The news media often uses the distinction: "If dog bites man that is not news. But, if man bites dog that IS news!"
This humorous, yet discriminating idea is an important concept. When a dog bites man it is natural and expected behavior - "Not newsworthy," as a good editor would say. On the other hand, if a man bites a dog, that is unusual and unexpected and worthy of news time or space.
In studying data and statistics, we must apply the same thinking. Some data, some variation, or some stories are so expected as to be not worthy of being mentioned. Other stories, due to their surprising, unusual and unexpected nature, are truly different and worthy of comment.
A widely reported story from the news services:
[ Thank you, Steven Maul, for making this unusual story available for me to use as a lesson. ; - ) ]
Two terms often used in describing the types of variation are signal and noise. Variation that is expected, usual and typical is called noise. Variation that is really unusual and unexpected is called signal. We must understand the difference in signal and noise before we can properly respond to either. One method to do this is the control chart which will be covered in a later lesson.
Back to Lesson 1 Go to Lesson 3 Home
Copyright © 2001, Michael L. McDonald
