E-5. Evaluation of the School Meal Data Collection Process - NASS Comments

APPENDIX E5. RESPONSE TO NATIONAL AGRICULTURAL STATISTICS SERVICE (NASS)

APPENDIX E5.
RESPONSE TO NATIONAL AGRICULTURAL STATISTICS SERVICE (NASS)

Response to National Agricultural Statistics Service (NASS) Comments on
Evaluation of the School Meal Data Collection Process

Overview
This appendix provides Westat’s response to the comments received from the National
Agricultural Statistics Service (NASS) on the Revised Draft OMB Package. The response to
the comments is organized as follows: A) response to the general comments B) response to
appendices comment; and C) responses to Part B comments.
Additional edits and suggestions from NASS within Part A and Part B were incorporated (as
tracked changes).

Response to General Comments
NASS General Comments:
This docket is well described and documented. One item that seems to not be discussed much
is what methodology will be used to summarize the data and create the recommended
guidelines and pinpoint errors. The responses to the interview questions are mostly open-ended
ones. Other than compiling the text responses, I would like to see discussion on how these
responses will be quantified and summarized.
Response:
Additional information on the methodology used to summarize the data has been added to
section B.2:
For the analysis, an iterative approach will be used, first analyzing the rich and detailed
information we collect during on-site qualitative interviews. Analysis will include
review of data elements with most frequent errors by each of the three forms and any
variations among school and SFA size. We will also analyze methods of data collection
and aggregation by form as well as school and SFA size in addition to most frequent
method and modes of data transmission for each form. Most frequent sources of 30and 90-day meal count volatility on the FNS-10 form as well as type and frequency of
data quality control measures will be explored.

1

Response to Appendices Comment
NASS comment:
In Part A and Part B Table of Contents, Appendix E is noted with 4 items. These four items
were not included in the Appendices folder that accompanied this package. If items are to be
noted, then I recommend making sure they’re included in the docket package.
Response:
Appendix E contains 5 items, all of which will be included in the docket package.

Response to Part B Comments
NASS comment:
B-1: The respondent universe is listed as the respondents in this study, which is problematic.
The universe would be all schools in the U.S. that meet criteria of having a meal plan program.
The sample for this survey would consist of 4 states and all parties noted in the section.
In constructing the sample, the sampling methodology is discussed generically, providing
more detail behind the selection would be beneficial.
Response:
The respondent universe is listed as all schools in the U.S. participating in the National School
Lunch Program. Detailed information on the selection criteria in the sample selection has been
added to Part A.2, Purpose and Use of the Information.
Nass comment:
B-3: The text states an expected 190/197 ~ 96% response rate. The guideline is referenced that
cooperation is mandatory, but an expectation of 96% is quite high. On top of the high response
rate, nothing is of note on how non-response will be handled. It is stated that 7 respondents
from the SFA Directors will be non-response, how is their non-response going to be adjusted?
Response:
Regarding the high response rate, participation is ultimately voluntary and we expect that
support and agreement to participate at the State Director level will be communicated to the
SFA Directors, which will then in turn communicate to School Food Managers.
Regarding non-response, we will effectively communicate and employ strategies of refusalconversion to those respondents who initially decline participation. If the respondent refuses to
participate, the next respondent fitting similar selection criteria will be invited to participate. A
final count will be kept on the number of non-responders.

2