Skip to main content
. 2021 Jan 26;74(1):23–29. doi: 10.4097/kja.20662

Fig. 1.

Fig. 1.

Comparison of parametric and non-parametric methods in sample size estimation. (A) parametric method : paired t-test with equal variance, (B) parametric method : Student’s t-test with equal variance, (C) non-parametric method : Wilcoxon’s signed rank test, and (D) non-parametric method : Wilcoxon’s rank sum test; δ0 (=0) is the null difference, δ1 (=1) is the actual difference, μ2 (=0) is the true mean of group 2, σ (=1) is the population standard deviation, N1 (= N2) is the number of items sampled from each population, and N is the total sample size.