| Questionnaire Design | Imputation |
| Sampling Frame | Weighting |
| Sample Design | Response Rates |
| Data Collection | Reinterviews |
| Data Editing | Manuals and Technical Reports |
NCES integrated several surveys, the Private School Survey, the Public School Survey and the Teacher Demand and Shortage Survey (SASS), which had previously been conducted separately. Principal and teacher surveys were added to make up the survey system called the Schools and Staffing Survey (SASS). Prior to the first round of SASS, NCES conducted a pretest of the content, followed by in-depth reinterviews in preparation for the first full-scale SASS.
School samples were selected from the Quality Education Data File (QED). Duplicate schools and schools that were out-of-scope were eliminated from the file before sampling. To improve coverage of private schools, the QED list was supplemented with an area frame, using 75 sample areas, and lists of schools from 17 non-Catholic organizations.
The main sample design objectives were to provide estimates of acceptable precision for the specified domains of analysis. Estimates can be made for public schools by state, by school level within state, by schools with more than 25 percent Indian enrollment and by Bureau of Indian Affairs-funded schools. Estimates for private schools could be made by affiliation and by school level.
Another objective was to balance the requirements of the school sample against the requirements of the samples of LEAs and teachers. The sample design called for selecting schools with a probability proportionate to the square root of the school's teacher size and, within each stratum, selecting a fixed number of teachers, subject to constraints on the total number of teachers selected in a school.
The SASS sample design also sought to control sample overlap between SASS and other NCES school surveys. To minimize the response burden on individual schools, the sample selection procedure used for SASS minimized the overlap among the SASS, the National Assessment of Educational Progress (NAEP) and the National Educational Longitudinal Study of 1988 (NELS) school sample while maintaining the initial probability of selection for each school in SASS averaged over all possible school samples for the three surveys.
The U.S. Census Bureau performed the data collection and began by sending an advance letter to LEAs of sampled schools in the fall. During November and December 1987, schools were asked to provide lists of their teachers for use in selecting the teacher sample. The first mailing of questionnaires to districts, principals and teachers took place in February 1988. Approximately six weeks after the initial mailing, a second questionnaire was sent to those who had not returned the first questionnaire. Telephone follow-ups to nonrespondents began a month later and, whenever possible, interviews were completed by phone.
The U.S. Census Bureau performed the data processing. Each questionnaire was coded according to its status - for example, whether the questionnaire contained a completed interview, a respondent refused to complete it, a school district merged with another district, or a school closed. The next step was to make a preliminary determination of each case's interview status, i.e., whether it was an interview, a noninterview, or if the respondent was out-of-scope (for example, if a sampled school had closed). A computer pre-edit program generated a list of cases where problems occurred as defined by edit specifications, depending on each survey. When necessary, the Census Bureau used extensive follow-up to collect data on key items from schools.
After the completion of range, consistency and blanking edits, the records were put through another edit to make a final determination of whether the case was eligible for the survey and, if so, whether there were sufficient data for the case to be classified as an interview. A final interview status recode value was assigned to each case as a result of the edit.
SASS used several methods to impute values for questionnaire items that respondents did not answer. These methods included the following: (1) deductive imputation (or using data from other items on the questionnaire); (2) hot deck imputation (extracting data from the record of a case with similar characteristics); and clerical imputation (reviewing the data record, the sample file record or other sources before deriving an entry consistent with the existing data).
Weighting of the sample units was carried out to produce national and state estimates for public schools, LEAs, administrators and teachers. Private schools, administrators and teachers were weighted to produce national and affiliation estimates. The weighting procedures have three purposes: to take account of the school's selection probabilities; to reduce biases that may result from unit nonresponse; and to make use of available information from external sources to improve the precision of sample estimates.
Weighted response rates are defined as the number of in-scope responding questionnaires divided by the number of in-scope sample cases, using the basic weight (inverse of the probability of selection) of the record.
| Component | Sample size | Weighted response rate |
|---|---|---|
| Public Schools | ||
| Teacher Demand and Shortage | 5,594 | 90.8% |
| Administrator | 9,317 | 94.4% |
| School | 9,317 | 91.9% |
| Teacher | 56,242 | 86.4% |
| Private Schools | ||
| Teacher Demand and Shortage | 3,513 | 66.0% |
| Administrator | 3,513 | 79.3% |
| School | 3,513 | 78.6% |
| Teacher | 11,529 | 79.1% |
SASS conducted a reinterview by telephone of about ten percent of schools in the sample. The reinterview consisted of selected questions from the school and administrator questionnaires. Reinterview results were analyzed using one index of inconsistency for items with a dichotomous response that is accounted for by response variance. A separate index of inconsistency was used for response items with more than two response categories.
| NCES 91136: | User's Manual: 1987-88 Schools and Staffing Survey - Public School Data |
| NCES 91137: | User's Manual: 1987-88 Schools and Staffing Survey - Public School Administrator Questionnaire Data |
| NCES 91127: | 1987-88 Schools and Staffing Survey: Sample Design and Estimation |
| NCES 9402: | Generalized Variance Estimate for Schools and Staffing Survey (SASS) |
| NCES 94340: | A Quality Profile for SASS: Aspects of the Quality of Data in the Schools and Staffing Surveys |
| NCES 2000308: | Quality Profile for SASS Rounds 1-3: 1987-1995, Aspects of the Quality of Data in the Schools and Staffing Surveys (SASS) |
