Re: objectivity via 3D charts

Tue, 15 Oct 1996 14:08:03 EDT

> analytic possibilities of
>"gradations of roles through time, and simultaneity of roles."
>'We could plot these relationships
>on a 3D chart...' or something like that. Perhaps I have become
>brainwashed by too close a contact with engineers over the years, but I
>have come to admire their crazy 3D charts graphing 3 variables at once,
>printed out as "carpet plots." ...

In some language and culture courses and methods courses, I saw 3-d
charts of descriptive language terms. One exercise (of names college
girls called college boys) was done in a "hard" format with styrofoam
rods -- not really as precise as computer printouts, but based on same
for prodding undergraduate understanding. Mapping/graphing such
variables initially had me skeptical -- even laughing, but later the
image of mapping certain ideas was a powerful mnemonic for what was being
discussed/illustrated. I subsequently saved a wonderful neatly bounded
data set (of nearly all cases rather than a sample) while in the field to
do such a thing with... never got around to it, but the data is in an
ascii file, waiting for the right program and the leisure time to play
with it. Still, this methodology brings home some disturbing thoughts
about objectivity. I, the presenter of the data, would be the one
establishing the categories, no matter how tested... true, the graphing
or mapping lets one see if such categories make sense. The book on
visual display of data which is often advertised in the back of journals
or book review sections which has been out for about five or six years
might shed some light on the depiction of info. The image presented by
the graphs gives one a visual check on reality that is out of my realm of
expertise, but I intuitively know is worthwhile validity check. Anyhow,
as one who appreciates how easy it is to lie with statistics, I have
mixed feelings.... Any statistical types, or seriously into graphing
types who can tell us more???

Patsy Evans