[PLUG] Huh?

Rich Shepard rshepard at appl-ecosys.com
Tue Dec 10 19:21:55 UTC 2002


On Tue, 10 Dec 2002, Paul Heinlein wrote:

> The meta-conclusion of the course was facinating. An economic
> conservative and a Marxian produce different accounts not because one
> is lying or inferior but because they ask a completely different set
> of questions. Ask a Marxian to tell you about Napoleon. Then ask a
> Freudian. :-) The art lies in understandining which questions will
> produce a better history.
 
  Which was exactly my point, Paul. In the book, "How to Lie With
Statistics" (there's now another one called "How to Lie With Maps") the
thesis is that one doesn't have to lie but only select the scale or range
that gives the answer you want.

> Consultants know this intimately. The real drama in any research is 
> framing the questions. After that, it's all boring old engineering.

  In addition to asking the question that will lead to a usable answer is
the task of collecting data and analyzing it properly. By "properly" I mean
objectively and without a priori knowledge of the answer.

  When calculating the Total Cost of Ownership, apparently, one can in- and
exclude factors until one comes up with the desired answer. But, the real
question, IMO, is not the TCO but what solution is the best for the
individual business; best regardless of time span. That does not have a
single, all-encompassing answer but one that varies by company. I strongly
suspect that linux would be the answer in almost all cases, but I'm neither
an economist nor an engineer. Ergo, I don't know what I'm writing about. :-)

  Do you know that if you lay down all the economists in the world, end to
end, you'll never reach a conclusion?

Rich





More information about the PLUG mailing list