Revised Part B WIC PeerCounseling 6 26 08

Part B
Statistical Methods
Part B of the justification for this information collection activity, the WIC Breastfeeding Peer
Counseling Study, addresses the five points outlined in Part B of the OMB guidelines.

B.1

Respondent Universe and Sampling Methods

Phase I will include all 86 States and Indian Tribal Organizations (ITOs) that receive WIC
breastfeeding peer counseling funding, 40 local WIC agencies, 20 local WIC clinics, and 40 agencies
or organizations that collaborate on Loving Support peer counseling with the local WIC agency. A
three-stage cluster sample will be selected, summarized in Exhibit B.1.
Stage 1. The first stage will be a web-based survey of the universe of all 86 States and ITOs that
receive FNS peer counseling grant funding. (Of 90 State agencies, all currently receive peer
counseling grant funds with the exceptions of the Commonwealth of the Northern Mariana Islands
(which entered the WIC program in 2006), Pleasant Point, Maine; Indian Township, Maine; and
Seneca Nation, New York.)
Stage 2. In this stage, we will select 40 local agencies implementing Loving Support peer
counseling. A review of State plans conducted as part of this study found that peer counseling in
many States is not offered in all local agencies. The State plans do not, however, always name the
agencies implementing Loving Support peer counseling. In order to develop the sample frame for
Stage 2, the web-based survey of State agencies will ask for names of all implementing local
agencies. We then will merge this information with local office caseloads from the 2006 WIC
Participant and Program Characteristics Study (PC 2006) files, a biennial report on WIC based on
State management information systems.
Our next step will be to stratify the local WIC agencies implementing Loving Support peer counseling
by combining the seven FNS regions into four “superregions”: East (FNS’ Northeast and MidAtlantic regions), South (FNS’ Southeast and Southwest regions, Midwest (FNS’ Midwest and
Mountain Plains regions), and West (left as is).1 The sample of 40 agencies will be allocated to the
four strata proportional to the number of pregnant WIC participants in each stratum. We will select
local agencies within each stratum with probability proportional to size, where the measure of size is
number of pregnant WIC participants. (Using the number of breastfeeding WIC participants in the
measure of size may give undue weight to agencies that do more effective breastfeeding promotion,
which is one of the outcomes we are measuring.) The number of pregnant participants appears to be a
good index or proxy for the population that is eligible for WIC peer counseling and other
breastfeeding support activities.
The sample for Stage 2 will thus be nationally representative of local WIC agencies with Loving
Support peer counseling programs. We will conduct site visits to these 40 local agencies and
interview staff involved in breastfeeding peer counseling services.
1

WIC caseload by “super region” was calculated using PC 2004. East represents 20 percent of WIC
caseload; South, 34 percent; Midwest, 21 percent; and West, 24 percent.

Abt Associates Inc.

Part B: Collection of Information Employing Statistical Methods

B-1

As part of Stage 2, we will also visit local WIC clinics associated with the local WIC agencies. We
estimate that there will be 20 such clinics that we will visit. While many local WIC agencies have
multiple WIC service-delivery sites or clinics, often the same peer counselors and WIC staff (e.g., the
breastfeeding coordinator) work across most or all of the service delivery sites (e.g., they spend one
or two days at one clinic, and the rest of their time at the other). However, our sample of local WIC
agencies also may include some agencies, such as those that operate in some larger cities, with clinics
that employ Loving Support peer counseling staff that work exclusively at one particular clinic and
may implement the Loving Support peer counseling program differently than the other sites within the
local WIC agency. In order to understand the complete picture of how Loving Support peer
counseling is implemented across the country, it will be important to gather information from these
clinics if they are part of the selected local WIC agency to understand the variation in how services
are delivered locally. When obtaining agreement for local agency participation for Stage 2, we will
ask local WIC directors if there are clinics or separate service delivery sites that implement Loving
Support peer counseling in a way that is different or independent from the way it is implemented
within the local WIC agency overall. At the time we schedule site visits to local agencies, we will
also work with the agencies to schedule visits to the 20 local clinics.
In addition, we will identify 40 community organizations collaborating with the Local WIC
agencies on breastfeeding promotion, and interview staff of one such organization per local agency.
We expect virtually all (more than 95 percent) States and ITOs to complete the web-based survey,
based on responses to previous survey requests at the State level. Based on our experience with
similar studies, we anticipate a 90 percent response rate from local WIC agencies and their
collaborating organizations. Local agencies receiving Loving Support peer counseling are likely to be
enthusiastic about the program and want to share their experience in an FNS-supported study. We
anticipate that virtually all WIC clinics asked to participate will respond positively, as they will be
under the authority of the local WIC agency recommending their participation.

B-2

Part B: Collection of Information Employing Statistical Methods

Abt Associates Inc.

Exhibit B.1
Overview of the Phase I Sampling and Data Collection Plan
86 States & ITOs

Web-based Survey

Stratify 86 States & ITOs
Into 4 Strata Based on
Geographic Regions

PPS Sample within Each of
the 4 Strata
40 Local WIC
Agencies

In-Person Interviews

Purposive Sample

In-Person
Interviews

a

20 Local WIC
Clinics

40 Local Collaborating
Organizations

In-Person or
Telephone
Interviews

“PPS” refers to probability proportional to size

B.2

Procedures for the Collection of Information

Procedures for the collection of information addressed below include:
•
•
•
•
•

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

As described in Section B.1, the sample design for Phase I of the WIC Breastfeeding and Peer
Counseling Study comprises two stages.
1. The first stage will be a web-based survey of the universe of all 86 States and ITOs that
receive WIC breastfeeding peer counseling funding.

Abt Associates Inc.

Part B: Collection of Information Employing Statistical Methods

B-3

2. The data collected in the first-stage survey will include a list of local agencies
implementing Loving Support peer counseling in each State and ITO. This list will
comprise the sample frame for the second stage. The agencies will be stratified by
“superregion,” within which 40 local agencies will be selected with probability
proportional to size (measured as number of breastfeeding WIC participants). The four
superregions are based on the seven FNS regions: East (FNS’ Northeast and MidAtlantic regions), South (FNS’ Southeast and Southwest regions, Midwest (FNS’
Midwest and Mountain Plains regions), and West (left as is). The sample of 40 agencies
will be allocated to the four strata proportional to the number of pregnant WIC
participants in each stratum. The sample for Stage 2 will thus be nationally representative
of local WIC agencies with Loving Support peer counseling programs.
The sample of 40 local agencies in the will each be asked whether they have any local
clinics that implement Loving Support peer counseling programs independent from their
local WIC agency. We plan to visit 20 of the clinics that are so identified.
Estimation Procedures

Abt Associates Inc. will be responsible for assigning sampling weights to the local WIC agencies and
clinics in the study sample. Two sets of weights will be developed for the 40 local agencies in the
Stage 2 sample. The first set will sum to the total number of local WIC agencies (LWAs)
implementing Loving Support peer counseling nationwide, while the second set will sum to the total
number of pregnant WIC participants served by these LWAs. Because the LWAs will be selected
PPS, the weights for local agencies that incorporate participant counts will be nearly uniform. That
is, every pregnant WIC participant will have about an equal probability of her LWA being selected.
In contrast, the weights that sum up to the total number of LWAs will vary across the 40 sampled
agencies, with smaller agencies receiving larger weights. The agency-level weights for the 40 local
agencies in the Stage 2 sample will be equal to the reciprocal of the selection probabilities, and will
sum to the total number of agencies receiving peer counseling fund grants.
Note that within the catchment areas of some of the 40 local WIC agencies, there may be variation in
how Loving Support peer counseling is implemented, with some clinics operating the program
differently than the sponsoring local WIC agency. In these cases, we will also have to weight
information collected from the local WIC agencies and from clinics in order to characterize the
implementation of Loving Support at that individual local WIC site, prior to then weighting the
information to construct the national estimate. In such cases, the local WIC agency’s and the clinic’s
operations each will be weighted proportional to the relative size of the amount of FNS peer
counseling fund grants used by each entity. For instance, if 25 percent of the FNS peer counseling
grant is used by a local clinic and 75 percent of grant is used by the local WIC agency, then
operations would be weighted so that the local WIC implementation is counted for 75 percent to
describe the local implementation and the clinic will be weighted to describe 25 percent.
Characteristics of States and ITOs, and local WIC agencies, will be presented as tabulations and
cross-tabulations. The agency-level analysis will be weighted to incorporate clinic-level information
as just described. Standard errors will be calculated using a statistical software package that takes
account of the complex sampling design, such as SUDAAN or SAS SURVEYMEANS.

B-4

Part B: Collection of Information Employing Statistical Methods

Abt Associates Inc.

Degree of Accuracy

The sample of 40 local WIC agencies will be used to generate national means and proportions related
to characteristics of peer counseling programs—for example, the proportion of pregnant women
served by local agencies in which peer counselors meet with WIC mothers while they are pregnant.
In other words, for observational units (agencies) of varying sizes there is more interest in the element
mean than the group mean (the proportion of agencies in which peer counselors meet with WIC
mothers while they are pregnant). When the observational units are selected with probability
proportional to size (PPS) sampling, an unbiased estimate of the element mean is the simple mean:

y=

1
a

a

∑y

a

,

a

where ya equals 1 if the αth sample agency has the characteristic of interest and 0 otherwise.
Hence for a characteristic with a mean of 0.50, the 95 percent confidence interval will be ±0.15.
Unusual Problems Requiring Specialized Sampling Procedures

No specialized sampling procedures are involved.
Use of Periodic Data Collection Cycles to Reduce Burden

This is a one-time survey data collection effort.

B.3

Methods to Maximize Response Rates and Deal with NonResponse

The procedures to be used to ensure a high rate of response for the study are largely non-statistical in
nature and focus on methods to ensure the cooperation of WIC staff at the State and local levels. Our
expectation, based on similar studies, is that we will achieve a 90 percent or better response rate using
these methods.
In eliciting cooperation from WIC staff, we have found that the following guidelines prove
successful:
•

use senior-level staff for recruitment and refusal conversion;

•

provide sufficient information about the study purposes, objectives, and methodology so
that potential participants have an informed basis for their decision;

•

provide a realistic appraisal of what contributions in time, information, space, and human
resources the participants will be expected to invest in the study effort and a statement of
anticipated benefits to them;

•

demonstrate knowledge and understanding of WIC and breastfeeding promotion
activities and a sensitivity to the problems facing WIC staff in trying to complete their
day-to-day activities; and

Abt Associates Inc.

Part B: Collection of Information Employing Statistical Methods

B-5

•

obtain the endorsement and support of State agencies and professional associations for
the objectives of the study.

In addition, use of highly trained data collectors helps minimize item non-response. These steps have
proven to yield an honest, collaborative relationship between the research team and participants in the
study.
For the State Agency Survey, the contractor’s survey monitoring team will each be assigned a portion
of respondents that they will be responsible for following up with as needed during the web-survey
data collection period. Our plan for following up with State agencies if they do not complete the
survey includes e-mail and phone reminders, and an offer to do the survey by phone. We will also ask
the FNS project officer for help if the State agencies continue to be unresponsive. It has been our
experience that reminding key informants of the importance of the data collection effort and offering
them different options to complete the survey will induce most State-level respondents to participate.
For the on-site data collection, our team of six experienced research staff will begin making contact
with sampled local WIC agencies during the recruitment phase. The assigned data collector will
follow up with each of the local agencies to clarify any questions about participating in the study as
part of the recruitment process. This will help to establish a working relationship with agency staff. In
cases where directors of sampled WIC agencies are reluctant or refuse to participate, we will work
with our FNS project officer in determining if the State WIC Director should be contacted to
encourage the sampled agency to participate. We will replace local WIC agencies that refuse to
participate with other agencies included in the sample as back-ups.

B.4

Tests of Procedures or Methods to Be Undertaken

Web-Based Survey

Two rounds of instrument testing are planned for the web-based State Survey. First, a paper version
of the survey was tested with staff from three WIC State agencies to make sure the questions result in
interpretable and intended responses.
Forms were e-mailed to respondents who were able to print them out and record responses. They
were also provided with a form where they could record the length of time spent on each survey
section. After providing the respondents with adequate time to complete the pilot survey, Abt staff
followed up with each WIC State agency staff participating in the pilot test to discuss their experience
filling out the survey and any comments on the draft instrument. The instruments were revised based
on feedback obtained from State staff, including clarification of instructions, changing some of the
questions, and modifications to question wording.
The second round of testing will occur once the on-line version of the State Survey is developed,
which will occur while the paper instruments are under OMB review. The purpose of the second
round of testing is to ensure that the presentation of the survey questions is clear on screen and the
data entry and editing procedures are user-friendly and minimize error. The second pilot testing of
the instrument will also be conducted by a group of nine or fewer State WIC staff. This test will be
followed by a similar telephone interview to get feedback about using the on-line instrument (ease of
navigation through the instrument, clarity of on-line instructions, layout and structure of the screens,
etc.).
B-6

Part B: Collection of Information Employing Statistical Methods

Abt Associates Inc.

Interview Guides

Abt study staff piloted test each of the interview guides with staff (e.g., WIC agency staff, WIC clinic
staff, and collaborating organization) from two local WIC agencies in Massachusetts and one in
Connecticut. Abt study staff worked with the FNS project officer to contact the Northeast Region and
the Massachusetts and Connecticut State Directors to identify WIC agencies that were willing to
volunteer for the pilot test. The local staff were contacted and site visits were arranged. Interviewers
made note of questions that needed clarification, questions that required adjustments, and those that
needed to be reworded, paying particular attention to phrasing that is common to the WIC community
that should be included to make the questions more meaningful. The time required for each interview
was also recorded. The interview wrapped up with a short session asking the respondents their
opinions of the interview, including what should be changed and what would make it better. The
results of the interview guide pilot were used to revise the guides and the data collection procedures
(e.g., what to include in phone calls setting up the site visits).

B.5

Individuals Consulted on Statistical Aspects and Individuals
Collecting and/or Analyzing Data

Individuals responsible for the data collection, analysis, and all statistical aspects of the study include
the Project Director, Joan McLaughlin (301-634-1833 or [email protected]); the
Task Leader for the Phase I implementation component of the study, Ann Collins (617-349-2664 or
[email protected]); and the Director of Analysis, Nancy Burstein (617-349-2796 or
[email protected]).
In addition, staff from FNS’ Office of Research, Nutrition and Analysis have reviewed the study
design and instruments. Dr. John Endahl is the FNS contact for this effort. He can be reached at 703305-2122 or [email protected]. Abt Associates is responsible for all data collection and
analysis for this study.

Abt Associates Inc.

Part B: Collection of Information Employing Statistical Methods

B-7

Addendum to the OMB Supporting Statement
WIC Breastfeeding Peer Counseling Study
Food and Nutrition Service, USDA
Responses to Questions from OMB
June 11, 2008
1. Please provide more detailed information regarding the Phase II study, especially the key
research questions and how they relate to the data being collected in the Phase I survey
sample.
Response: Phase II is planned as an impact study. The key question for this part of the study is

whether enhancing the peer counseling offered to WIC mothers in local WIC agencies with Loving
Support peer counseling programs results in longer durations of breastfeeding.
The Loving Support peer counseling program provides a lot of local flexibility on key aspects of the
program model. There is currently not enough known about how States and local WIC agencies are
implementing Loving Support to know what are “standard” and common practices and what are
potential enhancements that could be integrated to increase breastfeeding duration rates. Once we
have the results of the Phase I implementation study, FNS will be in a better position to proceed to
Phase II. A separate OMB package will be prepared for Phase II.
2. Please clarify why the state portion of this “study” is preferred over adding detail to the
current state reporting requirements. It appears as though some portion of Phase I is
designed to compensate for inadequate detail collected in mandatory program reporting
(so the proposed study may not actually be a “study”). Three factors contribute to this
impression. First, the proposal itself states, “The depth and quality of the information
obtained from State agencies varied and, for the most part, provided little detail about the
specific ways in which Loving Support is being implemented . . .” (pp. A2 – A3). Second,
Phase I would begin with a “national census of the 86 State WIC agencies receiving Loving
Support peer counseling funds.” Third, the content of the 18 “research questions” and
much of the proposed survey questions seem to be operationally oriented. Thus, the very
broadest question arises of whether some of the information sought by could be obtained
through refined administrative reporting rather than a study format.
Response: Congress specifically appropriated the funds for the Loving Support peer counseling

program and designated that these funds not be counted against the one-sixth of NSA funds that are
targeted to breastfeeding promotion. As such, FNS is working to obtain a level of detail on the
spending of these peer counseling grant funds that goes beyond what could be expected of States to
report on each year, and can be separated from the reporting on the breastfeeding promotion
expenditures typically documented in the State plans. In addition, FNS needs information that
consistently reports on the more detailed aspects of the Loving Support peer counseling on a one-time
basis for this study, so it would not be appropriate to ask states to report on this information annually.
This type of research is not uncommon for FNS. FNS conducts studies of program operations on a
fairly regular basis, as staff need systematic and consistent information that cannot easily and
efficiently be obtained through mandated reporting in State Plans.

Responses to OMB Questions

1

3. Please clarify the primary research goals and questions overriding the list of 18 largely
operational questions characterized as research questions.
Response: The primary research objective for the Phase I study is to obtain a comprehensive and

systematic picture of the implementation of the Loving Support peer counseling program.
4. Please provide a narrative that identifies and explains the components of the Loving
Support model, considers which components are mandatory vs. optional, which are
subject to interpretation and whether that interpretation is at the State or clinic level, etc.
Such a narrative could help in the assessment of how well the proposed survey is tied
with study objectives.
Response: The FNS guidance for the ten components of the Loving Support model is presented in

the table below, which we think more easily captures the information OMB is asking for in this
question. The left column identifies the mandatory components of the model, while the right column
describes the practices that are not mandatory but are recommended by research. Note that even
within the mandatory components outlined in the left column, much is left to the discretion of the
State and/or local agency, based on its WIC population, experience with peer counseling in the past,
local contexts, etc.
Ten Components of Loving Support:
Required Components of a Peer Counseling Program
I.
Appropriate Definition of Peer Counselor.
• Paraprofessional
• Recruited and hired from target population
• Available to WIC clients outside usual clinic hours and
outside the WIC clinic environment.

II.
III.

Designated breastfeeding peer counseling program
managers/coordinators at State and/or local level.
Defined job parameters and job descriptions for peer
counselors.

IV. Adequate compensation and reimbursement of peer
counselors.

Recommendations from the Research
Ideal Peer Counselor:
• Enthusiasm for breastfeeding
• Basic communication skills
• Previous breastfeeding experience (6 months)
• Similarities with WIC participants served
• Current or previous WIC participant
• Similar ethnic background
• Similar age
• Same language spoken.
Use sample job description provided in the Using Loving Support to
Manage Peer Counseling Programs Training Manual.
Job parameters:
• Settings for peer counselors should include telephone contacts
from home; and clinic, home and hospital visits
• Frequency of contacts with pregnant. and breastfeeding women
is important. Refer to Section 6 (Scope of Practice for Peer
Counselors) of “Using Loving Support to Manage Peer
Counseling Programs” Training Manual for recommended
contact frequency during the prenatal and postpartum period.
Job descriptions:
Use sample job descriptions provided in the Using Loving Support
to Manage Peer Counseling Programs Training Manual.
Many current peer counseling programs report paying peer
counselors the same general hourly part-time rate typical of other
entry level positions such as WIC clerical position. This is often
around $5.50 to $7 per hour. However, nearly all programs share
the belief that peer counselors should be paid more.
Other recommendations:
• Provide travel allowance for home/hospital visits/meetings
• Cover training expenses
• Provide benefits if possible
• Reimburse for telephone and other expenses.

Responses to OMB Questions

2

Required Components of a Peer Counseling Program
V. Training of appropriate WIC State/local peer counseling
management and clinic staff.

Recommendations from the Research
Cross-train so that peer counselors are familiar with WIC services
and WIC staff are trained in breastfeeding support.

• State and local program managers receive training in how to
manage peer counseling programs through “Using Loving
Support to Manage Peer Counseling Programs” training
curriculum.
• WIC clinic staff receive training about the role of the WIC
peer counselor through “Peer Counseling: Making a
Difference for WIC Families,” a PowerPoint presentation
included in the “Using Loving Support to Manage Peer
Counseling Programs” training curriculum.
• State and local staff involved in the training of peer
counselors attend the 2005 Regional train the trainer session
“Loving Support through Peer Counseling.”
• WIC clinic staff are trained in basic breastfeeding support
and receive the training “Loving Support through Peer
Counseling.”
VI. Establishment of standardized breastfeeding peer
counseling program policies and procedures at the State
and local level as part of Agency nutrition education plan.

Involve both State and local stakeholders in developing policies and
procedures for a peer counseling program.

VII. Adequate supervision and monitoring of peer counselors

The mentor/supervisor transition has been reported to be effective
for many successful programs. Refer to Section 8 (Supervision and
Monitoring) of “Using Loving Support to Manage Peer Counseling
Programs” Training Manual.
Conduct weekly phone contacts; regular review of contact logs; and
spot checks.

VIII. Establishment of community partnerships to enhance the
effectiveness of a WIC peer counseling program.

(Use sample Peer Counselor Contact Log and sample Weekly
Activity Report Form provided in the “Using Loving Support to
Manage Peer Counseling Programs” Training Manual).
Potential partnerships to consider:
Breastfeeding coalitions; businesses, community organizations;
cooperative extension program; international board certified
lactation consultants; La Leche League; home visiting programs;
private clinics; hospitals

IX.

Provision of the following to peer counselors:

• Timely access to breastfeeding coordinators and other
lactation experts for assistance with problems outside of
peer counselor scope of practice
• Regular, systematic contact with supervisor
• Participation in clinic staff meetings and breastfeeding inservices as part of the WIC team
• Opportunities to meet regularly with other peer counselors.
X.

Provision of training and continuing education of peer
counselors:

•

Peer counselors receive standardized training using “Loving
Support through Peer Counseling” training curriculum.

•

Peer counselors receive ongoing training at regularly
scheduled meetings.

Responses to OMB Questions

Provide formal instruction in addition to home study.
Provide opportunities to “shadow” or observe other peer counselors
and lactation experts.
Provide career path options (e.g., training/experience to become
senior level peer counselors; training to become IBCLC).

3

5. Please identify breastfeeding promotion activities that would be outside of the Loving
Support model. Would a listing of those activities in the survey be helpful for
respondents?
Response: State Survey question #3 lists breastfeeding promotion activities outside of the Loving

Support model, and provides an “other” option if they are doing something not included in the list:
State question 3. Please indicate the breastfeeding promotion activities that your State agency funds
for WIC participants. (Select all that apply)
‰
‰
‰
‰
‰
‰
‰

Media campaigns
Educational materials
Breastfeeding promotion training (other than Loving Support peer counseling training)
Make lactation consultants available to WIC participants
Sponsor certified lactation counselor training (or similar certification training)
Equipment (e.g., breast pumps)
Peer counseling or other counseling by clinic staff to WIC participants that is different
than Loving Support peer counseling
‰ Warmline or hotline
‰ Classes or support group meetings for WIC participants
‰ Other (Specify:)
The local WIC agency interview guide also contains a listing that can be used to prompt staff during
the interview:
Local Agency Question 5. Does your agency provide other breastfeeding promotion services or
programs (e.g., support groups, classes, educational materials) for WIC participants in addition to
Loving Support peer counseling?
‰ No
‰ Yes
**If yes, Please indicate the breastfeeding promotion activities available to WIC participants
in your local agency. (Prompt for funding amt and source, if known, collaborators, and major
activities.)
‰ Media campaigns and/or posting materials in public places, such as WIC clinics.
‰ Making lactation consultants and other trained specialists available to WIC participants
‰ Support groups or classes for WIC participants
‰ Equipment (e.g., breast pumps)
‰ Peer counseling or other counseling to WIC participants that is different than Loving
Support peer counseling
‰ Warmline or hotline
‰ Other (Describe:)

Responses to OMB Questions

4

6. Unless there is a statutory basis for assuring confidentiality, please change the supporting
statement and the instructions on the questionnaire more along the lines of FNS’s intent,
e.g., FNS does not plan to release identifiable data except as required by law.
Response: Based on our discussion with Stephanie on Tuesday, June 10th, we will add to Section

A10 and in our confidentiality section of the questionnaires that we will protect respondent
confidentiality to the extent allowable by law.
7. What is the meaning of the numbers (e.g., 12, 24, 8, and 4 in the “number of respondents”
column) in each “non-responses” row of the burden table? Also, note that there seems to
be an error in the State WIC agency, Total Annual Responses column of the burden table.
Response: The number in the “number of respondents” column in the burden table (attached

to the end of this document) in each non-response row represents the individuals that may
refuse participation in the survey. The number is calculated based on our estimated response
rates. For example, we estimate that 95 percent of the State agencies will respond. Thus, 4 of
each of the 86 State level respondents listed (State WIC Director, State Breastfeeding or Peer
Counselor Coordinator, or State Budget Officer) are estimated to minimally respond,
spending a burden of about 5 minutes on this study (4 State WIC Directors + 2 State
Breastfeeding or Peer Counselor Coordinator + 4 State Budget Officers = 12 respondents).
The numbers for the Local WIC agencies, clinics, and collaborating organizations are
calculated similarly, with the estimated 10 percent non-response for each entity.
There is an error in the State WIC Agency, Total Annual Responses column that we have corrected in
the table.
8. What are the key estimates that the study will produce? Please explain how the sample
sizes were derived and how they release to precision requirements for key estimates.
Response: The key estimates that the study will produce are proportions of State agencies and local

WIC agencies whose Loving Support Breastfeeding peer counseling programs have particular
characteristics.

An example of State agency level characteristics is the proportion of State agencies that fund
particular breastfeeding promotion activities:
•
•
•
•
•
•
•

Media campaigns
Educational materials
Breastfeeding promotion training (other than Loving Support training)
Making lactation consultants available to WIC participants
Equipment (e.g., breast pumps)
Peer counseling or other counseling to WIC participants that is different than
Loving Support
Warmline or hotline

Responses to OMB Questions

5

An example of local WIC agency characteristics is the proportion of local WIC agencies in
which the service delivery site for WIC peer counseling is:
•
•
•

the local WIC agency
local WIC clinics
other

For characteristics of State agencies FNS desires a full census of agencies receiving Loving
Support funds (n=86). For characteristics of local WIC agencies, FNS seeks an estimate
with a confidence interval of +/- 15 percentage points. For characteristics with a mean
proportion of 50 percent, this can be accomplished with a PPS (probability proportional to
size) sample of 40 local agencies.
9. Please explain the proposed procedure for replacing “local WIC agencies that refuse to
participate” and how this will can be done while preserving the probability nature of the
sampling.
Response: The contractor for this study, Abt Associates, will generate a list of 40 local WIC
agencies through the process described in Part B of the Supporting Statement (section B1, Stage 2).
These are the local WIC agencies that will be recruited first. Abt will also develop a backup list of the
remaining agencies using the same procedure, so that if one of the original 40 local WIC agencies
refuses to participate, then recruiters will go to the first local WIC agency on the backup list, then the
second if another refuses, and so on, thus still retaining the probability nature of the sampling.
10. Why are the collaborative organizations to be purposively sampled? What is the proposed
procedure? How will this purposive sampling limit generalizability of the key estimates?
Response: FNS encourages the establishment of community partnerships, but it leaves it to the
discretion of the States and local WIC agencies to determine which organizations are the most
appropriate for their local peer counseling programs. Once we have sampled the local WIC agencies,
and they have agreed to participate, we will ask them which organizations they collaborate with on
the Loving Support peer counseling program. Our experience in the pilot is that it is typically one
other organization or hospital. If there is more than one, we will ask the local WIC agency to identify
the organization that plays the most significant role in their Loving Support peer counseling program.
Thus, when we report the results from these collaborative organization interviews, they will pertain to
those organizations that have been the most helpful to the local WIC agencies in carrying out the
program. While they will not be representative of all collaborating organizations across the country,
they will be representative of local WIC agencies’ primary collaborating organizations.
Please reconsider which questions could be re-formatted as “yes/no” questions rather
than “check all that apply” questions. While the number of questions would of course be
longer as a result, the cognitive burden of a yes/no format has been found to be lower than
“check all that apply” questions. For example, the “yes/no” format has been found to
produce higher data quality than a “check all that apply format” when the response
options are lengthy. Further, in a yes/no format, one of each pair of mutually exclusive
response categories can be eliminated altogether (e.g., two pairs of responses in Question
7 on the WIC State Survey).

Responses to OMB Questions

6

Response: Following the pilot of the surveys, we asked for general feedback about the survey and

specific questions about the difficulty level of the questions asked. Although you note that a lengthier
survey with yes/no questions actually decreases burden, the respondents from our pilot test indicated
that they perceived the case to be otherwise. They told us that they wanted us to keep the survey as
short as possible, so we think lengthening the survey would be a problem for State staff. Issues about
the length of the survey are compounded by the fact that it is a web-based survey and more web pages
to scroll through could potentially increase the burden for those who have internet connections that
require more time to load each page. In addition, the pilot test respondents did not indicate that any of
the “check all that apply” questions were confusing or difficult. For these reasons, we would prefer
not to have to change these questions at this point, as we believe that we would need to test the
lengthier yes/no questions in a similar fashion. Given the goal of having some results available to
policy makers in the coming months, we recommend proceeding with the survey as currently
proposed.
11. Please avoid tabular formats where possible, shifting to series of person-specific
questions.
Response: Similar to the issues raised in question 10, our pilot testing revealed no issues associated

with the survey tables. Given the current schedule for the study, we would recommend that we
proceed with the survey as piloted.
12. On State Level Staff portion of WIC State Survey (p. 6), please use first name only as
identifier in person-specific questions. This approach helps keep staff information
organized and requires less personal information from the respondent than a full name for
staff. Consider which types of information are necessary for the study’s goals. For
example, “How long has person been in this position” may be secondary information
unnecessary for study.
Response: We will do as you suggest and use first name only as an identifier.

We have had the survey reviewed by FNS evaluation and WIC program staff, CDC staff involved in
breastfeeding promotion, and WIC state and local agency staff to try to keep the survey focused on
issues that are key to understanding Loving Support peer counseling and the context in which it is
implemented. The question on the length of time in the staff position is relevant for two reasons: (1) it
allows the evaluators to assess the persons knowledge of Loving Support peer counseling (staff that
have been in the position for a relatively short time will not know the program like an experienced
hand); and (2) it allows federal WIC program staff to assess the level of experience of staff
implementing this important breastfeeding promotion effort.
13. Is information available that would suggest that the “typical” level of staff for this table or,
alternatively, how is known that the respondent does not need to provide information on
fifty individuals, which would be very burdensome to complete?
Response: The range of staff for this table is about 2 to 4, including the State Director, Breastfeeding

or Peer Counselor Coordinator, and a budget/finance staff person.

Responses to OMB Questions

7

14. What is to be calculated from the staff information? FTE in all breastfeeding promotion?
FTE supported by Loving Support?
Response: FTE supported by Loving Support peer counseling grants.
15. Would it be possible to restructure the State Written Guidance (p. 12). Specifically, the
section could begin “Does State provide guidance in addition to or instead of the
recommendations found in Using Loving Support to Manage . . .” A “no” response would
result in a skip of several questions. If the response is “yes,” the respondent could
proceed to each component (role, etc., qualification, timing, frequency, etc) and be asked a
sequence of yes/no questions and, if “yes,” be provided the follow-up questions.
Response: The question will be re-worded along these lines.
16. Frequency of Contact information. Instead of providing either a table for completion or a
series of questions, could the contractor glean information if a state agency simply sent
the state’s “Frequency of Contacts Guidelines” to contractor? This approach could both
increase accuracy (enhance data quality) and reduce burden on respondent.
Response: The State guidelines are often provided in different formats (including e-mails) over

months and years. We believe that asking States to provide their written guidance would actually
require them to take them more time to gather and send it to us than if we ask these contact questions
via a survey at one time. Given our experience with asking for Loving Support Implementation plans,
State plan sections, and budgets, we anticipate that we would need one or more follow-up calls to
clarify information and to understand missing information.
17. If the goals of Phase 2 related to breastfeeding outcomes and if locally collected data will
be used, wouldn’t the questions about breastfeeding data (e.g., exclusivity and duration)
be better focused on data quality aspects like consistency of collection methods (e.g., self
reported on a form versus orally to a counselor etc) rather than mode of availability (e.g.,
Excel)?
Response: FNS is interested in gathering a broad array of information on breastfeeding data

collected by the states. As part of this effort, FNS has requested that Abt Associates provide a
separate deliverable on State tracking of breastfeeding data following the completion of the State
Survey. This deliverable will include the States focus (e.g., rates, duration, and frequency of
breastfeeding), various measures used by the State, how the States use the data, and other items that
will help FNS determine the status of data collection on this important topic.
18. Does FNS have the authority to mandate that States collect information on “ever
breastfed” and “duration” that is based on an FNS-provided definition that is common
across States?
Response: FNS does have this authority and has done so. However, this is a fairly new mandate and

FNS will be able to get a sense of how State agencies are doing through the results of this survey.

Responses to OMB Questions

8

WIC Breastfeeding Peer Counseling Study
Food and Nutrition Service, USDA
Responses to Second Set of Questions from OMB
June 17, 2008
1. Please explain why a complete census of the grantees is needed. Wouldn’t a sample of
grantees do in an effort to reduce burden?

The census of grantees serves three purposes. First, it provides us with a complete picture of the
ways in which Loving Support peer counseling funds are implemented. It is clear that State WIC
agencies are using the Loving Support peer counseling funds in a wide variety of ways, and
differently from one another; a sample of States is likely to produce an incomplete picture of Loving
Support peer counseling implementation. Our need to do a census, instead of take a sample, was
confirmed by review of State agencies’ Implementation and State Plans, discussions with individuals
involved in Loving Support peer counseling implementation, and our pilot test in four States.
Secondly, the census will provide FNS with a complete picture of funding for Loving Support peer
counseling; how funds are used varies, sometimes being matched by other State or grant funds;
sometimes distributed across all local WIC agencies and sometimes just to one. FNS needs to report
to Congress on how the funds appropriated for peer counseling have been used, and thus it needs to
make sure all State agencies are included in a census survey. Finally, it is necessary to do a census of
the states in order to select a representative sample of local WIC agencies receiving Loving Support
peer counseling funds. There is no pre-existing list available of the local sites receiving this funding.

2. Please provide a broad overview of how everything fits together. Explain how the
information in Phase I and Phase II will ultimately help FNS reach its end goal.

In Phase I, FNS is interested in learning how Loving Support peer counseling is being implemented
across the country, including the following: fund distribution and use, number and nature of staff on
the State and local level; written guidance provided to local WIC agencies; responsibilities, training,
and supervision of peer counselors, numbers of WIC participants served by peer counselors; types of
data collected by State and local WIC agencies; and whether other community organizations are
involved. Phase I will provide FNS with information on the current status of the program and how it
has evolved and which, if any, aspects of the Loving Support peer counseling model are “standard”
and where there may be important variations. Based on information from Phase I, FNS will identify
key enhancements of the Loving Support peer counseling model. FNS is then interested in knowing
from Phase II whether breastfeeding duration, a key focus of WIC breastfeeding efforts and one that
research suggests is improved through peer counseling, can be improved by enhancing the Loving
Support peer counseling program. Strategies suggested by research and that are not being done
universally will be targeted for the enhancements.
As you can see, before FNS can initiate Phase II, a better understanding of how State and local WIC
agencies are implementing Loving Support peer counseling is necessary. It is only then that the
strategies for enhancing the program can be identified, and the current level of implementation
identified against which the enhanced programs will be compared.

Responses to OMB Questions

1

3. How will Abt collect the data for the “separate deliverable on State tracking of breastfeeding
data?” Is FNS submitting an OMB clearance for this collection?

The “separate deliverable” referred to above is a memo that will be submitted to FNS on the results of
the questions asked in the State Survey on data collection (found in the last main section of the State
Survey, “State Data Collection about Breastfeeding and the Loving Support Peer Counseling
Program”). These questions are part of this request for OMB approval. FNS did not want to wait
until the final report was submitted to learn about the State data on breastfeeding, so it requested a
separate memo shortly after the results of the State Survey were obtained.

Responses to OMB Questions

2