Sample size depends upon:The effect size (usually the difference between 2 groups). The population standard deviation (for continuous data). The desired power of the experiment to detect the postulated effect (Power = 1 – b) The significance level.