The analyze and measure stages go together. The measure stage identified a large number of factors, or X's, that may have some impact on the output, Y. These are potential or likely factors for the output Y.

When we analyze these numerous potential factors, we observe that only a few of these have any significant impact on Y. The others are trivial factors in the sense that they have minimal impact on Y. In other words, the analyze stage can be referred to as the vital-few-trivial-many distinction stage.

There are a number of analysis tools that work with data, and there are other tools that analyze the process to reveal these vital few factors. These vital few factors can also be called root causes. Further validation and confirmation of these root causes is crucial as improvement projects involve investment of huge effort and resources. Hypothesis tests provide statistical validation for these root causes. The next stage, improve, then deals with actions to address these root...