Continued from my last Post:
Steve also writes, in connection with: "What I think
you may mean is how can look at our own experience and
see different value patterns at work. That is not
easy.", the following:
"A harder quiz...
What type of pattern of value are you participating in
when you choose
the healthy drink rather than a soda at the vending
machine? Is the
social status of drinking one drink or another driving
your preference?
Is it all about the the taste of it? Is it the cost?
Is it about the
need for energy? Have you made a rational decision
weighing the pros
and cons of each drink? Do these pants make my butt
look fat?"
====
As I said in my last Post, I'll have a try. Here it
goes for all that is worth:
When confronted with a choice between a healthy
drink and a soda, I start a decision making process(DM
in the foll.) This may result in choosing either the
healthy drink or the soda. Nothing resembling a
pattern yet. Suppose though that I'm confronted with
that choice a number of times; suppose I always (or
most frequently) choose the healthy drink. Here we are
getting close to a pattern of behavior: 'repeatedly'
confronted with a choice between the two, my DM
results always (or most frequently) in the same
outcome: healthy drinks.
I could build a model of that but it's sort of
trivial. Trivial, because, unless I were a behaviorist
(which I am not) I'd be a black box that chooses A
instead of B, because the computer program says to
choose A.
The harder problem is: how do we go from patterns
of behavior to patterns of value? Here we need a
further assumption: our behavior is governed by our
values (thus leaving out instinctive behavior). I
won't go into discussing the validity of the
assumption, let's just take it as "a rule of the
game". I further assume that you'd agree in that,
empirically, there are no patterns of value only
patterns of behavior.
Under that assumption, we could say that the result
of our DM's is determined by our scale of values.
Let's suppose further that there are only two
conflicting values at work: My Pleasure (P) as a
value (which leads to choosing the soda because it's
tastier to me) and My Figure (F) as a value (which
leads to not choosing soda because it's fattening).
Suppose that, for me, P is much higher than F (which
actually is).
Under those assumptions and simplifications I
could easily build a probabilistic (stochastic) model
of the system by giving to P a much larger weight than
to F, so that, out of 100 times of running the DM
process, 95 times chooses the Cola. That's pretty
close to a pattern: "again and again" (repeatedly) a
soda is chosen.
There might be another, far more interesting, way
of getting to a pattern of values. Suppose I keep only
P and F but now I confront myself with multiple
choices: soda vs. healthy drink; greasy hamburger vs.
salad; chantilly &chocolate cake vs. rice biscuits;
walking 10 miles vs. sitting to contemplate Nature
while sipping a G&T, etc. etc. We'd need many many
different choice-pairs, otherwise the results will be
inconclusive. Again we make a probabilistic model,
with P being much 'heavier' than F and, in each pair,
assigning a P or F dependency.
If out of, say, 50 pairs, close to 50 times
our model comes up with choosing an item exclusively
dependent on P, we could say that our model gives a
clear pattern of behavior which, under our
assumptions, may be translated into a pattern of
value.
Moreover, if we are 'satisfied with our model'
or if we think that we 'have discerned a pattern of
value' (which in our context amount to saying pretty
much the same) we could use the model or pattern to
Predict what I would choose in an unforeseen pair,
say, an asado vs. mashed carrots. This after all,
constitutes a much sought after application of models
and patterns to actual situations: to be able to
predict.
A scientist, if she's wise, uses Prediction loaded
with Caution; a model can predict a real- life event
if, and only if, the multiple assumptions and
simplifications used in designing the model are valid
for that event. The same holds of course for inferred
patterns but people tends to forget that; that's why
people loose money in the Stock Exchange and that's
perhaps why people may become so dogmatic about Right
and Wrong.
__________________________________________________________
Sent from Yahoo! Mail - a smarter inbox http://uk.mail.yahoo.com
Moq_Discuss mailing list
Listinfo, Unsubscribing etc.
http://lists.moqtalk.org/listinfo.cgi/moq_discuss-moqtalk.org
Archives:
http://lists.moqtalk.org/pipermail/moq_discuss-moqtalk.org/
http://moq.org.uk/pipermail/moq_discuss_archive/
