The target population is the universe of elementary and secondary school teachers who teach in public and private schools in the 50 states and the District of Columbia, in schools that had any of grades K–12 during the school year of the last SASS administration. This population is divided into two components: those who left teaching after that school year (former teachers) and those who continued teaching (current teachers).
The TFS sample of teachers includes those who left the position of a K–12 teacher in the year after SASS (leavers). It also includes those who continued to teach students in any of grades pre-K–12 or in comparable ungraded levels, including teachers who remained in the same school as in the previous year (stayers) and those who changed schools (movers). Prekindergarten is included so that sampled teachers who change assignments from teaching students in any of grades K–12 to teaching only prekindergarten students would not be considered leavers.
In SASS, the sampling frame for public schools is an adjusted version of the NCES Common Core of Data (CCD), and the sampling frame for private schools is a modified version of the NCES Private School Universe Survey (PSS). The sampling frame for the SASS teacher questionnaire consists of lists of teachers provided by schools in the SASS sample. A teacher is defined as a staff member who taught a regularly scheduled class to students in any of grades K–12 or comparable ungraded levels.
TFS surveys a sample of teachers who completed interviews in the previous year's SASS. The TFS sample is a stratified sample that is allocated to allow comparisons of teachers by five variables: status (stayers, movers, leavers, and unknown); school type (traditional public, public charter, and private); experience (new and experienced); grade level (elementary, middle, and secondary); and race/ethnicity (White, non-Hispanic, Black, Hispanic, and all other races/ethnicities). In the 2012–13 TFS administration, all responding SASS teachers in public schools who indicated that their first year of teaching was 2011 or 2012 were included in the sample. All other SASS responding teachers were stratified by the five variables in the following order: school sector, teacher status, experience, teacher's grade level, and race/ethnicity.
Within each TFS stratum, teachers with completed interviews in SASS are sorted by a measure of size (the SASS teacher initial basic weight, which is the inverse of the probability of selection prior to any corrections identified during data collection), main subject taught as reported by the teacher in SASS (i.e., special education, general elementary, mathematics, science, English/language arts, social studies, vocational/technical, and other), Census region, SASS private school affiliation stratum (for private school teachers only), school locale (based on the 1990 Census geography), school enrollment, and SASS teacher control number.
After teachers are sorted using the above variables, they are selected within each stratum using a systematic probability proportional to size (PPS) sampling procedure. Any teacher with a measure of size greater than the sampling interval is included in the sample with certainty (i.e., automatically included). Since TFS selection probabilities are not conditioned on anything, the selected sample sizes equal the allocated sample size.
The 2012–13 TFS data collection was an online collection, followed by e-mail and telephone reminders, a hard-copy mailing, and telephone follow-up. The U.S. Census Bureau was the data collection agent.
Reference dates. Most data items refer to teacher status at the time of questionnaire completion. Some items refer to the past school year, the past 12 months, or the next school year.
Data collection. In the fall of the year the survey is administered, the Census Bureau mails a Teacher Status Form to each school that had at least one teacher who participated in the previous year's SASS. On this form, the school principal (or other knowledgeable staff member) is asked to report the current occupational status of each teacher listed by indicating whether that teacher (1) is still at the school in a teaching or nonteaching capacity; or (2) has left the school to teach elsewhere or to enter a nonteaching occupation. If school staff indicates that a sample teacher has moved, and the teacher did not provide contact information on his or her SASS questionnaire, the Census Bureau tries to obtain the correct home address from the U.S. Postal Service.
Data collection for the 2012–13 TFS began as part of a sample selection operation in the fall of 2012 with the mailing of the Teacher Status Form to each school that had at least one teacher who had completed a Teacher Questionnaire in the 2011–12 SASS.
Data collection activities for current and former teachers began in January 2013. Initial contact included a letter to sampled teachers inviting their participation in TFS using an internet instrument; an e-mail to those sampled teachers who provided e-mail addresses inviting their participation in TFS; and, for Amish and Mennonite teachers, an invitation letter with a paper questionnaire. Following these initial contacts, a reminder letter and intermittent e-mails were sent to teachers who had not yet returned a completed questionnaire. Telephone follow-up began in February 2013 and was followed by additional reminder e-mails (six total) and second, third, and fourth reminder letters. The third and fourth reminder letters contained paper questionnaires. Reminder letters were also sent to teachers who had partially completed an online questionnaire, encouraging them to finish the survey. Data collection closed in July 2013.
Editing. Surveys undergo several stages of editing. TFS data that were provided on hard-copy versions of questionnaires are converted from paper to electronic format using manual data keying. All keyed entries are 100 percent verified by the keying staff, meaning that each field is keyed twice and the results are compared automatically for discrepancies and, subsequently, verified. All survey data are then reformatted into SAS datasets in order to begin the extensive preliminary data review process. During this stage, analysts split the TFS data into two files: a former teacher file (for leavers) and a current teacher file (for stayers and movers).
The next step is to make a preliminary determination of each case's interview status recode (ISR) value; that is, whether it is an interview, a noninterview, or out-of-scope for the survey. Records classified as interviews are submitted to a series of computer edits: range checks, consistency edits, and blanking edits. Next, the records undergo a final edit to determine whether the case is eligible to be included in the survey and, if so, whether sufficient data have been collected for the case to be classified as a completed interview. A final ISR value is then assigned to each case as a result of this edit.
Estimates from TFS sample data are produced using weighting and imputation procedures.
Weighting. The general purpose of weighting is to scale up the sample estimates to represent the target survey population. In TFS, the steps for weighting types of respondents are similar to those used for SASS. For TFS, a base weight (the inverse of the sampled teacher's probability of selection) is used as the starting point. Then, a weighting adjustment is applied that reflects the impact of the SASS teacher weighting procedure. Next, a nonresponse adjustment factor is calculated and applied using information known about the respondents from the sampling frame data. Finally, a ratio adjustment factor is calculated and applied to the sample to adjust the sample totals to frame totals in order to reduce sampling variability. The product of these factors is the final weight for TFS.
Imputation. In all administrations of TFS, all items missing values are imputed for records classified as interviews. In order to fill these items with data, questionnaires are put through three independent stages of imputation. The first stage involves using items from the same TFS questionnaire or items from the corresponding SASS school or teacher questionnaire to impute the missing data. In the second stage, any remaining unanswered items are imputed using "hot-deck" imputation (in which donor records are established and used to impute data). In the third and final stage, any remaining unanswered items are imputed clerically by Census Bureau analysts. The third stage is necessary when there is no available donor or the value imputed by computer is inconsistent with values in other items.
The National Teacher and Principal Survey (NTPS) is a redesign of the Schools and Staffing Survey (SASS), which the National Center for Education Statistics (NCES) conducted from1987 to 2011. The design of the NTPS is a product of three key goals coming out of the SASS program: flexibility, timeliness, and integration with other Department of Education collections. The NTPS collects data on core topics including teacher and principal preparation, classes taught, school characteristics, and demographics of the teacher and principal labor force every two years. In addition, each administration of NTPS contains rotating modules on important education topics such as: professional development, working conditions, and evaluation. This approach allows policy makers and researchers to assess trends on both stable and dynamic topics. The NTPS utilizes primarily internet and paper data collection instruments.
Although the NTPS will have a different structure and sample than previous SASS administrations, the focus remains on schools and their teachers and principals. The content of the 2011–12 SASS forms the basis of the NTPS content, though many questions will be shifted to different questionnaire instruments or will be answered through the use of extant data sources. Cross sectional analysis of trends is possible for SASS items that have been maintained in NTPS. TFS will be administered to a sample of teachers who completed the NTPS in the previous year.