[spam] Re: Suggested Projects for Class on PCT

[From Bill Powers (2007.01.01.0540 MST)]

Very hard to get the fingers to type 2007. Happy new year to
all.

Martin Taylor 2006.12.31.22.57 –

I think you might enjoy the
attached, from the current issue of American Scientist.

Thanks. Most interesting. One of the remarkable features of this paper is
that it shows how to improve treatments, on the average, without knowing
ANYTHING AT ALL about what is wrong with the patient. By examining
symptoms, one can determine risk factors for subgroups, and then use the
results of clinical trials within subgroups to determine treatments. This
gives results that are clearly better, on the average, than using just
one measure of past outcomes for everyone.
Phil Runkel has pointed out that subdividing a population has its own
risk factor, in that this strategy reduces the numbers within each
subgroup and increases the standard deviation of any measure. He calls
this “fine slicing.”
There is a spectrum of fine slicing with the whole group at one end and a
single individual at the other end. Statistical analysis does no good at
all with single individuals – however any judgment about an individual
is stated on the basis of group statistics, it inevitably includes terms
like “in the long run” or “on the average” or
“typically” which refer to the whole population, not one
person. The probably of a single event is zero if it doesn’t happen, and
1 if it does happen, and you don’t know which is going to be the case if
there is to be only a single event (this patient’s experience of
the treatment).

Obviously, the larger the clinical study is, the smaller is the average
effect of a treatment that can be shown to be significant. But by the
same token, the larger the clinical study needed to achieve significance
becomes, the less significant must the measured effect in any given size
of subgroup. The Am Sci paper shows that there is a tradeoff, with
improvements being achievable when the subgroups are based on estimated
risks. But there must be some number of subgroups at which the increasing
uncertainty of measurement offsets the gain due to segregating people
into a larger number of risk groups. And way over at one end of the
spectrum, where the doctor must decide on the treatment of a single
individual and the individual must decide whether to allow it, even the
improved forecasting does that individual no good at all (except perhaps
to make the person feel better about the prognosis, which of course is a
population prediction).

My complaint is not that we shouldn’t do studies like these and use
statistics to try to improve treatments. When that is all we have, it
would be foolish not to do what is possible. My complaint is that until
quite recently, that statistical approach to treatment has been
essentially all we do have, most particularly at the level of patient
treatment as opposed to research. A very large part of the
“education” of a general practitioner comes from the salesmen
representing drug companies, the “detail men” (If women insist
on being included in this dubious group, I will amend the wording, but I
have yet to hear a feminist complaining about phrases like “the evil
men do.”). And the drug companies, by and large, rely far more on
statistics than on science.

I see hints here and there that a new movement is afoot, called (I think)
“systems biology.” I’ve been advocating this for years, but
most of the money spent on research into cures has been of the old
statistical kind, with very little support for trying to understand how
the damned thing works. Systems biology, as I understand the hints, is
aimed at dealing with all the variables that are simultaneously
interacting in a living system, instead of looking for simplistic causal
relationships between one variable and a whole complex of symptoms.

The systems approach is what we need to deal with people at the specimen
level. It’s what we need in order to say what is wrong with someone who
exhibits the symptoms we carelessly label with words like
“schizophrenia” or “measles” or
“headaches.” It’s what we need to start recognizing that
side-effects result from real and important changes in variables in the
whole system that are caused by the treatment. You can’t affect just one
thing in a system.

We have to get beyond the statistical approach before we can start trying
to say what is wrong with a person who exhibits certain symptoms.
“Depression” is not a name for a disorder, it’s a name for a
set of symptoms that arises from some as yet unknown underlying
malfunction. The same holds true for most of the other names we give to
maladies. Because putative “sciences of life” have relied so
heavily on statistics, what they think they know about the human system
is at a terribly superficial level, like the level of the cargo cultists
trying to bring prosperity back by cutting crude landing strips through
the jungle. Treating symptoms is what we do before we have science. The
basic premise of statistics is that the future will (for no reason at
all) be like the past, which is no way to get anywhere.

Best,

Bill P.

···

[From Bill Powers (2006.12.29.1115 MST)]

Rick Marken (2006.12.29.0945) --

But when I am on surface streets and in a rush and I find myself behind some slow moving clod, I will get as close behind as I can get and -- if the person is obviously old and infirm or has a pro-Bush bumper sticker-- honk and flash my lights;-) Obviously, there are some higher level goals in there that have a lot to do with where I set my reference for how closely I follow in any particular situation.

Hmm. As an old infirm person, when someone honks and flashes lights close behind me, I assume it is a stupid redneck conservative who voted for Bush, and I slow down more and more, to a stop if necessary. I have been known, a couple of times, to block the road, get out, go to the driver behind me, and ask in a grateful manner "Were you trying to get my attention? Did you see something wrong with my car?" Both times I got an embarrassed denial and apology.

I take it that your reference conditions do NOT include avoiding expensive repairs to the front of your car. Or your front teeth.

Best,

Bill ("Road Rage") P.

[From Rick Marken (2006.12.29.1050)]

Actually, I forgot to mention that when I'm not in a rush myself I often get tailgaters, like myself, tailgating me, and, of course, I deal with it by slowing down even more (unless I suspect that the tailgater is armed). My point was that there is really no "tailgating" trait that determines whether or not one tailgates; it's all hierarchical control. Sometimes I'm an tailgater; sometimes I'm a tailgatee;-) It depends on what other goals I'm trying to achieve at the moment.

And I don't really honk and flash my lights. But, yes, I have tailgated slow pokes -- and I hate myself for it;-)

Best

Rick

Bill Powers (2006.12.29.1115 MST)--

Rick Marken (2006.12.29.0945) --

But when I am on surface streets and in a rush and I find myself behind some slow moving clod, I will get as close behind as I can get and -- if the person is obviously old and infirm or has a pro-Bush bumper sticker-- honk and flash my lights;-) Obviously, there are some higher level goals in there that have a lot to do with where I set my reference for how closely I follow in any particular >> situation.

Hmm. As an old infirm person, when someone honks and flashes lights close behind me, I assume it is a stupid redneck conservative who voted for Bush, and I slow down more and more, to a stop if necessary. I have been known, a couple of times, to block the road, get out, go to the driver behind me, and ask in a grateful manner "Were you trying to get my attention? Did you see something wrong with my car?" Both times I got an embarrassed denial and apology.

I take it that your reference conditions do NOT include avoiding expensive repairs to the front of your car. Or your front teeth.

Best,

Bill ("Road Rage") P.

Richard S. Marken Consulting
marken@mindreadings.com
Home 310 474-0313
Cell 310 729-1400

David Goldstein (2006.12.30.1004)

Rick,

If we were doing a Q-Methodology study, your statements might make it into
the Q-set. Namely,

"I usually keep a nice distance from
the car in front, proportional to speed: the faster I'm going, the
farther I stay behind the person in front. "

"When I am on surface streets and in a rush and I find myself behind some
slow moving clod, I will get as close behind as I can get and ... honk and
flash my lights."

I don't understand how you can say that:

"I hope to help students understand
that how a person is categorized on some dimension (such as their
attitude regarding tailgating) has no force on how they control (such
as the variables involved in driving)."

Your own statements seem to contradict this.

David

[From Bruce Nevin (2006.12.30 17:51 EST)]

Rick Marken (2006.12.29.0945) --

Speaking from personal experience, I usually keep a nice distance

from the car in front, proportional to speed: the faster I'm going,
the farther I stay behind the person in front. ... Obviously, there
are some higher level goals in there that have a lot to do with
where I set my reference for how closely I follow in any particular
situation.

The recommendation used to be a car length of separation for every 10
mph. If you're driving 70 mph, 7 car lengths. Now the recommendation is
a second for every 10 mph, measuring the distance travelled to reach a
stationary object such as a sign. If you're driving 70 mph, gauge how
far you have to be from something to reach it in 7 seconds, and leave
that distance between you and the car in front.

On multilane highways around here, if you drop back that far, others cut
in front of you, and to maintain the recommended distance you have to
drop farther back. This is a great annoyance to drivers behind you, who
perceive you as driving too slow for the lane you're in. If you're in
the slow-speed lane, it can easily drop your speed below the minimum
speed limit.

A number of goals involve perceived motion relative to other cars, as
distinct from relative to stationary landmarks:

- Going with the flow.
- Going with flow in the fastest-moving lane.
- "Progress in the journey" perceived as forward motion relative to
other cars.
- Another driver cutting in front of you as something to avoid or
prevent.

Other goals:
- Not annoying other drivers.
- Always being in the faster-moving lane
- Avoiding accidents (as my grandfather used to say, 99% are caused by
the other guy)

Drivers may control more than one of these goals at once. Many pairs of
these goals conflict, not always in immediately obvious ways.

A useful student exercise could be just to tease out these variables.

  /Bruce

[From Bill Powers (2006.12.31.0630 MST)] –

Rick Marken (2006.12.30.1700)]

(writing to David Goldstein
(2006.12.30.1004))

So even if my statements made it into the Q-set, the result of the
Q-sort, according to what you say here, would be a set of factors, on
each of which I would “stand” somewhere because of where my
statements were ranked. The factors are derived (as you note) from
the correlations between the Q-sorts for a group of people. So the
factors represent something like “traits”. One of those factors
might be called “cautiousness” because statements like my first
one above (“I usually keep a nice distance…”) load high on
it. So you might conclude from the factor analysis result
that I have a high level of “cautiousness”. Another factor
might be called “punitiveness” because it loads high on
questions like the second (“… honk and flash my lights”). So
I might also end up having a high level of
“punitiveness”. So the result of the Q-sort will be that
I have been categorized as high in “punitiveness” and
“cautiousness”. It looks to me like the Q-sort technique is
(like all psychometric techniques) an attempt to analyze people in terms
of personality “traits” based on the correlations between their
responses to tests (like the Q-sort).

You raise a good point here. It’s worthwhile asking what we mean by a
“trait” and how it figures into PCT. A trait is not a reference
signal, because as you have pointed out now and then, a reference signal
is varied by a higher-order system (except at the top level, where
it may still vary because of reorganization) as a means of counteracting
disturbances while making perceptions conform to higher-order reference
conditions, themselves variable.

And I have pointed out that even constant reference signals at a higher
level can require continually varying reference signals at a lower level:
“running”, for example, or “being a nice
person.”

So what is a trait? Isn’t it something that remains constant over many
different circimstances? Another word for trait is
“characteristic” (noun), and another word for characteristic is
property. A property of something changes only slowly if at all, and can
be used to predict behavior in many different circumstances. In PCT we
often call properties the “parameters of control.” So when we
think of traits in PCT, I think we should be thinking not about reference
signals, but things like loop gain, delay, perceptual organization, and
hierarchical structure. These things change, but only slowly and only
under reorganization.

This understanding of traits could lead to some interesting departures
from conventional uses of measures like Q-sorts. For example, a person
with a high loop gain (“I try to act immediately and strongly when
something goes wrong” might include drivers we call aggressive and
drivers we think of as timid. The aggressive driver is trying so hard not
to fall back relative to other cars that he squeezes into spaces that are
too small, blasts others out of his way with his horn, and so on. The
timid driver is trying so hard to avoid an accident that he shies away
from proximity to anything else – cars, bridges, birds, and so on. The
superficial details of the behavior seem very different, but the drivers
share a trait: high loop gain. In fact, these could be the same driver in
different circumstances – for example, driving his own car just after he
bought it, and driving his old beat-up armor plated Bummer. I mean
Hummer. If he has only moderate loop gain in the relevant control
systems, he will not become overly agressive in traffic in his old car,
nor will he become excessively timid in his new one.

I think trait psychology arose in yet another attempt to mimic physics,
but as usual without the intellectual means to do so. Traits are thought
of as inborn properties which can be used to predict behavior over many
circumstances. It would obviously be very useful to find human or animal
traits, if one’s business were that of predicting behavior. I think the
idea of measuring traits arose on something like that basis – the idea
of “personality,” for example, assumes that there is something
constant and unique in each individual that predicts that person’s
behavior in social situations.

What happened, as has happened in many areas of psychology, is that the
hope of being able to predict behavior foundered on the discovery that
the measurement of traits had such a large component of uncertainty that
the only way to get any consistency was to use repeated measures over
populations and apply the methods of statistics to look for faint
suggestions of regularity. Yet there may still be something like real
properties that differ from person to person, which can be discovered by
looking not for superficial behaviors that appear the same from one
situation or person to another, but for underying characteristics
that remain the same whatever the behavior happens to be. The PCT model
shows us numerous possibilities.

When a driver sees other cars around him, what, exactly, does he
see? Does he see spaces ahead into which he might be able to
maneuver, or threats which might do something unexpected at any moment?
When he considers himself as a driver, does he feel competent and
focused, or uncertain and fearful? When he sees that traffic is
uncommonly slow, does he see this as an opportunity to relax, or as an
obstacle to be overcome?

David G. composed a set of Q-sort items which were patterned on the
structure of the PCT diagram, making statements about perception,
comparison, action, and disturbances (as I recall – he can remind us). I
think this list might be refined some more, with the idea of focusing on
real traits rather than reference signals or behaviors that are subject
to a lot of variability. The key is to look for things that can change
only slowly and in the manner of reorganization. It’s like going up a
level, removing attention from the flashy details and looking for the
outlines of an underlying regularity.

The reward for doing this, I think, will be to reduce the standard
deviation in the data, an effect like turning the focus knob on a
telescope and changing the fuzzy blobs into crisp points and detailed
landscapes.

Best,

Bill P.