In Example 4.6.1. Resistant line for the IO-CPU time, we had
IO as the only independent variable which explained the variations of the
CPU time. In many practical problems, the dependent variable depends on more than one independent variable. In such cases, we need to factor the effect of such independent variables using a single model. When we have two independent variables, and median polish helps in building a robust model. A data display in which the rows and columns hold different factors of two variables is called a two-way table. Here, the table entries are values of the independent variables.
An appropriate model for the two-way table is given by:
Here, is the intercept term, denotes the effect of the i-th row, the effect ...