) "Measurement of the Sampling Frame and Cooperation Rates"; (7) "Nonresponse Modeling for the Public School Component"; and (8) "Highlights and Recommendations." Appendixes contain response rate tables and a description of tests of association between response status and characteristics. (Contains 80 tables, 64 figures, and 34 references.) (SLD)