#0584-0299 SS Part B Statistical Methods Aug 2006

Supporting Statement for the
FNS 380-1 Quality Control Review Schedule
OMB # 0584-0299
B.

Collections of Information Employing Statistical Methods

1.

Describe the potential respondent universe and any sampling or other respondent selection
method to be used.
The QC system and its reporting requirements are applicable to all of the 53 State agencies
administering the Food Stamp Program. Each State agency is required to select a sample of
households from its active case universe, the universe of households certified and issued
benefits. State sample selection choices result in an estimated yearly national sample of 57,146
cases. Historically, response rates in the form of completed case reviews have ranged between 90
and 95 percent.

2.

Describe the procedures for the collection of information.
The active case sample procedure is generally in the form of systematic random sampling.
Stratification is an option available to State agencies subject to FNS approval. State agencies
opting to use stratification divide their sample universe according to household case
characteristics, such as household size, geographic area, or program area (i.e., public assistance
or general assistance), prior to random selection.

3.

Describe the methods to maximize response rates and deal with issues of non-response.
The QC system attempts to control the potential bias of non-response with a completion standard
of 100 percent for both the active and negative case samples. This percentage is the ratio of the
number of completed reviews to the number of cases selected subject to review or to the
minimum number of cases computed from the sample size equations, whichever is larger. A
State agency's entitlement to the increased Federal share of administrative costs depends on the
satisfactory completion of all of the required active and negative case reviews. If a State agency
does not complete the required number of reviews, its error rates are adjusted by assigning two
standard deviations of the estimated error rates added to the regressed error rates to those cases
not completed in order to calculate the State agency's official error rates. Two standard
deviations correspond to approximately a 95 percent confidence interval. The use of standard
deviations of the mean is a common device for measuring the dispersion in a simple distribution
and represents a natural breakpoint.

4.

Describe any tests of procedures or methods to be undertaken.
No testing of information-collection procedures is planned.

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.
Nick Manthos, (703) 305-2474, and Kip Houge, (415) 705-1361, Senior Statisticians, Food
Stamp Program, Program Accountability Division, Quality Control Branch, are responsible for
analyzing the QC data.

2