有限总体校正计算器
在有限总体抽样时应用有限总体校正因子。
Corrected Standard Error
2.238307
Corrected Standard Error vs Uncorrected Standard Error
公式
How to Apply the Finite Population Correction
Formula
FPC = sqrt((N - n) / (N - 1))
Corrected SE = SE * FPC
When sampling without replacement from a finite population, the standard error is smaller than the infinite-population formula suggests. The FPC factor accounts for this. It is close to 1 when n is small relative to N, and approaches 0 as n approaches N (census).
计算示例
Standard error = 2.5, sample of 100 from a population of 500.
- 01Sampling fraction = 100/500 = 0.20 (20% of population)
- 02FPC = sqrt((500-100)/(500-1)) = sqrt(400/499) = sqrt(0.8016) = 0.8954
- 03Corrected SE = 2.5 * 0.8954 = 2.2384
- 04The standard error is reduced by 10.5%
常见问题
When should I apply the finite population correction?
Apply FPC when the sampling fraction (n/N) exceeds 5% (rule of thumb). For n/N < 0.05, FPC is close to 1 and can be ignored. For large populations (millions), the correction is almost never needed.
Why does the correction reduce the standard error?
When you sample a large fraction of the population, there is less uncertainty because you have already observed much of the population. In the extreme (census, n = N), the standard error is zero because you know the exact population parameter.
Does FPC apply to sampling with replacement?
No. FPC only applies to sampling without replacement. With replacement, each draw is independent of previous draws, and the population effectively remains infinite.
学习