Obtaining the sums of squares for each component of the experimental design was just the first step in testing our hypotheses. We now need to convert our sums of squares calculations into variance estimates. We do this by dividing each sums of squares needed in our F ratios by their appropriate degrees of freedom to obtain means squares estimates. These "mean squares" represent average squared deviations of an effect of interest around the grand mean. We know that the variance is defined as the average squared deviation of scores around a mean. Thus, our mean squares for each component are variance estimates.