Technical Notes

Overview of the NTPS

The National Teacher and Principal Survey (NTPS) is sponsored by the National Center for Education Statistics (NCES) of the Institute of Education Sciences within the U.S. Department of Education and is conducted by the U.S. Census Bureau. NTPS is a nationally representative sample survey of public and private K–12 schools, principals, and teachers in the 50 states and the District of Columbia. The NTPS was first conducted during the 2015–16 school year, and 2017–18 is the second NTPS collection.

The 2017–18 NTPS consisted of questionnaires for six types of respondents: public schools, private schools, public school principals, private school principals, public school teachers, and private school teachers. For the content of the questionnaires, see https://nces.ed.gov/surveys/ntps/question1718.asp. The information can be linked across teachers, principals, and schools by each sector (public and private). There is a separate data file for each type of respondent by sector (public school, private school, public school principal, private school principal, public school teacher, and private school teacher). For public schools, NTPS was designed to produce national, regional, and state estimates for elementary and secondary schools, principals, and teachers, including public charter schools and the principals and teachers within them.

For additional information on the specific NTPS-related topics discussed in this Technical Notes section, consult the Survey Documentation for the 2017–18 National Teacher and Principal Survey (Cox et al. forthcoming) or the User’s Manual for the 2017–18 National Teacher and Principal Survey Volumes 1–4 (Goldring et al. 2019), as well as the report from the 2017–18 NTPS on characteristics of public and private school principals (Taie and Goldring 2019). To access additional general information on NTPS or for electronic copies of the questionnaires, go to the NTPS home page (https://nces.ed.gov/surveys/ntps).

Sampling Frames and Sample Selection

The starting point for the 2017–18 NTPS public school sampling frame was the 2014–15 Common Core of Data (CCD) Nonfiscal School Universe data file.³ The sampling frame was adjusted from the CCD to fit the definition of a school eligible for NTPS. To be eligible for NTPS, a school was defined as an institution or part of an institution that provides instruction to students, has one or more teachers to provide instruction, serves students in one or more of grades 1–12 or the ungraded equivalent, and is located in one or more buildings apart from a private home.

The 2017–18 NTPS universe of public schools is confined to the 50 states plus the District of Columbia and excludes the other jurisdictions, Department of Defense overseas schools, and CCD schools that do not offer teacher-provided classroom instruction in grades 1–12 or the ungraded equivalent. Since CCD and NTPS differ in scope and their definition of a school, some records were deleted, added, or modified to provide better coverage and a more efficient sample design for NTPS. For a detailed list of frame modifications, see the Survey Documentation for the 2017–18 National Teacher and Principal Survey (Cox et al. forthcoming). After deleting, collapsing, and adding school records, the 2017–18 NTPS public school sampling frame consisted of about 86,800 traditional public schools and 6,800 public charter schools.

NTPS uses a systematic, probability proportionate to size sample, where size is defined to be the square root of the number of full-time-equivalent teachers in the school. Schools were oversampled based on school grade level, state, poverty status, enrollment, collapsed urbanicity, and charter status. These sampling procedures resulted in a total public school sample of about 10,580 schools (about 9,180 traditional public schools and 1,400 public charter schools). The principal or school head of each sampled school was selected.

Data Collection and Unit Response Rates

In 2017–18, NTPS employed a combined mail-based and internet survey approach, with subsequent telephone and in-person follow-up. The web was the primary mode of data collection for all questionnaire types for the 2017–18 NTPS. Paper questionnaires were introduced in the later mailings. Data collection began in September 2017 and ended in August 2018.

Unit response rates. The responses were weighted to produce national estimates. The weights were designed to reflect the probabilities of selection and were adjusted for differential nonresponse. The unit 9 response rate indicates the percentage of sampled cases that met the definition of a complete interview. The weighted NTPS unit response rate was produced by dividing the weighted number of respondents who completed questionnaires by the weighted number of eligible sampled cases, using the initial base weight (the inverse of the probability of selection)⁴ The weighted response rate using the initial base weight was 70.2 percent for public school principals.

Unit nonresponse bias analysis. Because the NCES Statistical Standards (4-4) require analysis of nonresponse bias for any survey stage with a base-weighted response rate less than 85 percent, the NTPS principal files were evaluated for potential bias. For further information on unit response rates and nonresponse bias analysis, see the Survey Documentation for the 2017–18 National Teacher and Principal Survey (Cox et al. forthcoming).

Variables Used and Item Response Rates

The variables from the survey used in this Statistics in Brief are listed in the text box below, along with the variable names used in the data file and the weighted item response rates. The analysis variables (with variable names) are defined after the text box. For additional information about the variables, see the User’s Manual for the 2017–18 National Teacher and Principal Survey Volumes 1–4 (Goldring et al. 2019).

Principals reported in 2017–18 for the previous school year (2016–17) about their participation in professional development activities while a principal at their current school to help understand how it might be affecting principal practices during the 2017–18 school year. Because principals were reporting about professional development activities undertaken as a principal at their current school in the previous school year, the analyses presented in this report are for principals with at least one year of experience at their current school.

Definitions of analysis variables. This report focuses on national estimates and bivariate relationships between the analysis variables and questionnaire variables. The following variables were used for analysis in this report.

Variable label	Variable name	Response rate
Professional development participation, any activities	A2700	99.2
Professional development evaluation consideration	A2703	98.5
Professional development, university courses	A2704	98.3
Professional development, visit other schools	A2705	98.2
Professional development, coaching of principals	A2706	98.3
Professional development, principal network	A2707	98.3
Professional development, workshops, conferences or training as a presenter	A2708	98.4
Professional development, other workshops	A2709	98.3
Professional development participation–analyzing/interpreting student achievement data	A2710	98.4
Professional development participation–human resource management	A2711	98.1
Professional development participation–student motivation and engagement	A2712	98.2
Professional development participation–technology instructional support	A2713	98.2
Professional development participation–management and policy	A2714	97.7
Professional development participation–improvement planning	A2715	98.2
Professional development participation–student social services	A2716	97.8
Professional development participation–safety/school climate	A2717	98.2
Professional development participation–effective instructional support	A2718	98.4
Total years of experience as a school principal	A0104	99.7
Years of experience at the current school	A0105	99.9

Urban-centric school locale code (URBANS12): Taken from the Public School Data Files, URBANS12 is a created variable collapsed from the 12 category urban-centric school locale code (SLOCP12) which was updated to incorporate Census population and geography information and recoded into four categories, as follows:

• City: includes city, large; city, midsize; city, small.
• Suburban: includes suburb, large; suburb, midsize; suburb, small.
• Town: includes town, fringe; town, distant; town, remote.
• Rural: includes rural, fringe; rural, distant; rural, remote.

Total years of experience as a school principal (A0104): Taken from the public school principal questionnaire, A0104 is a continuous variable recoded as a categorical variable into three categories: 1 or 2 years, 3 to 9 years, and 10 years or more.

Total years of experience as a school principal (A0104): Taken from the public school principal questionnaire, A0104 is a continuous variable recoded as a categorical variable into three categories: 1 or 2 years, 3 to 9 years, and 10 years or more.

Sources of Error in Estimates

Survey estimate is subject to two types of errors: nonsampling and sampling. Nonsampling errors are attributed to many sources, including definitional difficulties, the inability or unwillingness of respondents to provide correct information, differences in the interpretation of questions, an inability to recall information, errors made in collection (e.g., in recording or coding the data), errors made in processing the data, and errors made in estimating values for missing data. Quality control and edit procedures were used to reduce errors made by respondents, coders, and interviewers. In contrast, sampling errors result from the collection of data from a sample of the population rather than the full target population, and estimates of the magnitude of sampling error for NTPS data can be derived or calculated. Because of both types of errors, the survey estimates may differ from the values that would be obtained from the target population using the same questionnaire, instructions, and field representatives.

Statistical Procedures

Comparisons of estimates in the text have been tested for statistical significance using the Student’s t statistic to ensure that the differences are larger than those that might be expected due to sampling variation. All statements cited in the text are statistically significant at the p < .05 significance level (indicating that there is less than a 5 percent chance that the difference occurred by chance), using two-tailed statistical tests. Student’s t values were computed to test the difference between estimates with the following formula:

where E₁ and E₂ are the estimates to be compared and se₁ and se₂ are their corresponding standard errors. The threshold for determining significance at the 95 percent level for all comparisons in this report was t = 1.96. The standard errors of the estimates for difference subpopulations can vary considerably and should be taken into account when drawing conclusions about the estimates being compared. No adjustments for multiple comparisons were made in the analyses presented in this report.

³ For more information about CCD, see https://nces.ed.gov/ccd/.
⁴ For the formula used to calculate the unit response rate, see 2012 Revision of NCES Statistical Standards: Final (NCES 2014-097), https://nces.ed.gov/pubsearch/pubsinfo.asp?pubid=2014097.