Private School Universe Survey (PSS)
August 2013—July 2016
Supporting Statement
Part B
Request
for OMB Review
OMB# 1850-0641 v.5
Submitted by
National Center for Education Statistics
U.S. Department of Education
January 30, 2013
revised May 2015
SUPPORTING
STATEMENT
TABLE OF CONTENTS
B. COLLECTION OF INFORMATION EMPLOYING STATISTICAL METHODS 2
B3. Methods for Maximizing Response Rates 2
B4. Test of Procedures and Methods 3
B5. Statistical Consultants, Agency Staff, and Contractors 3
Appendix A. State List Request Letter
Appendix B. Association List Request Letter
Appendix C. School Communication Materials
Appendix D. PSS Questionnaire
Appendix E. Examples of Table Shells
The target population for the survey consists of all private schools in the United States that provide instruction to any of grades Kindergarten-12, or comparable ungraded levels (approximately 30,000 schools).
The universe frame to be used by NCES for the 2013-14 PSS will be composed of the 2013-14 PSS list frame and the 2013-14 area-search frame. The list frame is based on the 2011-12 PSS and is updated with current membership lists from private school associations and private school lists from state departments of education and social services (for kindergarten programs)1. The area list is developed after a systematic search of approximately 124 randomly selected primary sampling units (PSUs). The list of schools found in this search is unduplicated with the list frame to yield the area-frame schools. The area frame is used to represent schools missing from the list frame. The 2011-12 area-frame search was conducted in 124 PSUs and identified 354 unduplicated schools, weighting up to represent 1,941 private schools. Similarly, the universe frame to be used by NCES for the 2015-16 PSS will be composed of the 2015-16 PSS list frame and the 2015-16 area-search frame.
A combination of mail/Internet, telephone follow-up, and personal visit follow-up will be used to ensure an anticipated 95 percent response rate for the 2013-14 and 2015-16 PSS2. The first mailout will be followed by a reminder card (1 week later) and by a second mailout to nonrespondents (1 month later). Six weeks after the second mailout, noninterview cases will be turned over to the CATI facility for telephone follow-up and, when necessary, to field staff for personal visit follow-up.
NCES proposes to conduct a split-panel branding experiment in the 2015-16 PSS to test whether NCES or Census branding results in higher response rates. Two experimental panels of 5,000 cases (stratified by Catholic, Other religious and nonsectarian) would be selected for this experiment. The first panel will receive Census Bureau branding on all survey materials (e.g. Census Bureau name, logo, and commissioner’s signature) and would not be offered the Internet option (the internet instrument has the NCES branding). The second panel would be the control group and have the usual NCES branding (e.g. NCES name, logo, and commissioner’s signature) on all survey materials and would also not be offered the Internet option. The panel size of 5,000 will allow for differences of 2.5 percentage points to be detected (alpha 0.05). The balance of the cases would receive materials with the usual NCES branding and be offered the option to complete the survey using the Internet. The materials sent to all cases would state that the Census Bureau is conducting the survey for NCES.
Several procedures will be employed to ensure high response rates. The following elements of the data collection plan will contribute to the overall success of the survey.
1. Visible support from top-level federal education officials.
Top-level education department officials addressed the representatives of a number of private school associations at an NCES-sponsored meeting that took place in November 2012. The officials demonstrated their support for the survey by informing the representatives of the timing and objectives of the survey and by soliciting the representatives’ cooperation to ensure the survey’s success.
2. Interactive field monitoring with extensive follow-up (by telephone and personal visits) of nonrespondents.
The survey will use mixed survey modes—mail/Internet, telephone, and personal interviews, as needed, to achieve high response levels in 2013-14 and 2015-16. Past experience indicates that the majority of the school surveys will be completed by mail/Internet. Follow-up telephone or in-person interviews for the mail nonrespondents will be conducted to achieve an expected response rate of 95 percent. In the 2011-12 administration of PSS, we received 18% of responses by internet and, as been the case for the past two decades, about 58% of responses by a self-administered mode (internet/mail). We expect a similar distribution of responses by mode in the 2013-14 and 2015-16 administrations of PSS.
In addition, the survey content is based upon administrative information that is commonly on file in private schools. The ease of answering the items should facilitate the respondent’s willingness to complete the questionnaire.
The 2013-14 survey questionnaire and procedures will be based on those used in the 2011-12 PSS, with two questions added for the 2015-16 PSS (see amendment request memo). The revised Appendix D provides the 2015-16 PSS questionnaire. Since no new items are to be included in the 2013-14 PSS questionnaire, there is no need to test it. The two items added to the 2015-16 PSS have been used in SASS private school questionnaire for many years, and thus also do not require testing.
In addition to review within NCES, Aref Dajani and Sharon Stern of the U.S. Census Bureau reviewed the PSS sample design and related matters for statistical quality, feasibility, and suitability to the overall objectives of the survey. The following individuals have oversight of the data collection and analysis:
Stephen Broughman
PSS Project Director
ESSSP/ESLSD/NCES
U.S. Department of Education
Washington, DC 20006
(202) 502-7315
Andrew Zukerberg
Branch Chief
ESSSP/SSD/NCES
U.S. Department of Education
Washington, DC 20006
(202) 219-7056
Adam Rettig
PSS Project Director
Demographic Surveys Division
U.S. Census Bureau
Washington, DC 20233
(301) 763-2673
Aref Dajani
Chief, Longitudinal Surveys Branch
Statistical Methods Division
U.S. Census Bureau
Washington, DC 20233
(301) 763-1797
Carolyn Pickering
Chief, Education Surveys Branch
Demographic Surveys Division
U.S. Census Bureau
Washington, DC 20233
(301) 763-3873
The questionnaire is designed to collect information concerning the number of private schools, teachers, students, and high school graduates. In addition, the questionnaire contains several screening items that are necessary to establish that schools are bona fide schools or educational programs offering kindergarten programs and not day care centers, preschools, and the like. See Appendix D for a copy of the 2013-14 PSS questionnaire.
Item 1 seeks information to identify the respondent.
Item 2 lists statements that will be used to screen out schools that are not private schools or do not serve students in grades K-12.
Item 3 gives instructions to the respondent.
Items 4 and 5 obtain information regarding the grades offered, enrollment by grade, and total enrollment. These questions will be used to determine the current grades taught in private schools and will be used to classify sampled schools as elementary, secondary, or combined. The total enrollment (Item 5) and enrollment by grade (Item 4) will be used to estimate the total enrollment in private schools and the number of private school students by grade or grade span. This data will be used in the sample selection design for other school surveys.
Item 6 asks the racial/ethnic distribution of the student body. This is a basic school descriptor and an important policy-relevant measure of minority group composition in the school. This data will be used in the sample selection design for other school surveys.
Item 7 asks if a school is coeducational, all female, or all male and, if coeducational, how many male students attend the school. This is a basic school descriptor. This data will be used in the sample selection design for other school surveys.
Item 8 asks for the number of regular high school graduates during the 2012-13 academic year and the percent that went to 4-year colleges. This is a basic school descriptor and an important policy-relevant measure.
Item 9 asks schools to indicate the existence of a kindergarten program, and for those schools with a kindergarten program, whether it is a full- or half-day program and the number of days per week of the kindergarten program. The characteristics of kindergarten programs in private schools are important policy-relevant measures. This data will be used in the sample selection design for other school surveys.
Item 10 asks schools to indicate their number of full- and part-time teachers. This information will allow estimates of both the total number of private school teachers and the total number of full-time-equivalent teachers in private schools. This data will be used in the sample selection design for other school surveys.
Item 11 seeks information regarding school type. This information will be used as a classification variable in analysis.
Item 12 asks the extent to which private schools are supporting home schooling and whether the school is operated in a private home used as a family residence. This question will be used to determine the prevalence of schools in private residences and will also be used (in conjunction with Items 5 and 10) to screen out families that are homeschooling.
Items 13 and 14 ask schools to indicate whether they have a religious orientation and to indicate to what associations they belong. These data will be used to estimate the number of Catholic, other religious, and nonsectarian schools; and to determine if schools are members of school associations. The data will also be used as a classification variable in analysis and in the sample selection design for other school surveys.
Item 15 asks for the number of days in the school year for students. These data will be used to estimate the average length of the school year for students attending private schools.
Item 16 will obtain information about the length of the school day in private schools. These data will be used to estimate the average length of the school day for students attending private schools.
Item 17 asks schools to indicate the existence of a library/media center. This is a basic school descriptor and will be used in the sample selection design for other school surveys.
Item 18 asks schools if they participate in the National School Lunch Program and how many of their students are approved for the NSLP. This data will be used to measure the incidence of participation in the NSLP and as a classification variable for private schools.
Item 19 asks schools if any students at the school received title 1 services and, if so, how many. The data will be used to measure the incidence of participation in title 1 in private schools.
Item 20 will be used by NCES to study survey response patterns.
Item 21 will be used for updating the school name and address. This data will be used in sample surveys and future PSS collections.
Item 22 will obtain the physical or street address of the school, if different from the mailing address. This information will be used in assigning a geocode to the school location and for unduplication.
1 See appendices A and B for examples of letters used to request private school lists.
2 The response rate for the 2011-12 PSS was 92.1 percent.
| File Type | application/msword |
| File Title | SUPPORTING STATEMENT |
| Author | Steve Broughman |
| Last Modified By | Kashka Kubzdela |
| File Modified | 2015-05-19 |
| File Created | 2015-04-30 |