Feb 14, 2008
The Rocky Mountains
I definitely agree that the world needs more applied statistical and critical thinking skills. The watchout is that a lot of these stats folks get paid huge money for high expectations they won't be able to meet. Why is that? Because they may be brilliant, but without context inferential statistics can lead one down the wrong path. I remember watching a jr. process engineer come up with a great model to explain an event in a manufacturing facility. It seemed to have great power and an excellent fit. The problem was that he bucketed two independent variables into one because he didn't realize that there was an extra knob present. When the standard work and control schemes were reissued without mention of this operator controlled setting everything went really good until the cooler months came. At that time a heinous amount of scrap was being generated and consumer complaints peaked as well. When some of us with knowledge of the product and manufacturing system were sent back through and rebuilt the process map we found the obvious UHOH. It wasn't the PE's fault because the guy was new and should have never been given a project of that magnitude/authority without a mentor.
Last edited by a moderator: