APPENDIX Q.3.1 FNS RESPONSES TO NASS COMMENTS Assessment of Barriers to SNAP Benefits

APPENDIX Q.3.1 FNS RESPONSES TO NASS’ COMMENTS

FNS Assessment of the Barriers that Constrain the Adequacy of Supplemental Nutrition Assistance Program (SNAP) Allotments.

OMB control number 0584-NEW

Reviewer: Doug Kilburg, NASS Mathematical Statistician

Supporting Statement Part A

Nothing found to Note a Comment.

Supporting Statement Part B

For part B.1.1. Defining the universe as people on your sampling frame is problematic. List the characteristics of people you want to interview.

The universe is defined as individuals and households participating in the SNAP program (that is, receiving SNAP benefits), who have been in the program for at least the previous six months (at time of sampling). We will select a sampling frame from this universe which stratifies participants according to State of participation and household size.

In part B.1.2., your statements about not needing to build a sampling frame contradicts your definition of your universe in part B.1.1.

B.1.1. has been changed and no longer contradicts B.1.2.

For figure 1, is the number of States per FNS region proportionate to the number of SNAP participants per state by coincidence, or do you mean the number of selected States per FNS region will be proportionate to the SNAP participants per state?

States are not chosen to be proportionate to the number of States within an FNS region. Rather, States are selected, using probability sampling, based on the number of SNAP participants per state. States with more SNAP participants are more likely to be selected for the sampling frame. Because of this, a region with more SNAP participants is also likely to have more States selected, because we are selecting based on SNAP participation. But the sampling is not designed to select States exactly proportionally to FNS regions. We have rewritten the second box in Figure 1 to indicate that 30 States from a stratified State list with participants per State (optional: using number of SNAP participant households as measure of size (MOS)”

On page 4, how will Westat identify the sample of 8 to 12 locations to recruit participants?

The locations will be selected as a purposive sample. Locations with clusters of survey respondents will be prioritized. Two locations with a large base of Spanish-speaking respondents will also be chosen. A balance of rural and urban locations will also be sought, as well as geographic diversity (locations from multiple geographic regions will be selected). Finally, locations with varying levels of household food security will be selected.

I didn’t see the comments from the federal registrar in Appendix Q1.

The comments are in Appendix R1.

On page 4, it is specified that Westat will work with the Agency. Is this the FNS or USDA?

Westat will work with FNS. This has been clarified in the Statement.

In B.2.2 Estimation Procedures, CHAID is discussed that it will be used to determine non-response. This sounds like item non-response. How will survey non-response be handled?

CHAID will be used in the weighting process for handling unit (survey) nonresponse, not for item nonresponse. The purpose of CHAID analysis is to identify variables that better predict unit response propensity that can be used to form the nonresponse adjustment classes (cells). Item nonresponse will be handled by imputation.

With all the different samples described, how will overall weighting be handled? Particularly, how will respondents via craigslist be weighted into the regular survey results with no sampling weight?

Weights will be created for the cases from the probability sample, based on the probability of inclusion. The main sample will not be pooled with the in-depth interview sample — these will be analyzed separately.