Two-step analysis is also called as "two-step design". This is the method to analyze the data of experiment.
Second step of two-step analysis is made from the concept of robust design .
As a concept of design,
two-step analysis is
"At first, you should make a technology to reduce unevenness. Second you should do to get target average."
At first, we look for the condition that unevenness is minimum. Then we look for the condition to get the best average.
The case, that same factor affects both unevenness and average, is not unusual.
NEXT Kinds of Unevennes