ABSTRACT
Student’s t-test is used to test the equality of the means from two different populations (Rothmann, 2005). Use of Student’s t-test for comparing more than two groups can cause Type I error. This can be better understood from the example below:
Absolute weight of the liver of female mice in a 13-week repeated dose administration study is given in Table 11.1. Table 11.1. Liver weight (g) of female mice in a 13-week repeated dose administration study
Group N Mean ± SD Tukey’s multiple range test
Repeated comparison with Student’s t-test
A B C A B C A 10 1.083±0.057 - - - - - - B 10 1.098±0.077 NS - - NS - - C 10 1.154±0.050 NS NS - S NS - D 10 1.273±0.062 S S S S S S
NS-Not signi¿ cant; S-Signi¿ cant (P<0.05). Repeated analysis by Student’s t-test revealed a signi¿ cant difference
between Groups A and C. Actual increase in liver weight in Group C compared to Group A is only 6.6%. In this case, the signi¿ cant difference between Groups A and C detected by repeated comparison with the t-test is caused by Type I error. When the groups were compared using Tukey’s multiple range test, no signi¿ cant difference was observed between Groups A and C (Tukey’s multiple range test is the ideal test in this situation, since the number of groups to be compared is more than two).
