Sample size analysis

Topic overview

Our journal club enjoyed Nataraja et al.'s randomized trial: lavage vs suction in laparoscopic appendicectomies [1]. Section 1.4.1 considered two distributions (means 4.4 and 5.5 days, sd 1.29), and asked how large, equal sized, samples needed to be such that the chance of getting a p-value  > 0.05 was about 20%:— this being the chance of making a Type II error, of wrongly accepting the null hypothesis when the parent populations differ. The Type II error is less well understood than the Type I error, represented by the p-value, being the chance of wrongly rejecting the null hypothesis when there is no difference between the parent populations.

Keywords

Hashtags