You bring up a very important point when you mention the importance of clearly defining the "idea" before designing an experiment. It is important to remember that the use of statistical analysis in a practical business setting is a "means to an end." As managers we are primarily concerned with analyzing information that helps us make solid business decisions. Statistical techniques including hypothesis testing, provide managers with the tools to navigate an extremely competitive environment. Therefore it is important that the tool is designed to provide the right information for the decision at hand. Thomas Stewart (2011), provides a very good analogy of this by comparing a business to a large train, and the manager/CEO to a train engineer. While the engineer sits in the locomotive essentially "driving" with controls that are familiar to him, a more important question looms. What if the instruments are broken, not connected to anything, or just completely wrong? This is one of the primary reasons that the sequence is so important. Managers must be confident that the data (instruments) are "connected to something." In other words, the analysis must be tied to the "right questions." Question is: Can anyone think of a business decision (either at your organization or something in the news), where it is obvious that management made the decision based upon information that failed to ask the right questions. Unfortunately, these situations are often seen in market research (new product testing, customer feedback, etc.)