0584-0547 Advance Modernization
I have reviewed the attached OMB docket from FNS and here are my comments/suggestions.
1. This study is using convenience sampling quite often to determine the groups to be interviewed. In one case, this may be the best method available because there is no list of eligible nonparticipants of SNAP. They are aware of the potential bias caused by selecting people that are using emergency food centers. However, they are using convenience sampling to select which sites they visit to conduct their interviews. They have several criteria that want to use in selecting the sites so I would think that all sites could be stratified based on these criteria and a random selection could be made from each strata. All sites need to have a chance of selection in order to have a representative sample. I would also suggest visiting more than 4 sites per state especially if SNAP is county-administered since the variance of modernization may be greater.
2. The docket only describes a qualitative data analysis based mostly on interviews and focus groups. I would think that some quantitative analysis could be conducted especially since they are already collecting data from the sites. Would it not be beneficial to see if there were any measurable differences before and after modernization such as reduced error rates, faster approval time, and decreased costs? This would show which of the studied states' efforts were the most effective.
Otherwise, it looks like they have planned the study pretty well especially since they contracted Mathematica to conduct all the data collection and analysis. Sorry it took me so long to review it.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Lindsay C. Drunasky
Statistical Methods Branch
USDA-National Agricultural Statistics Service
202-720-8951
| File Type | application/msword |
| Author | hancda |
| Last Modified By | hancda |
| File Modified | 2010-11-10 |
| File Created | 2010-11-10 |