45 CFR Appendix C to Part 1356 - Appendix C to Part 1356—Calculating Sample Size for NYTD Follow-Up Populations

prev | next
Appendix C to Part 1356—Calculating Sample Size for NYTD Follow-Up Populations
1. Using Finite Population Correction

The Finite Population Correction (FPC) is applied when the sample is drawn from a population of one to 5,000 youth, because the sample is more than five percent of the population.

Sample size with FPC = ( Py ) ( Pn ) + Std. error 2 Std . error 2 + ( Py ) ( Pn ) N

• (Py)(Pn), an estimate of the percent of responses to a dichotomous variable, is (.50)(.50) for the most conservative estimate.

· Standard error = Acceptable level of error Z coefficient

• Acceptable level of error = .05 (results are plus or minus five percentage points from the actual score)

• Z = 1.645 (90 percent confidence interval)

Standard error, 90 percent confidence interval = . 05 1.645 = . 0303951

• N = number of youth from whom the sample is being drawn

2. Not Using Finite Population Correction

The FPC is not applied when the sample is drawn from a population of over 5,000 youth.

Sample size without FPC, 90 percent confidence interval = ( Py ) ( Pn ) Std. Error 2 = ( . 50 ) ( . 50 ) ( . 0303951 ) 2 = 271

[73 FR 10372, Feb. 26, 2008]