# Learning to control

[From Rick Marken (960226.2130)]

Me:

could you present the kind of learning data that you feel HPCT should
be able to predict?

Shannon Williams (960226.15:30) --

If a person is presented with a problem (he is having difficulty making
some perception match its reference), and I present some partial
solution to the problem, I want to be able to predict if my partial
solution gives him enough information to solve his problem.

OK. That helps a lot.

The kind of problems you are talking about clearly involve control. Take
a water jar problem, for example. You have 3 jars and are told that they
can hold 8, 5 and 3 units of water, respectively. You are also told that
the first jar is full while the others are empty; so the contents of jars
is 8,0,0, respectively. The goal is to divide the water in the first jar
equally between the first and second jar so that you end up with the
contents of the jars being 4,4,0, respectively. You are to reach this goal
by pouring water from one jar to another.

Control is involved because you want to bring the current perceptual state
of the contents of the jars (8,0,0) to the reference state (4,4,0). You do
this by producing intermediate perceptual states that follow the constraints
of the problem. For example, you might start by pouring from jar 1 into
jar 2, changing the perceived contents of the jars from (8,0,0) to (3,5,0).
The next move might be to pour from jar 2 into jar 3, resulting in (3,2,3).

A person who knows how to solve water jar problems will change the initial
state of such a problem into the goal (reference) state in something
close to the minimum number of moves (in this case, about 7). A person
who does not know how to solve such a problem may make hundreds of
moves before hitting on the solution; I saw one person spend nearly an
hour making moves and never getting the solution.

A water jar problem like this involves both some ability to control (the
ability to correctly change from one problem state to another, for example;
just knowing what state results when you are in state (6,2,0) and you
pour from jar 1 into jar 2 involves the ability to do some complex
controlling--in imagination and reality) and learning. You can't tell
whether any particular action (move) is part of a control or a learning
process just by looking at it. For example, is the person who moves the
problem from state (6,2,0) to state (6,0,2) cleverly controlling for
having 1 remaining in jar 1 that can then be added to 3 from jar 3; or
is the person making moves randomly to see what will happen next
(trying to learn what is important to control about the problem)?

I can't see how you could know this (whether the problem solver is
controlling or learning to control) without doing something like the
Test to determine whether the person intended to produce a particular
state (6,0,2) and, if so, to produce it as part of a process of controlling
for a more complex variable (producing 1 to add to 3) or whether it was
produced as part of a process of exploration, where the result might have
been intended, but not as part of a process of controlling for anything
other than producing the final goal state.

To the extent that the problem solver is in control of the problem --
that is, the person knows which perceptions to control and how to control
them -- then you can predict each move and predict the effect of
disturbances to any state of the problem. To the extent, however, that the
problem solver is not in control of the problem -- that is, the person has
little idea what perceptions to control or how to control them -- then it
will be difficult to predict anything other than the fact that moves will
be made nearly randomly; nearly, because the learning process is assumed
to be a _biased_ random walk; control processes -- and the moves that support
them -- that produce an approach to the ultimate goal (problem solution)
will persist longer than those that produce an apparent move away from
the ultimate goal (indeed, there is conventional problem solving data that
suggests that this is the case; a means-ends strategy, which makes the
current state of the problem look more like the goal state, seems to persist
-- and interfere with the ability to find the correct control strategy --
throughout problem solving).

I have always wanted to apply control theory to this kind of problem
solving. The first thing I would do is allow a subject to become skillful at
solving a certain class of problems, like the water jar problem. A person
is skilled when they can solve any version of the same problem (in this
case, any set of 3 jar capacities) in the minimum number of moves. Once
the person is "in control" of the problem, I would start testing to see
what is being controlled; I would try to determine what this (and several
other) skilled problem solvers control in order to solve the problem.

Once I knew what variables were controlled by the skilled water jar
problem solvers, I would start studying how novices learn to solve
these problems. I would do this by testing for control of the variables
that were controlled by the experts and I would do this throughout the
learning process. My goal would be to determine whether people learn to
control a particular "expert" variable suddenly (all or none) or gradually;
whether control of a particular "expert" variable is tried, abandoned and
re-tried; whether several "expert" variables must be tried all at once in
order to gain control of the problem.

These are really interesting questions (to me) but I think it would be
hard to answer them before I knew what variables are controlled by skilled
problem solvers.

Best

Rick

[From Fred Nickols (2002.03.23.1050)] =

Rick Marken (2002.03.22.1320)]

> Fred Nickols (2002.03.22.1200) --
>
> If the fidelity and transfer issues ring a bell, let me know and I'll ask
> around on a couple of the training lists to see if anyone can point to
> specific studies.

Thanks. That would be _great_. I think it's more like a fidelity issue, though
that's not quite it either. The question is whether _actual_ skill (in
terms of
control gain, say) is higher after training in a "high disturbance" rather
than
a "low disturbance" environment. The ambient disturbance level of the actual
environment in which the skill is exercised may be low. But I'm wondering
whether when you train to have some skill whether you end up more skilled (in
the control loop gain sense) when you train in a high rather than a low
disturbance environment.

Hmm. Looks like we might be mixing commonplace with technical meanings of
"disturbances." Let me try an example. If I have it right, I'll use it to
post the query to the other lists.

Suppose I'm training a manager to process his or her in-basket. As far as
fidelity is concerned, if the manager has a private, quiet office, a
faithful training environment will provide that kind of ambience. On the
other hand, if the manager works in an open cubicle in a noisy environment,
subject to all manner of distractions and interruptions, I need to
replicate that environment if the training is to be faithful (i.e., marked
by fidelity).

Next, regardless of their actual working environment, suppose I train all
managers in a noise-filled, interruption-laden environment. The question
you seem to be asking is if they actually work in quiet environments, will
their in-basket skills be more developed than if we had trained them in a
quiet setting?

This seems to me to be a commonplace use of the term "disturbances" more so
than in the usual PCT sense. In the PCT sense, I thought "disturbance"
referred to an externally driven change in the level or perception of some
controlled variable. It seems to me that a noisy environment might or
might not upset any controlled variables -- depending on what we would
ordinarily call the ability of the performer to focus and not get
distracted. Could someone improve upon that ability via training? Probably.

Anyway, do I understand what you're after?

Now, for a relevant postscript.

One of the things I enjoyed about being a consultant was that it afforded
me opportunities to go places and do things that would otherwise have been
denied me -- or at least never have come my way.

One of these was the opportunity to work with Tom Peters, the famed
management guru.

In a session I set up for some Bell System (pre-divestiture) managers, Tom
told the story of an experiment in which two groups of clerks were
subjected to a continuous barrage of background noise, interruptions,
etc. One group had a "shut off the noise" button (i.e., if they pressed
the button, all the background noise would cease). Tom, in his usual
diplomatic fashion, pointed out that in a science (psychology) noted for
its "non-findings" there were no surprises: the group with the
shut-off-the-noise button produced twice as much with half the errors. He
went on to add that the truly amazing finding, a finding that has since
been replicated many times, is that NOT ONCE did the group with the
shut-off-the-noise button ever use it. Tom's point was that the perception
of even a modicum of control did wonders for performance and productivity.

Somehow that strikes me as relevant to your query. Whatcha think?

Regards,

Fred Nickols
740.397.2363
nickols@att.net
"Assistance at A Distance"
http://home.att.net/~nickols/articles.htm

[From Rick Marken (2002.03.25.1400)]

Fred Nickols (2002.03.23.1050)--

The question
you seem to be asking is if they actually work in quiet environments, will
their in-basket skills be more developed than if we had trained them in a
quiet setting?

No. I am asking whether the gain of the "in-basket" control system is greater
after training with large rather than moderate disturbances to the perception
controlled by the in-basket control system.

In the PCT sense, I thought "disturbance"
referred to an externally driven change in the level or perception of some
controlled variable.

Not quite. A disturbance is a variable that influences the physical state of
affairs perceived and controlled by the control system. Whether or not a
disturbance actually produces a change in the state of the controlled variable
depends on how well the controlled variable is being controlled.

Anyway, do I understand what you're after?

It doesn't seem so. But thanks for trying.

Somehow that strikes me as relevant to your query. Whatcha think?

It doesn't seem relevant to me.

Best

Rick

···

--
Richard S. Marken, Ph.D.
The RAND Corporation
PO Box 2138
1700 Main Street
Santa Monica, CA 90407-2138
Tel: 310-393-0411 x7971
Fax: 310-451-7018
E-mail: rmarken@rand.org

[From Rick Marken (2002.03.26.1110)]

Fred Nickols (2002.03.26.1146)--

Okay, I'll try again. If I'm running an in-basket exercise and there are
two basic approaches: (1) a sort of routine, mundane, run-of-the-mill
approach and (2) a Holy Cow! I'm swamped! approach, you're asking if
learning to handle the "I'm swamped" in basket exercise would leave the
trainee with a better (higher gain?) and thus better able to handle the
routine work.

Yes. This is more like it. Assuming that the number of items in the in basket is
what is being controlled, then the question is whether control is better after
training with a a large (but not overwhelming) disturbance compared to training
with a more moderate disturbance (where the size of the disturbance would be the
amplitude of variations in the rate at which items are added to the in basket).

Wait a minute. Does a disturbance influence the "physical state of affairs
perceived and controlled by the control system"

Of course. Disturbances are physical variables that have effects on other physical
variables. Check any PCT diagram.

Best regards

Rick

···

--
Richard S. Marken, Ph.D.
The RAND Corporation
PO Box 2138
1700 Main Street
Santa Monica, CA 90407-2138
Tel: 310-393-0411 x7971
Fax: 310-451-7018
E-mail: rmarken@rand.org

[From Fred Nickols (2002.03.26.1146)] --

Rick Marken (2002.03.25.1400)

Fred Nickols (2002.03.23.1050)--

> The question
> you seem to be asking is if they actually work in quiet environments, will
> their in-basket skills be more developed than if we had trained them in a
> quiet setting?

No. I am asking whether the gain of the "in-basket" control system is greater
after training with large rather than moderate disturbances to the perception
controlled by the in-basket control system.

Okay, I'll try again. If I'm running an in-basket exercise and there are
two basic approaches: (1) a sort of routine, mundane, run-of-the-mill
approach and (2) a Holy Cow! I'm swamped! approach, you're asking if
learning to handle the "I'm swamped" in basket exercise would leave the
trainee with a better (higher gain?) and thus better able to handle the
routine work.

In real life, for example, I might be under some pressure at all times to
process whatever's in my in basket. Usually, I process whatever's there
and thus keep my level of perceived stress at some acceptable
level. However, on occasion, my in-basket is suddenly filled to
overflowing with high-priority matters and my perceived stress level is now
outside my tolerance limits. I have a big disturbance and, probably, an
error. So, perhaps I change my schedule for the day, buckle down and work
my way through whatever's in the in basket. Perhaps I negotiate a
different due date for some of what's there. Perhaps I delegate some of
it. Point is, I act to bring the perceived stress level back inside
acceptable limits.

Is this closer to what you're talking about?

> In the PCT sense, I thought "disturbance"
> referred to an externally driven change in the level or perception of some
> controlled variable.

Not quite. A disturbance is a variable that influences the physical state of
affairs perceived and controlled by the control system. Whether or not a
disturbance actually produces a change in the state of the controlled variable
depends on how well the controlled variable is being controlled.

Wait a minute. Does a disturbance influence the "physical state of affairs
perceived and controlled by the control system" or is that a hypothesis or
surmise on our part given that all we know of those affairs consists of our
perceptions? And, under PCT, I thought we controlled our perceptions of
controlled variables, not the variables themselves. So, theoretically at
least, it seems to me that a disturbance could in fact alter the "variable"
and that would not be a problem as long as the perception of that variable
remained within limits. (Why do I have the sinking feeling that someone is
about to give me a quick lesson in PCT 101?)

> Anyway, do I understand what you're after?

It doesn't seem so. But thanks for trying.

No problem. As you can see, I've just taken another crack at it.

> Somehow that strikes me as relevant to your query. Whatcha think?

It doesn't seem relevant to me.

Now that I think I understand your query better, it doesn't seem relevant