ABSTRACT
A different distribution — the t-distribution — is used if the sample size is small. The t-distribution has a similarly shaped curve to the normal distribution, but is more widely spread out and flatter. For the use of a t-test to be valid, the data should be normally distributed. Although the test is described as ‘robust’, meaning that it can withstand moderate departures from normality, severely skewed data are unsuitable. For small samples, the Wilcoxon signed-rank test can be used instead of the paired t-test, and the Wilcoxon rank-sum test or Mann-Whitney U-test can be used instead of the independent t-test. These methods are covered by many more detailed texts. There are several different t-test formulae which are used in different situations. This test is used to assess the difference between two paired measurements. It tests the null hypothesis that the mean of the difference is zero.
