Part B QRIS 2017-2020 System Clearance

Quick Response Information System

(QRIS) 2017-2020 System Clearance

Supporting Statement Part B

OMB # 1850-0733 v.34

National Center for Education Statistics (NCES)

U.S. Department of Education

Institute of Education Sciences

Washington, DC

July 2017

Section B. Description of Statistical Methodology

B.1. Respondent Universe and Sample Design

The Quick Response Information System (QRIS) consists of the Fast Response Survey System (FRSS) and the Postsecondary Education Quick Information System (PEQIS). The respondent universes and sample designs for FRSS and PEQIS that are covered by this generic clearance request are described below.

Fast Response Survey System

FRSS is designed to conduct brief surveys of state education agencies, public elementary and secondary schools, private elementary and secondary schools, public school districts, and school and public libraries. In the sections that follow, the approaches that will be used to design and select samples from the various sectors of interest are described. State education agencies are not discussed here, given that they would be surveyed on a 100-percent basis.

Efficient probability sampling designs are an integral part of the FRSS. For those sectors that are surveyed frequently in FRSS (e.g., school districts and public schools), a general approach to sampling is designed and modified as necessary to meet the specific goals of the study. For example, for many FRSS surveys, stratified probability-proportionate-to-size (PPS) sampling designs are used to ensure that (a) estimates with specified levels of sampling precision can be obtained for key subgroups of interest, and (b) both categorical and quantitative variables can be estimated reliably.

For some of the less frequently surveyed sectors, it is desirable to select a sample that is tailored to the specific needs of the individual survey. This specialization will be most efficient when pertinent data are available for sample selection purposes. Examples of situations that will necessitate designing and drawing special-purpose samples include surveys that are restricted to a particular subgroup (e.g., districts with summer migrant education programs, districts with English learner programs, or specific types of postsecondary institutions), surveys that require concurrent fielding of different questionnaires in the same sector (e.g., library services for children and young adults), and related surveys involving different sets of respondents that must be linked through an overlapping sample design (e.g., the three surveys on educational technology conducted in 2008-09).

Public Elementary and Secondary Schools

Since each new survey to be conducted under the FRSS will have unique analytic requirements, it is not possible to specify the exact form of the sample design that will be appropriate and efficient for a particular study. This can only be done after the study objectives and data requirements have been clearly delineated. However, past experience suggests that many design features that have been employed successfully in prior studies will also be applicable to future FRSS studies.

First, the frame to be used to select the required samples will be constructed from the most recent NCES CCD Public School Universe File available at that time. As of this submission, the most current file available is for the 2014-15 school year. As shown in Table 1, over 87,000 regular public schools are included in the 2014-15 CCD file, including about 51,000 elementary schools, 16,000 middle schools and about 20,000 secondary /combined schools. The school-level variables that are available in the CCD Public School Universe File include type of school, enrollment, number of teachers, instructional level, grade span, type of locale, region, race/ethnicity of students, eligibility for free/reduced-price lunch, and others. Such information is critical for designing efficient samples for the QRIS. As in prior studies, it is anticipated that instructional level, enrollment size, and one or more other variables will be used to define the primary strata for sampling purposes. It should be noted, however, that it may be necessary to merge information from other data sources (e.g., the National Teacher and Principal Survey - NTPS) to obtain relevant auxiliary variables useful for stratification or weighting purposes. Also, some variables in the CCD file may have missing values (e.g., enrollment by race/ethnicity and the number of students eligible for free/reduced-price lunch). If used for stratification purposes, the missing values will be imputed or treated as a separate category.

Next, decisions must be made on how to allocate the total sample to the primary strata. For example, proportional allocation (in which the sample is allocated to a stratum in proportion to the number of schools in the stratum) is generally efficient for estimating the percentage of schools having a given characteristic (e.g., percentage of schools with students enrolled in dual credit courses). On the other hand, allocation in proportion to enrollment is generally more effective for estimation of aggregate statistics that are correlated with enrollment (e.g., the number of enrollments of high school students in dual credit courses). Although many prior FRSS samples have been designed primarily to estimate the latter type of aggregate statistics, there has generally been an equal interest in estimating percentages. For this reason, allocation in proportion to the square root of enrollment offers a compromise solution that is often used in many QRIS surveys. Again, the specific requirements of a particular survey will dictate how the sample is allocated to strata.

Table 1. Number and enrollment of regular public elementary and secondary schools in the 2014-15 CCD Public School Universe file, by instructional level

Instructional level	Number of schools*		Total enrollment
Elementary		51,007	23,934,296
Middle		16,421	9,526,183
Secondary/combined		19,795	15,280,189
TOTAL		87,223	48,740,668

* Counts reflect only regular schools with non-zero or missing membership in the 2014-15 CCD file. For example, special education, vocational education, and other alternative schools are excluded.

Within the primary sampling strata, schools can also be selected at varying rates depending on the goals of the study. This can be accomplished either by using a PPS systematic sampling algorithm (Hansen, Hurwitz, and Madow, 1953), or by forming appropriate size classes (strata) of equal aggregate measure of size and selecting equal numbers of schools from each size stratum. The latter procedure is often used for reasons of simplicity.

The sample size for a typical FRSS survey of public schools is approximately 1,200. This will yield a sample of 1,080 respondents assuming a response rate of 90 percent. Based on the illustrative sampling allocation given in Table 2, the standard error of an estimated percentage for the total sample can be expected to be in the range of 1.4 to 1.7 percent. For a 50 percent item, a standard error of 2 percent corresponds to a coefficient of variation (cv) or relative standard error (RSE) of 4 percent. Moreover, the sample is large enough to provide reasonably reliable estimates for broad subsets of the population (e.g., one-way classifications by type of locale or size class). Table 2 illustrates the levels of precision that can be expected for a sample of 1,080 responding public schools under the given allocation. The standard errors presented in this table are given for illustrative purposes only; the actual standard errors to be obtained for any given QRIS survey may be smaller or larger than those shown, and will differ depending on how the sample is allocated to strata to meet the specified goals of the study.

Table 2. Illustrative standard errors for an estimated percentage based on a sample of 1,080 public schools, by instructional level, Census region, type of locale, and enrollment size

		Estimated percentage equal to:
Subset of sample	Expected sample size*	80%	50%	30%
Total sample	1,080	1.4	1.7	1.6
Instructional level
Elementary	450	1.9	2.4	2.2
Middle	198	3.0	3.7	3.4
Secondary/combined	432	2.2	2.7	2.5
Region
Northeast	198	3.2	4.0	3.7
Southeast	257	2.8	3.5	3.2
Central	256	2.8	3.5	3.2
West	369	2.3	2.9	2.7
Type of locale
Cities	310	2.5	3.2	2.9
Suburban	410	2.2	2.8	2.5
Towns	140	3.7	4.7	4.3
Rural areas	220	3.0	3.8	3.4
Enrollment size
Less than 300	161	3.2	4.0	3.7
300 to 499	245	2.6	3.3	3.0
500 to 999	411	2.0	2.5	2.3
1,000 or more	263	2.7	3.3	3.0

* Assumes an initial sample of 1,200 schools and a response rate of 90 percent.

Private Elementary and Secondary Schools

The general approach described previously for public schools will also apply to private schools. Samples of private schools will be selected from the most current NCES Private School Survey (PSS) Universe File. Note that the PSS frame consists of two parts: a list frame and an “area frame.” The latter is actually an area probability sample that represents schools not included in the list frame. Consequently, the schools in the area sample must be weighted to represent the unlisted portion of the private school universe. There are an estimated 33,000 private schools of all types represented in the 2013-14 PSS file, of which approximately 23,000 are coded as “regular” schools. In most surveys conducted under the QRIS, only the regular schools are of interest, However, other types of schools such as special education or vocational education schools can be included in the survey if appropriate. Table 3 summarizes the weighted distribution of regular private schools in the 2013-14 PSS frame by school size and private school typology. To select the sample of private schools, stratification by instructional level (elementary, secondary, and combined), private school typology (Catholic, other religious, and nonsectarian), and other characteristics may be employed. Within each primary stratum, the private school frame can be sorted by enrollment size, geographic region, or other characteristics available in the PSS file to induce additional implicit stratification. Depending on the goals of the survey, the total sample can be allocated to the primary strata in different ways (e.g., in proportion to the number of schools, in proportion to enrollment, or in proportion to the aggregate square root of the enrollment). Table 4 illustrates the levels of precision that might be expected for a sample of 1,080 responding regular private schools (based on an initial sample of 1,200 schools and a response rate of 90 percent).

Table 3. Number of schools in the 2013-14 PSS private school frame by school size and private school typology

			Private school typology
School size		Total*	Catholic	Other religious	Nonsectarian
Less than 50 students	8,325		243	7,329	753
50-149 students	5,845		1,456	3,623	766
150-299 students	4,914		2,568	1,861	485
300-499 students	2,309		1,261	713	336
500-749 students	1,088		566	343	179
750 or more students	722		301	242	179
Total	23,203		6,394	14,110	2,698

Note: Detail may not sum to totals because of rounding or missing values.

*Weighted count of regular schools. Special education, vocational education, alternative, and other types of nonregular schools are excluded. There are an estimated 10,000 nonregular schools in the 2013-14 PSS file.

Table 4. Illustrative standard errors for an estimated percentage based on a sample of 1,080 private schools, by instructional level, private school typology, type of locale, and school size

		Estimated percentage equal to:
Subset of sample	Expected sample size*	80%	50%	30%
Total sample	1,080	1.5	1.9	1.8
Instructional level
Elementary	522	2.0	2.5	2.3
Secondary	279	2.9	3.6	3.3
Combined	279	2.9	3.6	3.3
Private school typology
Catholic	458	2.2	2.7	2.5
Other religious	480	2.1	2.6	2.4
Nonsectarian	142	4.0	5.0	4.6
Type of locale City	418	2.4	3.0	2.8
Suburb	390	2.5	3.1	2.8
Town/rural	272	2.8	3.5	3.2
School size
Less than 150	364	2.1	2.7	2.5
150 to 499	468	1.9	2.4	2.2
500 or more	248	2.8	3.5	3.2

* Assumes an initial sample of 1,200 schools and a response rate of 90 percent.

Public School Districts (LEAs)

The sampling frame to be used to select the required samples of public school districts will be constructed from the most recent NCES CCD Public Elementary and Secondary Local Education Agency (LEA) Universe File available at that time. As shown in Table 5, about 13,000 regular public school districts are included in the 2014-15 CCD file. We anticipate that for most of the district surveys to be conducted under FRSS, stratification by size class, region, metropolitan status, poverty status, or other variables will be used to improve the precision of overall estimates, and to ensure minimum sample sizes for the analytic domains of interest. Further, we expect that a probability-proportionate-to-square-root-of-size design will be efficient for the goals of the study. However, this basic design will be modified as necessary to meet the particular objectives of a given study.

The sample size for a typical FRSS district survey is approximately 1,200 districts. Assuming a response rate of 90 percent, an initial sample of 1,200 districts will yield approximately 1,080 respondents. If the sample is allocated to strata in rough proportion to the aggregate square root of enrollment, the expected sample sizes would be those shown in Table 6. With these sample sizes, survey-based estimates for the total sample and for selected subgroups are expected to be reasonably precise. Table 6 illustrates the levels of precision to be expected for a sample of 1,080 responding public school districts

Table 5. Number of regular public school districts in the 2014-15 NCES CCD LEA Universe Survey by enrollment size class

Enrollment size class	Number of districts†
Less than 1,000		6,336
1,000 to 2,499		3,222
2,500 to 9,999		2,944
10,000 to 24,999		610
25,000 or more		288
TOTAL		13,400

†Includes district types 1 (local school district not part of a supervisory union) and 2 (local school district component of a supervisory union). Counts of districts with 0 or missing enrollment as reported in the CCD universe file are excluded.

Table 6. Illustrative standard errors for an estimated percentage based on a sample of 1,080 public school districts, by size class, region, and metropolitan status

Subset of sample	Expected sample size*	Estimated percentage equal to:
		80%	50%	30%
Total sample	1,080	1.4	1.8	1.6
District size class
Less than 2,500	453	2.1	2.6	2.4
2,500 to 9,999	367	2.3	2.9	2.6
10,000 or more	260	2.7	3.4	3.1
Region
Central	327	2.6	3.3	3.0
Northeast	220	3.2	4.0	3.7
Southeast	200	3.3	4.2	3.8
West	333	2.6	3.2	3.0
Type of locale
City	220	3.1	3.8	3.5
Suburban	236	3.0	3.7	3.4
Town	364	2.3	2.9	2.6
Rural	260	2.7	3.4	3.1

* Assumes an initial sample of 1,200 districts and a response rate of 90 percent.

Libraries

As an important provider of educational services, libraries (both school and public libraries) have been the focus of past research efforts. For example, public libraries have been surveyed in FRSS 66 (survey on programs for adults in public library outlets) and FRSS 47 (survey on library services for children and young adults). In addition, Westat, the current contractor to ED for FRSS and PEQIS surveys, conducted surveys of public and private school libraries for ED as part of the Assessment of the Role of School and Public Libraries in Support of the National Education Goals. If required for a QRIS survey, samples of public libraries will be drawn from the most recent Public Library Survey (PLS) universe file (the PLS is now conducted by the Institute of Museum and Library Services) available at that time.

Postsecondary Education Quick Information System (PEQIS)

The samples of postsecondary institutions for PEQIS studies will usually use the existing postsecondary education panel. This panel currently consists of a stratified probability sample of approximately 1,650 2- and 4-year public and private institutions. The sampling frame for the PEQIS panel was constructed from the 2009-10 Integrated Postsecondary Education Data System (IPEDS) Institutional Characteristics file. Institutions eligible for the PEQIS frame included 2-year and 4-year (including graduate-level) institutions that are both Title IV eligible and degree granting, and are located in the 50 states and the District of Columbia: a total of 4,485 institutions. The PEQIS sampling frame was stratified by instructional level (4-year, 2-year), control (public, private not-for-profit, private for-profit), highest level of offering (doctor’s/first-professional, master’s, bachelor’s, less than bachelor’s), and total enrollment. Within each of the strata, institutions were sorted by region (Northeast, Southeast, Central, West) and by whether the institution had a relatively high minority enrollment. The sample of 1,647 institutions was allocated to the strata in proportion to the aggregate square root of total enrollment. Institutions within a stratum were sampled with equal probabilities of selection. The PEQIS panel may be refreshed or a specialized postsecondary sample (e.g., all 2-year public institutions, or all 4-year institutions that award degrees in education) may be selected prior to a subsequent PEQIS survey to meet the particular objectives of a given study. Table 7 summarizes the PEQIS universe counts and sample sizes by level and type of control.

Table 7. Distribution of institutions in the postsecondary (PEQIS) universe and panel by level and type of control

Level	Control	Number of institutions in PEQIS frame	Number of institutions in PEQIS panel
4-year	Public	672	442
	Private, not-for-profit	1,537	435
	Private, for profit	562	128
2-year	Public	999	544
	Private, not-for-profit	85	11
	Private, for profit	630	87
Total		4,485	1,647

Less-than-2-year and non-degree-granting institutions are not included in the PEQIS universe and panel because of the great volatility of these types of institutions. These schools, many of which are proprietary, open and close at a much faster rate than other kinds of postsecondary institutions. This means that any portion of the PEQIS panel allotted to less-than-2-year and non-degree-granting institutions would be outdated very quickly -- that is, it would no longer represent an up-to-date universe of these schools. Further, NCES does not anticipate that there will be many survey requests that include these institutions. Thus, NCES decided that when a survey was requested through PEQIS that included less-than-2-year or non-degree-granting institutions, the most recent IPEDS Institutional Characteristics file would be used to draw an up-to-date supplementary sample of these institutions to be used for that survey. This approach means that the basic PEQIS panel will remain up-to-date (i.e., will accurately reflect the current universe of sampled institutions) for a longer period of time, and the supplementary samples of less-than-2-year and non-degree-granting institutions will also be up-to-date for the specific surveys for which these supplementary samples are drawn. NCES believes that this approach is the best compromise between the efficiencies of a standing panel of postsecondary institutions and the need for any such panel to reflect the current universe of institutions.

Sampling weights reflecting the overall probabilities of selection under the design employed for a FRSS or PEQIS study will be computed and assigned to each sampled unit. Nonresponse weight adjustments will be used to correct for unit (whole questionnaire) nonresponse in the survey. Item imputation techniques such as “hot deck” or regression methods will be used to estimate missing values of individual survey items as required. To estimate levels of precision, sampling variances will be computed using the jackknife or other appropriate replication method. Estimates of population characteristics produced from FRSS and PEQIS surveys have generally achieved coefficients of variation (CVs) in the range of 2 to 4 percent for most national estimates, with somewhat higher CVs for subgroup estimates.

B.2. Statistical Methodology

The statistical methodology is described in detail in Section B.1.

B.3. Methods for Maximizing the Response Rate

Telephone followup for nonresponse will begin about 3 weeks after questionnaires have been mailed to the institutions. Experienced telephone interviewers will be trained in administering the questionnaire and will be monitored by Westat supervisory personnel during all interviewing hours. The response rate for most of the quick response surveys with single-stage samples completed to date through FRSS ranged from 85 to 99 percent, with most surveys around 90 percent, and on PEQIS ranged from 90 to 96 percent. Similar response rates are anticipated for future FRSS and PEQIS surveys. Ratio-weighting within adjustment cells will be used to partially compensate for the expected approximately 10 percent nonresponse to each survey.

B.4. Tests of Procedures and Methods

Following the procedures for NCES quick-response surveys (PEQIS and FRSS) in effect during the current QRIS generic clearance (1850-0733), feasibility calls and a pretest (for which OMB clearance is requested under the NCES system clearance for Cognitive, Pilot, and Field Test Studies, OMB #1850-0803) are conducted for each survey prior to OMB review under this QRIS generic clearance (1850-0733). The purpose of the feasibility calls and pretest is to determine what problems respondents might have in providing the requested information and to make appropriate changes to the questionnaire, if necessary. Responses and comments on the questionnaire are collected by fax and telephone during the pretest, and the results are summarized as part of the documentation for the survey.

B.5. Reviewing Statisticians

QRIS surveys are conducted by NCES. NCES contracted Westat to carry out developmental studies; mail the questionnaires; collect data by Web, mail, and telephone; edit, code, key, and verify the data; and produce tabulations and survey reports. Adam Chu, Senior Statistician at Westat, was consulted about the statistical aspects of the PEQIS panel design and the FRSS samples. Chris Chapman, the Associate Commissioner of the Sample Surveys Division within NCES, is the Contracting Officer’s Representative for QRIS, and supervises staff working on QRIS surveys.