The analysis of variance may be described as an hypothesis test which is used to make comparison between variables of two or more independent groups. The null hypothesis is always of the notion that there is no difference in the means. While the alternative hypothesis is the opposite, for two independent groups, the alternative hypothesis is that both means are different, or not equal or not the same. However. When we have more than 2 independent groups, then the alternative hypothesis is stated as : 'the means are not all equal'. This means that the means of each group does not all have to be different, but the mean of one group may be different from that of the other groups or the mean of two groups are different from the other groups and so on.
