SFSP AG_3198_C_13_0016 OMB Supporting Statement for Clearance to Conduct Surveys and Interviews Part B

Supporting Statement for Evaluation of the Summer Food Service Program (SFSP) Participant Characteristics

(OMB Control Number: 0584-NEW)

Supporting Statement for

Evaluation of the Summer Food Service Program (SFSP) Participant Characteristics

Part B

(OMB Control Number: 0584-NEW)

April 14, 2014

Prepared for:

USDA, Food and Nutrition Service

Office of Research and Analysis

3101 Park Center Drive, Room 1014

Alexandria, VA 22302

COR: Chanchalat Chanhatasilpa

[email protected]

CS: Ashley Owens

[email protected]

Prepared by:

Optimal Solutions Group, LLC

University of Maryland M Square Research Park

5825 University Research Court, Suite 2800

College Park, MD 20740-9998

Point of Contact:

Mark Turner, Ph.D., Project Director

Email: [email protected]

Phone: 301-918-7301

Table of Contents

PART B. Collections of Information Employing Statistical Method

1. Universe and Sampling Respondent Selection...……………………………………….. 3

2. Describe the procedures for the collection of information 5

3. Describe methods to maximize response rates and to deal with issues of non-response. 6

4. Describe any tests of procedures or methods to be undertaken. 8

5. Provide the name and telephone number of individuals consulted on statistical aspects of the design. 9

PART B. Collections of Information Employing Statistical Methods

1. Universe and Sampling Respondent Selection

Describe (including a numerical estimate) the potential respondent universe and any sampling or other respondent selection method to be used. Data on the number of entities (e.g., establishments, State and local government units, households, or persons) in the universe covered by the collection and in the corresponding sample are to be provided in tabular form for the universe as a whole and for each of the strata in the proposed sample. Indicate expected response rates for the collection as a whole. If the collection had been conducted previously, include the actual response rate achieved during the last collection.

Developing the sample frame information for the study will require data collection at several stages. State agencies will be contacted to provide the lists of sponsors and sites to develop the sample frame for the selection of sponsors at the end of the spring of 2015. The sponsors selected for the sample will be contacted to provide their respective site lists in early summer to develop the sample frame for the selection of sites. Two school districts in SFSP eligible areas will be contacted to provide the lists of eligible children, (participating in school lunch or breakfast programs) to select respondents for the semi-structured interviews with parents or caregivers of SFSP participants and eligible nonparticipants. Twenty-five parents or caregivers of SFSP participating children and 25 parents or caregivers of SFSP-eligible nonparticipating children will be randomly selected from the lists. The parents/caregivers surveys will not be representative and will be a sample of convenience. Due to the low response rates reported by 2003 study (45%), financial incentives will be used to increase response rates.

To develop the sampling frame and to answer some research questions, administrative data will be collected from the states, the sampled sponsors, and two schools or SFA sites. The data will be used to develop the sampling frame of sponsors, sites, and parents/caregivers of participants and eligible nonparticipants. Administrative data on sponsor and site characteristics will be used to address research Objectives 1, 2, and 3, as well as to provide comparisons with the ERS/MPR 2003 results (research Objective 6). The administrative data for this study will be collected from three sources:

1. The states’ administrative data will include a request for a complete listing, contact information, and types of current SFSP sponsors and sites; planned dates of meal service; tabulations of SFSP participation for the previous summer (e.g., number of children and number of meals served), meal types, and the number of meals expected to be provided during the summer.

2. The administrative data from the sampled sponsors will include a request for a complete listing, contact information, and types of current SFSP sites; planned dates of meal service; and meal types and the number of meals expected to be provided during the summer.

3. The administrative data from the two schools or SFA sites will include the lists and the contact information for their NSLP and SBP participants and eligible nonparticipants.

Based on the 2003 study the expected universe and response rates for this request are presented in Exhibit 5.

Exhibit 5. Sampling Universe, Sample, and Response Rates

Type of sample	Respondent universe	Sample size	Expected response rates	Expected non-participants
State agencies	54	54	100%	0
Sponsor organizations	4,397	300	90%	30
Sites	35,530	350	90%	5
Parents/caregivers of participants	n/a list of 200	25	80%	5
Parents/caregivers of eligible nonparticipants	n/a list of 200	25	80%	5

Note: the sample of parents/caregivers will be a sample of convenience without an expectation of representativeness.

2. Describe the procedures for the collection of information

Statistical methodology for stratification and sample selection,estimation procedure, degree of accuracy needed for the purpose described in the justification, unusual problems requiring specialized sampling procedures, and any use of periodic (less frequent than annual) data collection cycles to reduce burden.

Statistical methodology for stratification and sample selection

Based on the obtained sponsors and site lists, a stratified sampling approach will be used to select nationally representative samples of sponsors and sites by USDA region, location (urban/rural), and with probability proportional to size, where the measure of size will be average daily attendance at SFSP sites. The overall design for the study involves two-stage stratified sampling which utilizes the USDA regions as the primary sampling strata, and selects sponsors and sites at each stage with a probability proportional to size. The sampling frame for the first stage of this effort will be derived from listings of sponsors, and data on program volume obtained in conjunction with the census of state administrators. The second stage (selecting program sites) will be based on information obtained from program sponsors and will target up to two sites per responding sponsor. The first-stage selection of sponsors will involve two phases. In the first phase, a sample of sponsors expected to return from the preceding year will be selected from the previous year’s lists. New sponsors will be selected from a list of new sponsor training attendees in the second phase.

These sampling procedures will target the selection of a total of 300 sponsors and 350 sites. Under standard assumptions, power calculations indicate that the target size for the primary sample of 300 sponsors and 350 sites should be sufficiently large to detect the influence of environmental or policy factors individually using analysis of variance (ANOVA) or within a general linear modeling framework, even when effect sizes are small (0.1-0.2). Calculations also assume a design effect of 1.5 and at least an 80% response rate among sponsors and sites. Non-sampling errors arising from unit and item nonresponses will be dealt with through weighting and imputation where appropriate. The sponsors and sites data will be weighted by calculating the sampling weight as the inverse of the probability of selection. The sampling weight for the sites will depends on the probability of selection of the sponsor at the first stage.

3. Describe methods to maximize response rates and to deal with issues of non-response. The accuracy and reliability of information collected must be shown to be adequate for intended uses. For collections based on sampling, a special justification must be provided for any collection that will not yield "reliable" data that can be generalized to the universe studied.

The study will utilize a mixed-mode data-collection approach to maximize the survey response rates and minimize the respondent burden. The following data-collection techniques will be used:

• Use USDA official introductory letters and e-mails to establish the legitimacy and the importance of the survey and to make respondents aware that this is a USDA initiative.

• Use introductory letters and e-mails from state agencies to make sponsors and sites aware of the importance and legitimacy of the survey. The introduction will describe the study’s purpose and value, the privacy and confidentiality of the survey data, and its use in the analysis.

• Send the color brochure by mail and by e-mail to explain the study’s value; to outline the research objectives, data-collection approach, dates, and the information being collected; and to provide relevant contact information.

• Send up to five e-mail reminders at various times and on various days of the week to increase the odds of the survey invitation’s being noticed and acknowledged.

• Send up to five reminder postcards after the introductory letter to increase the number of contacts with the respondents.

• Provide self-addressed, prepaid envelopes, and closely monitor the status of postal and e-mail survey invitations and survey activity to track respondents’ activity, and send targeted and friendly e-mail reminders to non-respondents.

• Conduct up to five reminder telephone calls with non-respondents. The respondents will have the option of taking the web survey, responding to the telephone survey, rescheduling for a more convenient time, requesting a paper version of the survey, or identifying another suitable point of contact.

4. Describe any tests of procedures or methods to be undertaken.

Testing is encouraged as an effective means of refining collections of information to minimize burden and improve utility. Tests must be approved if they call for answers to identical questions from ten or more respondents. A proposed test or set of tests may be submitted for approval separately or in combination with the main collection of information.

The five types of survey instruments (state agencies, sponsors, sites, parents/caregivers of participants, and parents/caregivers of nonparticipants) were pretested with two state administrators, six sponsors, six site supervisors, two parents/caregivers of the participants, and three parents/caregivers of eligible nonparticipants. The state and sponsor respondents were also asked to estimate the burden required to provide the lists and administrative data on the sponsors and the sites that they operate.

The pretesting involved cognitive interviews to estimate respondents’ burden and to identify respondents’ difficulties in interpreting and responding to the questions. This approach utilized probes to determine whether respondents were having difficulty understanding the meaning of the questions or words, problems remembering the information needed to answer questions, issues with unnecessary questions or response options, or trouble with missing questions or response categories. The items that were reported by the respondents or were noticed by the interviewers as being problematic, difficult, or time-consuming to answer, were further discussed to determine the reasons for the difficulties.

Based on the pretesting results, the surveys were revised to minimize the respondents’ burden. The revisions involved rewording some questions to improve comprehension and responding, omitting some items that were confusing or difficult to answer, adding skip question options to avoid asking questions that were not applicable, revising some of the response scales, adding a few specific questions and items, and revising the surveys’ overall layout and format. The final versions of the surveys are attached in Appendix E-1, F-1, G-1, H-1 and I-1.

5. Provide the name and telephone number of individuals consulted on statistical aspects of the design and the name of the agency unit, contractor(s), grantee(s), or other person(s) who will actually collect and/or analyze the information for the agency.

The names and telephone numbers of individuals consulted on statistical and methodological aspects of the design are:

EXPERT, ADVISORY PANEL- SFSP

Expertise	Name	Position	Affiliation	E-mail
Sampling, survey methodology	Laura Stapleton	Associate Professor in Measurement	Statistics and Evaluation in the Department of Human Development and Quantitative Methodology at the University of Maryland	[email protected]
Subject matter, food security programs, program administration	Judi Bartfeld	Director	Research on Poverty (IRP)--USDA's Food Security Research, Innovation, and Development Grants in Economics (RIDGE) program--University of Wisconsin	[email protected]
Subject matter, food security programs, program evaluation, social services	Signe Anderson	Child Nutrition Policy Analyst	Food and Research Action Center (FRAC)	[email protected]
Subject matter, food security programs, program evaluation, social services	Kate Sims	Outreach and Policy Associate	Food and Research Action Center (FRAC)	[email protected]

Statistical Contact

For questions regarding the study or questionnaire design or statistical methodology, contact:

Mark Turner, Ph.D., Project Director

Optimal Solutions Group, LLC

University of Maryland, M Square Research Park

5825 University Research Court, Suite 2800

College Park, MD 20740-9998

Telephone: 301-918-7301; e-mail: [email protected]

7