That Dog Don't Hunt

This one goes right along with bootleg data.

Let's say you are in the middle of your MSA (measurement system analysis), and you are using a database that holds your average build time for a circuit board, the combobulator (the opposite of the discombobulator). This database was developed years ago and is the primary source of information for plant layout and process setup decisions. Of course it is made up of a super secret formula that Joe came up with six months before he left, and it has little to no useful documentation.

Now you are bringing me results based on this database reporting system, and telling me that your gauge passes with a Likert scale score of three by using the database a listening post? That dog don't hunt.

Bootleg Data

Bootleg data. Another Finneyism. A few ways to tell you might have bootleg data:

• If you spend more time on the animations for your powerpoint than you do collecting your data, you probably have bootleg data
  
• If your R-squared number for your regression equation is 11.3%, you probably have bootleg data
• If you "knew" what the problem really was before you even started your project, and all data you have collected so far confirms it exactly without exception, then you didn't really have a project OR you probably have bootleg data
  
• If your boss "knew" what the problem really was before you even started your project, you probably have bootleg data
  
• If you skipped your MSA (measurement systems analysis), you don't even know if you have bootleg data or not