ANOVA or analysis of variance is when a group of statistical models are utilised in the data analysis of variations between the group means and the processes related to them. A T-test is when two statistical models are utilised in the analysis of variations between the two means and a small sample size is associated to them. The T-test is limited to the testing of two means but ANOVA can handle more than two means. These means are derived from groups that have been formed within the sample population itself based on a certain detail or for the requirements of the process analysing the population. For instance consider the splitting up of 20 men into 4 groups to find out which exercise has the maximum effect on them.