We now can make calculations that represent the total amount of variance found in each of the groups. This is also referred to as the total square sum of scores (divided by n). First the sum of X for each group is separately squared and then the resultant answer is divided by the group size. These calculations need not be done separately when each of the group sizes is the same; this is true for the case at hand (nl = 8, n2 = 8, and n3 = 8). Because most data sets do not entail equal group sizes, however, each of the three calculations is done separately. X..1 are 120, 75, 55. The next step is really what the formula prior to Steps 4 through 6 asked for: the total square sum of scores divided by n. In simpler terms, to complete Step 7, add together the results of Steps 4, 5, and 6.. Step 8 is an intermediate operation and requires that we sum all the observed X's, square this value, and then divide it by the total number of subjects (TV). In other words, take the value of Step 1, square it, and then divide this