- The document discusses small sample inference and hypothesis testing for means and proportions. For small samples, the t-distribution rather than the z-distribution must be used because the sample standard deviation s is a less precise estimate of the population standard deviation σ.
- The t-test corrects for this by using t-scores and the t-distribution table rather than z-scores and the normal distribution table. It demonstrates a t-test example for a small sample study on anorexia treatment.
- For small sample inference on proportions, the binomial distribution must be used rather than approximating it as normal. It provides an example hypothesis test on gender bias in manager trainee selection.