Minitab determines categories using the following iterative methods:
Defining the last category
Conceptually, defining the last category is similar to defining the first category, but Minitab works backward starting from the largest data value.
The last category is " xj ", where xj is the largest data value greater than (1 + the data value of the first category), such that the category has an expected value greater than 2. The probability and expected value for the last category are pj and N*pj respectively, and the observed value is the number of data values xj .