Science of groups, Limits of patience

[From Rick Marken (920107.0900)]

Von Bakanic says:

The following passage sent immediate
error signals in my own understanding of the science of groups:

< ... Psychology is really the science of groups; unfortunately it thinks
<its about individuals and this is mainly self-deception. Now that I think
<of it, the only places PCT has been able to have ANY impact on conventional
<psychology is where psychology is REALLY about individuals....

Since I wrote the passage I'll try to assuage your error signals. My claim
that "psychology is really the science of groups" was based on the fact that
most psychological "facts" are based on statistical averages over several
people. When a psychologist says that "a person in condition A makes riskier
decisions that a person in condition B" s/he is talking about the AVERAGE
person -- because this claim is based on average results. The
statistical inference procedures that are used to test this result are not
asking "what is the probability of this result not being true for every
individual?"; they are asking "what is the probability of this result
not being true of the population from which this group was drawn?".
A statistically significant result is one where there is a sufficiently
small probability that the null hypothesis is true (so we reject the null
hypothesis) -- and the null hypothesis (in a t test, for example) is that
there is actually NO DIFFERENCE between the POPULATION MEANS; it is a
hypothesis about GROUPS -- explicitly. This has all been pointed out before;
there is an excellent old book about this by Bakan ("On Method"). So
psychology is about groups (at least, based on the way they go about their
business) whether they like it or not.

If psychology is really the science of groups, what pray tell is
sociology? Self-aware conventional psychology? Does this imply that
sociologists are a lost cause as PCT converts?

Sociology is EXPLICITLY the science of groups -- and sociologists are often
happy to use psychological data because they know (even if the psychologists
don't) that it is about groups. Group behavior is interesting; the PCT
perspective on it is simply that it is the result interaction
between autonomus, hierarchical control systems. The CROWD program is a
GREAT illustration of what PCT can do for sociology -- it shows how the
interesting group behavior discovered by sharp eyed, clever sociological
observers like Clark McPhail and Chuck Tucker, can be modelled as the
result of interactions between control systems controlling for rather
simple perceptions. PCT will point sociologists to the individual for
explanations of the group phenomena they observe; PCT will point psychologists
to components of the individual to explain the individual behavior they
(could)observe; PCT will blur the distinction between disciplines like
psychology and sociology and biology, but so what? It's all control.

And what is implied in the
use of "REALLY". Is this a truth claim? I thought you folks were
interested in perception, not "reality". Can anyone clear up my confusion?

I remember some time ago Bill Powers explaining this in terms of
the steps toward nirvana (I think). First it's reality; they you learn
PCT and find out that it's all perception (which it is); and then you
become an expert in PCT and can talk about it as reality again (because its
a lot more awkward to say "I would like to perceive you picking up that
perception of an apple and putting it in my perception of me hand" than
"please give me the apple".

From Greg Williams (930107) --

As I've said before, Newton was the Skinner of his day:

Kind of insulting to Newton, I think. I think Skinner was the
P.T Barnum of his day.

I found that quote, and I still have considerable sympathy for Skinner's
pragmatism and humility.


I can only come up with a mathematical description
relating observable inputs and outputs, which, I contend, can be as
predictive as the most predictive PCT model.

Is the point of this that SR models can be just as good as control
models of control? What's the point here, Greg. It looks like there
is virtually nothing that could be done to convince you (it is you,
now, right, not Dr. Diabolo?) that the PCT model is essentially better
than an SR or output generation model of control. I feel like I'm
listening to the local flat earther. One reaches a point where it becomes
clear that nothing is going to sway either side. All one can say at this
point is "enjoy your beliefs and have much success". I think this is what
happens in "real" science too; at some point each side just has to say
(along with Bob Dylan again) " well, it's just like you go your way and
I'll go mine". I think it's clear that people who find value in
operant analysis, SR models, planned output models, etc etc are not
going to change their minds about it; NOTHING WILL CONVINCE THEM (as
Greg has made so clear with respect to SR models). What evidence we
(PCT) have of problems with these models have already been presented;
I think it's time to just go off and do whatever follows from one's own
point of view (or model) of behavior. I'm going to do the conflict research
(as time permits) and not worry about the fact that "conventional"
psychologists would not find it interesting at all. As I said in the
intro to "Mind Readings" I've had it with showing what's wrong with
conventional approaches to psychology (a useless effort) -- now I'll
just do PCT psychology -- at least there are a couple of people who
will talk about it with me -- and appreciate it. It's too bad there are
not more people who are interested but that's the way it goes.

Of course, when my patience returns I might be willing to DISCUSS the
problems of SR and output generation models of behavior -- but, I will
do it knowing full well that it will be of absolutely no use.



[Martin Taylor 930107 14:00]
(Rick Marken 920107.0900)

Who was it said "It ain't so much what you don't know as what you
know that ain't so"...

Of course, when my patience returns I might be willing to DISCUSS the
problems of SR and output generation models of behavior -- but, I will
do it knowing full well that it will be of absolutely no use.

I think it is always of some use. Every return to the theme clarifies
matters on minds with both presuppositions. Weren't you of the other
persuasion once yourself? Some of the discussions might just hit the
right button for some people.

I'm working on the "information theory leads to PCT" paper. Since I have
to go right back to the basics of what "probability" means, and thence
to statistics in the real world, it becomes clearer that although PCT
hangs everything together in (so far as I can see) the only realistic way,
nevertheless there are useful findings to be obtained from both S-R and
preplanning studies. The problem with each is that the findings are often
hard to use in real life situations, outside the tight laboratory controls.

As to whether the results from one person are useful in dealing with another,
that depends entirely on what it is that is being measured. I have 30-year
old group (as you would say) data on the timing of reversing figures which
are every bit as tight as any PCT prediction of tracking. I used it to
argue that instructions can bias what people perceive as opposed to what
they say they perceive (Canadian J. Psychol, 1963, 17, 210-223, esp fig 4).
The worst deviation of any point from the prediction line is about 7%, and
most are so much closer that I can't determine the error of prediction. And
the results from the two groups are closer to each other than they are to
the prediction line, so there was something wrong with the prediction, even
if the error was very small.

Although it does not involve group data, you (Rick) might also be interested
in a follow-up study on perception of a reversing figure by individual subjects,
in which we found that one subject seemed to devote 26 "elementary perceiving
units" to the perception of the figure, while the other devoted sometimes 32
and sometimes 33 units. The data fits were perhaps not formally statistical,
and the subjects were treated individually, but this is a pure S-R study
that I think says something important about the perceiving system of the
subjects. Now I think it says something about the distribution of perceptual
control among non-orthogonal ECSs.

In the past, metaphors like gold nuggets among the garbage have been used
on CSG-L. The metaphor is reasonable, but the question is whether the
nugget-to-garbage ratio is low enough to allow a real psychologist (PCT
oriented) to discard without consideration what fantasy psychologists
(non-PCT oriented) have done. I think there is a lot of good stuff that
can be used. Bill and Rick and others think not. We shall have to see,
but it will take a lot to convince me that the N/G ratio is zero.