The following real life experience emphasize Richard Smith's statement "I
maintain a very strong opinion that comparing one months activity to another
just doesn't tell you much."

In 1971 while a graduate student I started working as a "casual" (i.e. part
time, temporary) employee at the Federal Records Center in Ft. Worth with my
primary job being pulling/refiling Income Tax Returns that were less than
3-years old.

I went to my first formal evaluation feeling good about my performance
because I was more mature and conscientious than most of my peers.  Mr.
Garza pulled out a file that documented the number of files each of us had
worked and had statistics showing how many files the average person worked
per hour.  My average was well below normal.  I was shocked and disappointed
in myself even when Mr. Garza stated that though my numbers were below the
norm they were still acceptable.

But Mr. Garza went on to explain why my numbers were low.  First, there was
my gender.  As a group women were faster than the men because of their finer
motor skills.  (He also noted that men as a group were superior to women
when the job called for muscle such as shelving, shifting, or pulling entire
groups of boxes.)

Second, I came to work early in the morning and worked the batches of
requests and refiles that had been left over from the day before.  Others
came in later after the new batches arrived and, human nature being human
nature, worked the easiest ones leaving those of us who came early with the
more difficult, time consuming ones.

He emphasized despite the fact that my statistics were below the mean
average my performance more than adequate.

So the bottom line is that statistics is that can identify performance
outliers that merit further analysis but competent supervisors must be aware
of factors that do not show up on graphs before judging performance of
individuals or groups.

Paul R. Scott CA, CRM
Records Management Officer
Harris County, TX

List archives at http://lists.ufl.edu/archives/recmgmt-l.html
Contact [log in to unmask] for assistance