ABSTRACT

In this chapter, the authors aim to choose expressions to make our analyses neater, which means making their results simpler and easier to understand. The most important way for an analysis that involves two separately changeable factors to be neater is to have ‘the effects of the factors be additive’ which means if they change the values of one factor from C to D, the change in response is the same whether the value of the other factor is E or F. If, for a particular data set, some other modification of the standard row plus column fit does a good job of absorbing the interaction in that analysis, using the corresponding extended fit and unabsorbed interaction.