It has been over a half century since I sat in a class room taking a statistical course.  At the time, the department head was a very good friend, Dr. Harry Smith.  Actually Harry and I were both employed at Procter & Gamble, thereby giving me many opportunities to discuss the statistical results of our research as well as the assumptions and theories relating to the research.  Of particular concern, was that, in more cases than not, I was forced to accept a hypothesis that I knew for a fact was wrong. That is, I had to accept the Null Hypothesis when I knew that in the world there were never two things alike.  That significance was only a function of the amount of time, effort and monies spent exploring the differences between the problem variables.  I came to learn that “All things significant may or may not be important,” and “All things not significant may or may not be important.”

Over the past 50 years, I have seen little in publications relating to the topics of significance and importance.  That is, until the January Quirks Publication.  A young man by the name of Richard McCullough, president of Marco Consulting in Palo Alto has written an excellent article on the topic.  I encourage everyone to read his article.

Following are a few of the many points he makes in the publication.  Mr. McCullough can be reached at (650)691-1332 or richard@macroinc.com.

Mccullough’s First Law of Statistical Analysis:  If the statistics say an effect is real, it probably is.  If the statistics say an effect is not real, it might be anyway.

MacLachlan’s Law:  Torture any data set long enough, and it will confess.

McCullough’s Second Law of Statistical Analysis:  Never, ever confuse significance with importance.

Corollary to the Second Law:  If it doesn’t make sense, don’t do it.

The Rolling Stones Law:  You can’t always get what you want.  But if you try, sometimes, you just might find, you get what you need.

McCullough’s Third Law of Statistical Analysis:  If you can’t tell what it is, it ain’t art.

McCullough’s Law of Small Samples:  Give me a sample small enough and all means will be statistically equal.

McCullough’s Law of Large Samples:  Give me a sample large enough and all means will be statistically significant.

McCullough’s Law of Cluster Analysis:  The name of the cluster is always more important than the cluster.

McCullough’s Law of Statistical Fashion:  Give a kid a hammer, and the world becomes a nail.

McCullough’s Law of Marketing Impact:  Any direction is better than no direction.