On Sun, 15 Feb 1998, Rol Fessenden wrote:
> You are right, this is mathematically -- and also in the real world --
> obvious. Without meaning to demean you, it is also irrelevant. In a
> world in which continuous improvement is the norm, everyone is pushing
> forward in small increments in their capabilities. Someone who is not
> pushing forward begins eventually to lag behind. It may be that the work
> does not galvanize that person, it may be any one of thousands of reasons.
> The point is still relevant: this person is lagging behind his or her
> peers in performance. While your point above is theoretically accurate,
> in the real world, no one is making revolutionary strides forward. It
> typically takes several years for a person to noticeably lag behind. That
> means that for several years they were either not motivated, or not able
> to continue to improve when everyone else did continue to improve. This
> person needs one of two things, either a) more motivation or personalized
> (focused) attention, or b) a different job.
What you seem to be describing is an individual who, in Deming's terms, is
performing below the system. Deming provided a number of statistical
methods for determining whether or not an individual is performing below
the system, and agreed with you that such an individual's performance
needs to be addressed using your a) or b). I am not aware of any cogent
arguments against this.
But the important point is that you *cannot* determine whether or not
someone is performing below (or above) the system by looking only at his
rank in a single round of performance evaluations, and the latter is what
the systems that many of us have objected to try to do. By definition, in
any round there's going to be one person who will be outperformed by all
his peers. If he's performing within the system than 1) assuming nothing
changes and that successive performance evaluations are applied
independently, i.e. no "halo effect," his rank in the next evaluation
could be anything at all and 2) individualized attempts to improve his
performance (as opposed to attempts to improve the system) will be at best
useless and could in fact worsen his performance (by getting to fix
something that isn't broken; see the results of Deming's funnel
experiment). If he's performing below the system, than if nothing
changes, he will likely continue to perform below the system and
individual help is warranted. But a system based on ranks alone cannot
distinguish between the two cases.
--Eric Bohlman <ebohlman@netcom.com>
Learning-org -- Hosted by Rick Karash <rkarash@karash.com> Public Dialog on Learning Organizations -- <http://www.learning-org.com>