OMB83C Change Memo v.27

Memorandum United States Department of Education

Institute of Education Sciences

National Center for Education Statistics

DATE: November 12, 2019

TO: Robert Sivinski, OMB

THROUGH: Kashka Kubzdela, OMB Liaison, NCES

FROM: Maura Spiegelman, Project Officer for the National Teacher and Principal Survey, NCES

SUBJECT: National Teacher and Principal Survey of 2020-2021 (NTPS 2020-21) Preliminary Field Activities Update Change Request (OMB# 1850-0598 v.27)

The National Teacher and Principal Survey (NTPS), conducted every two to three years by the National Center for Education Statistics (NCES), is a system of related questionnaires that provides descriptive data on the context of elementary and secondary education. Redesigned from the Schools and Staffing Survey (SASS) with a focus on flexibility, timeliness, and integration with other ED data, the NTPS system allows for school, principal, and teacher characteristics to be analyzed in relation to one another. NTPS is an in-depth, nationally representative survey of first through twelfth grade public and private school teachers, principals, and schools. Kindergarten teachers in schools with at least a first grade are also surveyed. NTPS utilizes core content and a series of rotating modules to allow timely collection of important education trends as well as trend analysis. Topics covered include characteristics of teachers, principals, schools, teacher training opportunities, retention, retirement, hiring, and shortages. The NTPS 2019-20 preliminary activities were approved in July 2019 (OMB# 1850-0598 v.26). This request is to update the details of the NTPS 2019-20 sampling plan. This request does not introduce changes to the estimated respondent burden or the costs to the federal government.

Revisions made to Part B are summarized below (red font reflects text deletions and additions).

1. Page 4 – updates to the sampling universe based on more recent universe data and newly adopted approaches to defining school levels:

   1. We no longer anticipate having to manage overlap with schools sampled for SSOCS, so we have removed the reference to it from this package;

   2. We have updated the respondent universe with the 2017-18 Common Core of Data (CCD); and

   3. ED revised the way that it categorizes public schools by level, and we have updated Table 1 and its accompanying matter to reflect that.

The most recent final Common Core of Data (CCD) file available from NCES at the time of sampling in spring 2020 will be used to construct the public school frame.¹ The respondent universe for charter schools will be identified as those public charter schools that meet the NTPS definition of an eligible school found on the CCD. The universe has been adjusted to remove kindergarten-terminal schools, which are not eligible for NTPS. Table 1 presents the number of public schools on the 2017-18 NTPS public school universe, which are based on the 2014-15 CCD, by urbanicity and school level. The CCD that will be used to construct the sample for NTPS 2020-21 is not yet available at the time of this submittal. The NTPS 2020-21 school sample will be drawn in April-May 2020 and we will begin to contact sampled schools in June 2020. Another NCES survey, the 2020 Survey of School Crime and Safety (SSOCS), will be collecting data from schools from February through July 2020. We currently do not anticipate a significant overlap between the 2020 SSOCS and NTPS 2020-21 samples, however, once drawn, if the NTPS 2020-21 sample proves to overlap significantly with the 2020 SSOCS sample, NCES will consider adjusting the NTPS 2020-21 probabilities based on the 2020 SSOCS sample. In that case, NCES would submit a change request to OMB in spring 2020 explaining the adjustment.

Table 1. Estimated respondent universe by school level and urbanicity for the proposed NTPS 2020-21 public school sample, based on the 2017-18 NTPS Public School Universe

School level2
Region	Primary	Middle	High	Combined	Total
Central City	15,308	3,699	5,407	1,727	26,141
Suburban	17,933	5,136	5,901	1,220	30,190
Town	6,138	2,340	3,481	876	12,835
Rural	12,221	3,189	6,014	3,538	24,962
Total	51,600	14,364	20,803	7,361	94,128

SOURCE: 2017-18 NTPS; 2014-15 CCD.

Table 1. Respondent universe by school level and urbanicity for the proposed public school sample, based on the 2017-18 Common Core of Data (CCD)

School level2
Region	Primary	Middle	High	Combined	Total
Central City	15,265	3,853	5,998	948	26,064
Suburban	17,742	5,495	6,150	801	30,188
Town	5,714	2,392	3,623	507	12,236
Rural	11,794	3,493	6,655	2,824	24,766
Total	50,515	15,233	22,426	5,080	93,254

SOURCE: 2017-18 CCD.

If the 2018-19 CCD is not available as of early January in 2020, the most recently available CCD as of that date will be used instead.

² The definition for the school level categories may change for the NTPS 2020-21, pending a final decision from NCES about creating a consistent definition across NCES studies. If this change occurs and impacts the final sample design for the 2020-21 NTPS, NCES will submit a change request to this submission by April 2020.

² Rather than defining school level for public schools by the highest and lowest grades offered at the school, starting with the 2017-18 CCD, ED revised the logic to derive LEVEL with the primary goal of recategorizing as many of the schools in the “Other” category (for NTPS, these schools are categorized as Combined) into categories that are more meaningful for data users. In particular, in Table 1, primary schools are schools in which grades K, 1, 2, 3, or 4 are offered and the number of primary grades is greater than number of middle grades; middle schools are schools in which grades 5, 6, 7, or 8 are offered and the number of middle grades is greater than number of elementary or secondary grades; high school are schools in which grades 9, 10, 11, or 12 are offered and the number of high grades is greater than number of middle grades; and combined schools are schools with both primary and high grades.

2. Page 5 – Because there was no change to definitions for school level categories for private schools, we have left Table 2 unchanged and deleted the mention of the possibility of that change.

Table 2. Respondent universe by school level and urbanicity for the proposed private school sample, based on the 2017-18 Private School Survey (PSS)

School level3
Region	Elementary	Secondary	Combined	Total
Central City	4,621	1,070	2,555	8,246
Suburban	4,649	854	2,903	8,406
Town	1,234	148	731	2,113
Rural	2,735	447	1,882	5,064
Total	13,239	2,519	8,071	23,829

SOURCE: 2017-18 PSS.

³ The definition for the school level categories may change for the NTPS 2020-21, pending a final decision from NCES about creating a consistent definition across NCES studies. If this change occurs and impacts the final sample design for the 2020-21 NTPS, NCES will submit a change request to this submission by April 2020.

3. Page 8 – Because there is no longer any intention to change the oversampling plan for private schools, we have deleted the mention of the possibility of doing so.

Sampling – Private Schools

The NTPS private school sample will be roughly the same as the private school survey from Schools and Staffing Survey (SASS) 2011-12: a school sample size of 3,000 and a number of completed school interviews of about 1,750. For the 2017-18 NTPS, the sample size was 4,000 (expecting 2,266 completed school interviews) to achieve sufficient power for an embedded experiment.

The sampling plan oversamples as follows, and any changes to oversampling, should they occur will be provided in a change request by April 2020:

• Elementary schools are sampled at a rate proportional to the measure of size;

• Secondary schools are sampled at a rate proportional to 3 times the measure of size;

• Combined schools are sampled at a rate proportional to 1.2 times the measure of size;

• The oversampling rates for Nonsectarian schools are increased by an additional factor of 1.25 (e.g. the oversampling rate for Nonsectarian secondary schools is 3 ×1.25 = 3.75); and

• The oversampling rates for Baptist schools are increased by an additional factor of 1.1 (e.g. the oversampling rate for Baptist combined schools is 1.2 ×1.1 = 1.32).