From Greg Williams (930326 - 2)
Rick Marken (930326.0800)
I was assuming (and Bill was too) that Ken Hacker shares many of
the same goals of PCTers -- such as, the goal of gaining some
knowledge about the processes underlying the phenomenon of human
behavior, both individual and collective.
I believe that assumption is correct (maybe we could ask Ken -- but,
uh, that's another thread!), but my concern was Ken's explicit
complaint that some other scientists' aims (and hence their methods)
were being deprecated by some "arrogant" folks on this net. I could go
back and get the quote if you want.
I don't see where Bill, Mary or I (we've been the only participants
in this discussion) have supposed that our goals are the only or the
most important ones. What gave you that impression?
I got the impression from such comments as (paraphrasing here) non-PCT
behavioral scientists and/or pseudo-PCTers don't know what a real
model looks like. This sort of line, TO ME, implies that "real models"
are where it's at, and anything else is, well, to paraphrase again,
not actual knowledge. I guess I tend to read between the lines some,
and what I get is the "flavor" that modeling individuals is much more
important for Bill, Mary, and you than other methods of science
(especially statistical). Now, that's fine, as long as it is more
important because it is the appropriate (i.e, efficient) technique for
reaching (or attempting to reach) YOUR OWN goals. The "arrogance"
which I believe Ken was concerned about is in hinting (if not
explicitly saying) that OTHERS' goals SHOULD BE less important FOR
THEM than goals similar to yours. It is not arrogance to point out
that certain techniques will fail or work poorly for achieving
someone's aims; it is arrogance to claim that those aims are not as
"important" as your own. And the latter is, I think, what Ken was
perceiving.
By the way, I think its fine for people studying population
level phenomena (insurance companies, polling agencies, etc) to
use sampling statistics. Their goals are different from mine (in
PCT) and that's fine with me. I have no interest in trying to
derive actuarial phenomena from individual behavioral models.
Then you aren't arrogant, even though you might sound that way to some
netters. (:>)
Bill Powers (930326.0830)
Oh, heck. Well, is it arrogance to suppose that one's own
definition of arrogance is the only, or even the most important
one? What is the most important goal, then? And where can I look
out there to find it?
PCT itself claims a radical individualistic stance for the notion of
"importance." We each decide what views, goals, and definitions are
most important TO US. Others certainly might have differences of
opinion about the importance of any of these. One manifestation FOR ME
of arrogance is assuming that others should ascribe the SAME
importance to one of these as you do. As long as you say something
like "if you want to do so-and-so, I think you'll find that a good way
to do it is via thus-and-thus," then you are in the nonarrogant
realm. When you say something like "this is what you should want to
do, and doing something else isn't as good," you are simply saying
(arrogantly) that your own goals are better (i.e., "more important"
or "real knowledge") than the goals somebody else prefers.
Seems to me that I've commented a number of times on the fact
that people who deal with populations do fine (for themselves)
with statistics, because statistics concerns the properties of
whole populations, not individuals. When that's your concern, who
cares if a few of the ants get stepped on?
Are you suggesting that the concerns of people who deal with
populations are in some sense unseemly, because "a few" individuals
generally get stepped on? Are you suggesting that these people could
use PCT to do a better (to you) job in some way? If so, how? I said in
my last post that it appeared extremely unwieldy to develop population
measures from individual models. Perhaps your preference is for the
population dealers to simply quit dealing?
I would suggest that there are some human goals which can be
met more efficiently by using descriptive statistics than by
using the (again, "Grand Method") of individual models.
Whose goals are you talking about?
Insurers, epidemiologists, some government workers, statistical
physicists, many agricultural researchers and engineers, etc.
If you own an insurance company, it's most efficient to use
statistics to set the rates, to raise the rates on people who make
claims, and to reject applicants who belong to high-risk populations.
That's the most efficient way to satisfy your personal human goal of
making money out of selling insurance to a population.
Personal human goals are the only ones we've got. You seem to be
implying sleaziness or selfishness again here. Do you really know that
much about the people who use population statistics to say that what
they (some? many?) are doing is "bad" ethically?
When will your modeling and associated methodolgy be
sufficiently sophisticated to predict when I will COOPERATE and
when I won't?
Probably never. Predicting behavior is not what PCT is about. PCT
is about understanding behavior, and what it is being used to
control.
OK, the challenge becomes: show me you understand my behavior and what
it is being used to control. But first, YOU tell ME what a passing
grade at such "understanding" would consist in. You speak of highly
accurate PREDICTIONS in tracking experiments; why not also in more
complex experiments? It sounds to me as if you are going back on the
promise of Glory Days (sometime in the future) when PCT experimenters
can predict .99+ correlations in more than tracking trials. At any
rate, I'd like to hear more about "understanding" as opposed to
predicting.
If I ask you to do a tracking task and you do something
else, it will be clear that your goal is not cooperation, but
something else. I might be able to find out what the something
else is by interacting with you long enough.
I'm ready to try the experiment when you are.
And when PCT has leapt that hurdle, the next one is to make
models for THOUSANDS of individuals and combine them some way
to predict population measures. Lots of luck.
Why would I want to do that? I'm not trying to make a living by
selling insurance or proving that I am -- on the average-- a
successful doctor or educator or politician. My interest is in
understanding the next individual I meet, by some means that
doesn't involve formalized prejudice.
YOU don't want to do it, but MANY OTHERS would, IF PCT could provide
more reliable population measures. The big question is: can it? When?
If you don't think it should (or can) be used that way, then how is
PCT going to help these folks? Perhaps you think it SHOULDN'T help
them, because they are misguided?
No, it is better to follow the example of physics and stick
with descriptive statistics for generating SOME kinds of
knowledge.
Better for everyone?
Obviously not. That's why I said "SOME"! You can ask individuals what
they want to do. (And you can tell them that they shouldn't want to do
THAT, but that's not likely to get you very far. You'll probably just
be labeled "arrogant.")
But I am not arrogant enough to think that modeling individuals
is the only path to knowledge. It IS the only path to SOME
kinds of knowledge. But some people don't need that kind of
knowledge.
Not for their professions, I agree. What about for getting along
with their families and friends and the salesperson and the
waitperson and themselves? It seems to me that by relying on
statistical generalization in such person-to-person
circumstances, people create more problems for each other than
they solve.
That is quite possible. But it remains to be seen whether or not the
individual modeling method can do any better. This is because it has
not been attempted to the degree where a verdict can be given. I'm
game -- have been for several years -- to try it out, but if I were to
prejudge the outcome and say that (sometime in the future) PCT will
allow solution of interpersonal problems which are now exacerbated by
statistical generalization (yes, even the fallacy of applying
population measures to individuals, since "good" results can
sometimes come from mistakes!), I would be arrogantly predicting the
future to match my own preferences.
I think what you're saying basically makes sense: use statistics
for appropriate purposes, and models for other appropriate
purposes. Partly this is just a practical matter of what we
currently know how to do. Weather modeling doesn't work very
well, and perhaps can't, so weathermen also use statistical data.
Even when a model would in principle be the best tool, if you
don't have a model developed well enough to use you fall back to
relying on generalizing from experience just as people have
always done.
Yes. And I'm also saying that it is still to be determined whether PCT
models will ever be efficient for understanding some population
effects.
What people really want is a way to predict the behavior of
others within the very small subpopulations of people with whom
they are likely to interact.
WHICH people want this? ALL people? I think YOU want it, and are
extrapolating to what you think others SHOULD want. I don't think all
others "really want" this. Where's your evidence?
The PCT model is that kind of model. At present, it can say such
universally true things only in simple circumstances like the
rubber-band experiment.
Where's your evidence that the PCT approach is going to be
sufficiently generalizable? That's the kind of evidence people want in
order to not characterize a statement like "Here's the way you should
be doing it" as arrogant.
You might say that where we can't use the PCT model, why not use
the statistical approach, because it's all we have?
The problem is convincing folks that the PCT model will do a better
job FOR THEM.
As ever,
Greg