A statistical power analysis was used to calculate the required sample size. A medium effect size should be substantial so that it can be detected by the naked eye. The repeated-measures design in this study had two groups of 69
participants each, for a total of 138 participants. Each participant was measured three times. The between-participant standard deviation (SD) was 18.78 (effect size = 0.27) and the within-participant SD was 16.26 (effect size = 0.15). This design achieved 87% power when a F-test was used to test the group factor at a 5% significance level. A multistage sampling method was chosen (n = 140; 70 participants per group).