Skip to main content

Table 1 a priori sample size calculations to detect small, medium, and large effects for endothelial function research

From: Statistical Inferences Using Effect Sizes in Human Endothelial Function Research

Effect size Paired samples t test (n = number of pairs) Independent samples t test (n = per group)
Small (d = 0.28) 102 202
Medium (d = 0.69) 19 34
Large (d = 1.21) 8 12
  1. Tests were calculated using 80% power, and assuming a two-tailed test and α = 0.05. Researchers may use alternative levels of statistical power (Fig. 6), however, 80% typically represents the minimally acceptable level of statistical power (Quintana [12])