April Fools in Degrees (from Chris Cherpas)

[From Chris Cherpas (960403.0911 PT)]
     [re: >Bruce Abbott (960401.0820 EST)...and replies]

Reading Bruce Abbott's April Fools post "fooled me in degrees." The
"Subject:" read "Reinforcement Wins" -- already leading me expect
something tongue-in-cheek. However, at this early point, I either
did not have any awareness of the date of the post being 4/1/96,
or I did see the date in the vicinity of "Reinforcement Wins" but did
did not "put together" the association between a dubious Subject
and the fact of an April 1st post.

Bruce's first paragraph refers to recent data (we believe to be factual,
although now those posts are in question!) he had posted about an
experiment. After reading this, my doubts reduced and the balance
was definitely tipped in the directiong of being fooled.

But Bruce's next paragraph begins:

The study was cleverly designed to pit reinforcement against control in the
following way.

I doubt if Bruce would seriously refer to his own experiment as
"cleverly designed." My degree of being fooled dropped considerably
at this point.

A procedure is then described which, as Bill P. points out, can be read
as containing a crucial contradiction. However, I read this as just
loosely describing an "adjusting" schedule, in which a single response
_is_ immediately followed by food delivery (i.e., there is no time
delay between a response and a food delivery), but that responses
also incremented a counter which would subsequently have to time out
for the next response to be (immediately) followed by a food delivery.
The ambiguity could be resolved.

This kind of experiment is actually quite interesting in that it pits
a global economic maximizing model against a local sub-optimal model
of adaptive behavior. I didn't think Bruce was doing this experiment,
but since he _does_ run "cleverly designed" experiments, I began to be
fooled again -- this time _hoping_ it was true, since this is a general
area in which I did my dissertation. But if it were true, there still
had to be something fishy going on since I knew that Bruce would
not believe that PCT would just _assume_ that overall food consumption
rate was a controlled variable.

By the time Bill P. was "quoted" as saying "Skinner was right" I knew
that I had been had, but, frankly I _still_ wanted to check out this
interesting procedure with Bruce! After all, this sounded like a
"Melioration versus Maximization" (a favorite) type of experiment.

The rationalized moral for me: behaving (me) rationally under uncertainty
involves a sub-optimal process more like Melioration than Maximization.

The simple moral: I've just _got_ stop being such a fool!

Best regards,