Selecting Your Sample Size

Using the sample pool, SFAs will calculate the required sample size (number of applications that must be verified) using one of the following methods:

  • Error Prone (Standard) Sampling Method
  • Random Sampling Method (Alternate 1)
  • Focused Sampling Method (Alternate 2)

The Error Prone (Standard) Sampling Method must be used by an SFA unless it qualifies to use one of the alternate sampling methods. The SFAs approved to use an alternate sampling method are listed in the SFAs Required to Conduct Income Verification tab.

When determining the sample size, regardless of the verification method used, SFAs must always round up to the nearest whole number. For example, if your sample size is 3.18 you must verify 4 applications. You may not verify any more or any less than the required sample size allowed.  Please note, at least one application must always be verified.

Error Prone (Standard) Sampling Method

New SFAs and SFAs that had less than 80 percent response rate to verification in the prior school year MUST use the Error Prone (Standard) Sampling Method to conduct verification in the current school year. SFAs coming off Provision 2 or CEP must also use the error prone sampling method.

For Error Prone (Standard) Sampling Method, SFAs must verify the lesser of three percent (3%) or 3,000 of the approved applications on file as of October 1. The applications to be verified must be selected randomly from applications that meet the definition of an Error Prone application.

*Error prone applications are those with a reported income within $24 per week, $100 per month, or $1,200 per year of the applicable Income Eligibility Guidelines.

If the SFA does not have enough error prone applications to produce a three percent (3%) sample, the SFA must use income applications and/or applications that report a SNAP/TANF/FDPIR number selected at random to meet the required sample size under the error prone sampling method.

SNAP/TANF/FDPIR applications will not be selected unless there are too few applications that meet the error prone criteria.

Note: When approving applications, SFAs should flag those that meet the error prone definition (see definition above) in order to facilitate this process.

Alternate Sampling Methods (Random Method & Focused Method)

SFAs that had an 80 percent (80%) or greater response rate in the previous school year may choose to select an alternate option for verification. (Please see the list of schools with 80% or greater response rate based on the previous school year's Income Verification.) These schools may choose to use the Error Prone Sampling Method or may choose one of the following sampling methods:

Random Sampling Method (Alternate 1)

SFAs must verify the lesser of three percent (3%) or 3,000 of the total number of approved applications on file as of October 1. The applications to be verified must be selected randomly.  The random sample should include both income applications and categorically eligible applications for children receiving SNAP/TANF/FDPIR benefits whose names do not appear on the direct certification matching list.

Focused Sampling Method (Alternate 2)

SFAs must verify the lesser of one percent (1%) or 1,000 of the total number of approved applications selected from income applications nearest the income eligibility guidelines (error prone applications); PLUS the lesser of half percent (½ %) or 500 of approved applications which reported a SNAP/TANF/FDPIR case number on an application instead of income information.

Example:

The Focused Sampling Method could be accomplished as follows:

SFA with 900 approved applications which include 600 applications that reported a SNAP/TANF/FDPIR case number

1.  Count all approved applications, including applications reporting a SNAP/TANF/FDPIR case number to determine the number required to fill the one percent of income sample size

.01 (1%) x 900 = 9

2.  Separate applications into two groups:  

  • Income Applications (non-SNAP/TANF/FDPIR)
  • Applications reporting a SNAP/TANF/FDPIR case number

3.  From the income applications (non-SNAP/TANF/FDPIR) group, select the sample of households (9) from error prone applications. If you do not have enough error prone applications, select the remainder at random from the income applications.

4.  From the applications with a SNAP/TANF/FDPIR case number, determine the number required to fill the .5 percent sample size

.005 (.5 percent) x 600 = 3

5.  Choose the applications (3) from those that reported a SNAP/TANF/FDPIR case number.

6.  Perform the confirmation review on all selected applications and continue with all steps of verification.

*Note: 12 applications were verified in this example (9 income and 3 SNAP/TANF/FDPIR)