RE: Fighting The Merit Cycle- Help!

[Myron Tribus <mtribus@home.com>]

The question has been raised: "How do I get management to accept that 
statistical variation is important?"

Here is a method I used about 18 years ago and it has never failed. 
First I present the following table showing the number of errors 
attributed to 8 workers over a 12 week period.

Week    1   2   3   4   5   6   7   8   9   10  11  12    sum

Mary    1   0   0   0   0   0   0   0   0    0   0   0     1
Jack    0   0   0   0   0   0   0   0   0    2   0   0     2
Fred    1   0   1   0   0   1   1   0   1    0   0   2     7
Bert    0   0   1   0   0   0   1   0   1    0   1   0     4
Jean    0   0   0   0   0   0   0   0   1    0   0   0     1
Sue     1   0   2   0   1   0   0   0   0    0   0   0     4
Eva     2   1   2   2   0   2   0   0   0    2   0   1    12
John    0   0   0   0   1   0   1   0   2    0   1   0     5

Then I tell the manager that one of his Supervisors has come to him 
with the above table and seeks his advice.  What would he advise this 
Supervisor to do?

I get various suggestions such as give Eva more training, let her sit 
next to Mary and learn, fire her, put her on warning, etc., etc.

Then I tell the manager that the numbers were actually generated on 
my computer, using a random number generator to produce a Poisson 
distribution and that after the results of eight runs of 12 trials at 
a fixed error rate, the results were assigned to the set of 
fictitious names.  What did they think about their advice now?

Several said to me: "I know that the numbers came from a random 
process, but nevertheless I would still talk to Eva"

I presented this before a gathering of about 200 statisticians in 
the UK on the occasion of a celebration of the birth of Sir Ronald 
Fischer.  Not one of the people in the audience suggested a 
statistical analysis of the data to see if Eva were outside a 3 sigma 
limit!   Furthermore, one gentleman in the audience said; "Have you 
noticed that there is a  30 day periodicity in Eva's performance?"


My point is: "No sinner was ever saved in a single sermon."  Don't 
rely on one experience (red beads or whatever) to do the job.  It is 
hard and often frustrating work.  Learn to enjoy the battle.

Myron Tribus,  350 Britto Terrace,  Fremont, CA 94539
Ph:510 651 3641  Fax: 510 656 9875   e-mail: mtribus@home.com
If it goes without saying, it goes better if you say it (An old Hindu proverb).
=========================================================================