In an effort to summarize how to think about sums of squares and degrees of freedom and how this translates into a model that can be implemented in SAS, Dr. Rosenberger walks you through this process in the videos below. Pay attention to the subscripts and these are the keys to understanding this material.
Part One
Part Two
Typo at 2:30: The equation should be \( \sum_i \sum_j \sum_k \sum_l (\bar{Y}_{i \cdot\cdot\cdot} - \bar{Y}_{\cdot\cdot\cdot\cdot} )^2 = bcn \sum_i (\bar{Y}_{i \cdot\cdot\cdot} - \bar{Y}_{\cdot\cdot\cdot\cdot} )^2 \)