Jeff Vancouver's new article

[From Dick Robertson, 000718.1505CDT]

Congratulations to Jeffrey Vancouver on your article, "Automaticity,
goals and environmental interactions," in July AMERICAN PSYCHOLOGIST.
Very nicely and succinctly argued. I note the two articles following
yours showed not a clue, as usual. I hope they read yours.

Best, Dick R.

[From Jeff Vancouver 000719.1200 EST]

Thanks Dick,

I actually have 2 out this month. In the July issue of Organizational
Behavior and Human Decision Processes (OBHDP) is the article in which I test
for the controlled variable on a sample of undergraduate psychology students
interacting with my simulation of a nurse manager's job. Of course, it will
not get the circulation of AP and it is much longer.

As an aside, the OBHDP article orginally had a system dynamics model, but
one of the reviewers thought it was too simple. No argument from me that it
was simple, but it seemed a good place to start. Anyway, I wrote another
article based on a system dynamics model and with a slight variation on the
OBHDP article in terms of study design. I submitted it to the Academy of
Management conference and will be presenting it there in early August (which
is why I cannot make the CSG conference - too many conferences already).
However, the article form was just rejected from the Journal of Applied
Psychology (JAP). I am now going to revise it for OBHDP.

One more tidbit. I also had an article accepted by JAP based on a paper I
gave at last year's Academy of Management conference. This one "pitted" PCT
against Bandura's Social Cognitive Theory. The idea was based on Bill P.s
comment in AP (1991) to Bandura (1989, AP). Bill argued that high
self-efficacy beliefs may lead one to believe they reached their goal
(reference point) more readily than low self-efficacy beliefs, leading to a
negative relationship between self-efficacy and performance. This is what I
found. I have not talked about this paper to you all because I did not use
the TCV. I did not think you would like it. Indeed, this effect is weak
and, I think, relatively unimportant in the grand scheme of understanding
PCT and human behavior. Yet, the audience at academy ate it up and JAP
liked it enough to publish it. They liked the idea that I showed something
opposite of what is popular. Indeed, at the conference Ed Locke was there,
as well as the lead author in a recent meta-analysis in Psychological
Bulletin that showed a strong positive relationship between self-efficacy
and work performance. I had specifically singled out the meta-analysis as
problematic because it was based on correlational studies. Anyway, both lit
into me. It was great fun! They were livid. The audience loved the show.
The article will probably come out next summer.

I am hoping that Ed Locke is at the next one. I am much more directly
questioning his views in that paper. If he is in the audience, he will
speak up and that will more likely reveal the need for what I did, which the
JAP reviewers and editor did not appear to understand.

News from Ohio University,

Jeff

···

-----Original Message-----
From: Control Systems Group Network (CSGnet)
[mailto:CSGNET@POSTOFFICE.CSO.UIUC.EDU]On Behalf Of Richard Robertson
Sent: Tuesday, July 18, 2000 4:07 PM
To: CSGNET@POSTOFFICE.CSO.UIUC.EDU
Subject: Jeff Vancouver's new article

[From Dick Robertson, 000718.1505CDT]

Congratulations to Jeffrey Vancouver on your article, "Automaticity,
goals and environmental interactions," in July AMERICAN PSYCHOLOGIST.
Very nicely and succinctly argued. I note the two articles following
yours showed not a clue, as usual. I hope they read yours.

Best, Dick R.

[From Rick Marken (2000.07.19.1500)]

Jeff Vancouver (000719.1200 EST) --

I submitted it to the Academy of Management conference
and will be presenting it there in early August (which
is why I cannot make the CSG conference - too many
conferences already).

Too bad. Picking the Academy of Management over CSG was a very
poor choice, in my opinion.

I have not talked about this paper to you all because I did
not use the TCV. I did not think you would like it.

"Like" is not the issue. If you didn't use the TCV then you
learned nothing about what people are controlling. So from a
PCT point of view you learned precisely nothing.

Indeed, this effect is weak

Correction. Less than nothing.

Yet, the audience at academy ate it up

I bet!

Anyway, both lit into me. It was great fun! They were livid.

Sounds like an argument between the three blind men to me.

Hasta luego

Rick

···

--
Richard S. Marken Phone or Fax: 310 474-0313
MindReadings.com mailto: marken@mindreadings.com
www.mindreadings.com

[From Bill Powers (2000.19.1723 MDT)]

Rick Marken (2000.07.19.1500), replying to Jeff Vancouver --

Nice hatchet job, Rick. Feel better?

Bill P.

[From Rick Marken (2000.07.19.2000]

Bill Powers (2000.19.1723 MDT)--

Nice hatchet job, Rick.

The devil (in the form of that pesky reference signal for
PCT correctness) made me do it;-)

Feel better?

Not any more.

Hasta luego

Rick

···

---
Richard S. Marken Phone or Fax: 310 474-0313
Life Learning Associates e-mail: marken@mindreadings.com
mindreadings.com

[From Bill Powers (2000.07.20.0730 MDT)]

Nice hatchet job, Rick.

The devil (in the form of that pesky reference signal for
PCT correctness) made me do it;-)

But what was PCT-incorrect in the papers Jeff wrote or presented? What he
_said_ he was trying to get across sounded pretty much OK to me, though I
can't judge until I've seen what was actually published. The Test for the
Controlled Variable doesn't appear in chapter 1 of B:CP: it doesn't even
make sense until you have some concept of how control works. When you try
to tell neophytes about control processes, is the TCV what you start with?

Mary wrote to Edwin Locke (among those neophytes who were livid about
Jeff's presentation) some years ago, and got a reply from him in which he
said he could tell that B:CP was incorrect just from the title -- he had no
need to read it. Kind of infuriating, no? Well, strictly speaking, no, but
it's the sort of statement that disturbs certain concepts concerning
scientific interactions and in my case at least seems to call for energetic
countermeasures, which I might take if I could get my hands around his
neck. Jeff might be inclined to ask exactly what it was in his papers that
you found lacking or wrong, and more to the point, what you would have said
instead to improve the papers. Could you answer him?

Best,

Bill P.

[From Bruce Gregory (2000.0720.1005)]

Rick Marken (2000.07.19.1500)

"Like" is not the issue. If you didn't use the TCV then you
learned nothing about what people are controlling. So from a
PCT point of view you learned precisely nothing.

Let's imagine a Martian observing her first baseball game. After a few innings she decides that one perception controlled by players in the field is "catching any ball that comes my way." She is unable to apply the TCV, but nevertheless, it seems to me that she has learned something. She also infers that when a ball is hit into the air, the player closest to where the ball eventually winds up runs to position herself where she can catch the ball. Again the Martian cannot use the TCV, but again I maintain she has learned something (even though it is possible that the players are controlling some other perceptual variable). It is only when she wonders how the players "figure out" where the ball will been when it reaches the ground that the TCV becomes an essential tool.

BG

[From Bruce Nevin (2000.0720.1118)]

Bruce Gregory (2000.0720.1005)--

···

At 11:05 AM 07/20/2000, Bruce Gregory wrote:

Let's imagine a Martian observing her first baseball game. [...] She ... infers that when a ball is hit into the air, the player closest to where the ball eventually winds up runs to position herself where she can catch the ball.

etc.

It is possible to note disturbances to postulated CVs without being the originator of those disturbances.

         Bruce Nevin

[From Rick Marken (2000.07.20.0830)]

Bill Powers (2000.07.20.0730 MDT)--

But what was PCT-incorrect in the papers Jeff wrote
or presented?

Nothing, as far as I know. My "hatchet job" was not aimed at
Jeff's papers but, rather, at what he said about them in his post.
Jeff already knows that I liked the paper on nurses controlling
their schedules, which is apparently the one published in the
July issue of _Organizational Behavior and Human Decision
Processes_. I also got a copy of Jeff's _American Psychologist_
comment from him yesterday and thought it was excellent as well.
I've been in touch with Jeff on CSG net and by private e-mail a
bit and I hope I've made it clear to him that I think his recent
PCT-based work has been outstanding, from my point of view.

My "hatchet job" was aimed at two comments Jeff made in his post
that just happen to disturb some variables I'm controlling with,
perhaps, higher than necessary gain. The first was Jeff's comment
about meeting attendance. To me, there is nothing more important
than CSG Meeting attendance for anyone who is seriously interested
in PCT. The fact that Jeff would chose some other meeting over
CSG was _very_ disappointing to me.

The second was Jeff's comment about the TCV. Jeff said:

I have not talked about this paper to you all because I did
not use the TCV. I did not think you would like it.

This disturbed me because it suggests that the test is just
one way to go about studying control systems; the other way
is to use the conventional IV-DV approach. My petulant reply
was meant to communicate the fact that the TCV is not just a
ritual you carry out in order to be seen as a "true" PCTer.
Rather, it is the _only_ reasonable approach to studying
the behavior of systems that are controlling their own
perceptual inputs.

The PCT beef with with the conventional IV-DV approach is not
based on its being the wrong ritual (with the TCV being the
right one) It is based (as you know) on that fact that the
functional relationships between disturbance (IV) and output
(DV) discovered using IV-DV methods (even when carried out
on individuals) reveal characteristics of the environment,
not of the organism (this, of course, is your hugely
important discovery of the _behavioral illusion_).

You say:

The Test for the Controlled Variable doesn't appear in chapter 1
of B:CP: it doesn't even make sense until you have some concept
of how control works. When you try to tell neophytes about control
processes, is the TCV what you start with?

But this was not my point. Of course I don't start with the TCV
when introducing PCT. And neither does Jeff in his excellent
essays on PCT (such as the one in the current _American
Psychologist_). I did not mean to suggest that Jeff (or anyone)
should start explaining PCT in terms of the TCV. I was reacting
(as I noted above) to the fact that Jeff had done research
based on control theory (Jeff says "based on Bill P.s comment
in AP (1991)") using conventional IV-DV methods and found (as
usual) a "weak [statistical] relationship" between IV and DV.
Actually, I wasn't even bothered by the fact that Jeff had done
that kind of research. I was just bothered by Jeff saying that
he didn't think we would "like it", as though doing research
using the TCV was some kind of odd preference of PCTers.

Anyway, I am sorry if I offended you, Jeff. I think you are
doing _very good work_ on PCT and doing a great job of
bringing PCT to the attention of conventional psychologists.
It's really because of that that I am so disappointed
that you are not coming to the CSG meeting.

Hasta luego

Rick

···

--
Richard S. Marken Phone or Fax: 310 474-0313
MindReadings.com mailto: marken@mindreadings.com
www.mindreadings.com

[From Bruce Gregory (2000.0720.1221)]

Bruce Nevin (2000.0720.1118)

It is possible to note disturbances to postulated CVs without being the
originator of those disturbances.

Indeed. Bit I thought the purpose of TCV was to eliminate plausible but incorrect conjectures about the CV. Was I misinformed?

BG

[from Jeff Vancouver 000720.1210 EST]

I suppose I should weigh in on this as it is me that is being talked about
so much.

First, let me say that whether negative or positive, I control for attention
and I am loving this. Because I also control for praise, I like that stuff
doubly.

Second, I was not offended by Rick's comment. I did not take it too
seriously. I now believe Rick respects me (a perception for which I have a
high gain), but I generally use my mental model of Rick, and not his actual
statements, to create my perception of Rick's respect for me. Otherwise, I
would have to deal with disturbances to that system all the time.

I also know (my mental model spits out) that he thinks the TCV is the _only_
method for studying control systems (like humans) and on that score we
respectfully disagree (although Rick's conception of what constitutes the
TCV is broader than mine, so we may not be as far apart on this issue as it
might appear). I also predicted that is why he would not like my paper. I
do not seems to have erred there. However, I should probably make the paper
available. That way we can get beyond prediction to information. Toward
that end, I have put the paper and its figures on my web site (though not
generally accessible). You can reach it at:
http://oak.cats.ohiou.edu/~vancouve/efficacy.htm. It is not downloadable,
but you can read and print it from your browser. Some of the figures need
to be clicked on to see/print.

Regarding conferences. I have no doubt that CSG would be a good experience
for me. I am less sure that my presence would have much useful impact on
you all. Instead, the people I want to influence are not you as much as it
is my colleagues in psychology and management. I can better do this at
their conferences than yours. If I had the resources, I would do both. I
also consider Rick's displeasure the highest form of flattery. Maybe next
year.

This logic also relates to the significance of the unseen paper we have been
talking about. I may have misspoke when I said it was relative unimportant
in terms of understanding humans, although the effect (if there is one)
would be more indirect. But if it gets 2 research psychologists to take
more seriously PCT and start doing PCT research, then the time my students
and I spend on it would be more worthwhile to PCT then our simply doing more
PCT work that has little impact on the greater scientific community. As I
have said over and over, the game is getting the scientific community to buy
the theory. Without buy in, it is merely a footnote. The point of the
paper I am presenting at Academy (and is in OBHDP) is the next step - to
show how to do PCT research. (And yes, Rick, the OBHDP is the paper you saw,
without the system dynamics model.) It is a harder sell because they are
not understanding why they should do PCT research; what it gets them. To do
that, I have to think like they do and act in ways they can understand.
Like the sane in an insanity ward.

Later

Jeff

P.S. Rick, I would appreciate you adding my home page to the web site. That
is: http://oak.cats.ohiou.edu/~vancouve/. Thanks. Please do not add the
paper's address. That would violate APA's copyright rules.

[From Bruce Nevin (2000.0720.1403)]

From Bruce Gregory (2000.0720.1221)--

Bruce Nevin (2000.0720.1118)

>It is possible to note disturbances to postulated CVs without being the
>originator of those disturbances.

Indeed. But I thought the purpose of TCV was to eliminate plausible but incorrect conjectures about the CV.

Of course I agree. The additional point here is that it's possible to test for a controlled variable when you are not controlling the source of the disturbance, but only observing it. Based on her observations, your Martian could validly conclude that one CV is "batted ball reaching the ground". Gravity is the primary disturbance as ball-players try to prevent fly balls from reaching the ground. They are also controlling many other variables, but that's always the case. For some CVs it may not be possible to measure a disturbance without being the originator of it. For others, such as the trajectory of projectiles, there are few if any alternative explanations when you see people repeatedly move so as to intercept the trajectory.

In 1971 at a demonstration in DC I saw a stone or brick flying from somewhere to my right over people's heads across a considerable distance and arcing down pretty accurately toward the chief of police, maybe 50 yards away from me. A tall, dark-haired young man who appeared to me to be one of the demonstrators ran across and threw himself into the air with his arms outstretched. My guess is he intended to block it with his body. Maybe he had put some thought into shielding others from harm with his body, with noble pacifist intent. His left cheek intercepted the trajectory nicely, and he completed his own trajectory to the ground crumpling over and clutching his face. I don't know what became of him, maybe they arrested him. As it happened, the action was caught on camera and appeared on the evening news, so I saw it a second time in a bar that evening. Whatever his purposes were at higher levels, can there be any doubt that this young man was controlling the trajectory of that projectile to stop it? Of course we could be more sure if we see him doing it again and again. (Hopefully, he would gain in skill.) Your Martian has that benefit watching baseball games.

         Bruce Nevin

···

At 01:21 PM 07/20/2000, Bruce Gregory wrote:

[From Rick Marken (2000.07.20.1250)]

Jeff Vancouver (000720.1210 EST) --

I was not offended by Rick's comment.

Thanks, Jeff. No offense was intended.

I also know (my mental model spits out) that he thinks
the TCV is the _only_ method for studying control systems

It's the only _valid_ method for studying control systems,
yes. And there's no need to rely on your mental model of me
to know that I think this is true. I've said it straight out
in many forums (including this one earlier today and, of
course, in the "Dancer..." paper to which you refer in the
_American Psychologist_ article). As an experimental
psychologist, I believe the fact that the TCV is the only
valid approach to the study of living control systems is
the most important lesson PCT has to offer the psychological
research community.

and on that score we respectfully disagree

Yes. But I still don't understand why (we disagree, not why
we are respectful;-). My reason for believing that the TCV
is the only valid method for studying living control
systems is based on an understanding of the behavioral
illusion. So my reason is described (and illustrated with
a demo) at:

http://home.earthlink.net/~rmarken/ControlDemo/Illusion.html

Why do you think the TCV is _not_ the only valid method for
studying living control systems? What other valid methods
are there? Why are they valid?

P.S. Rick, I would appreciate you adding my home page to the
[CSG]web site.

With pleasure. It will be there later today.

Best

Rick

···

--
Richard S. Marken Phone or Fax: 310 474-0313
MindReadings.com mailto: marken@mindreadings.com
www.mindreadings.com

[From Jeff Vancouver 000720.1615 EST]

Rick Marken (2000.07.20.1250)]

Why do you think the TCV is _not_ the only valid method for
studying living control systems? What other valid methods
are there? Why are they valid?

Rather than answer that question, maybe you could read the paper and tell me
if you find anything worthwhile in the methods. As I mentioned, your
definition of TCV is sometimes broader than mine (although, it is sometimes
narrower I think too).

One caution, your use of the term "valid" is going to be an issue. By valid
I assume you mean a method that reduces alternative explanations. If that
is true, valid is a matter of degree. It may be that in some contexts (to
answer certain questions) a method will produce only one interpretation, but
that is rare (e.g., the TCV is open to the possibility that you are blocking
more than you think you are blocking), and the same method may not produce a
single interpretation when used in a different context (i.e., interpreting
TCV results is very difficult if reorganization is occuring, if I read
Runkel right, although I think that whether Runkel said it or not). I
attempt to employ methods that are relatively free from alternative
interpretations, while being practically employable. Conflicting goals can
be a pain.

That being said, I am grateful to you for pushing me to use the Test. It is
a very nice (likely best) method in certain contexts.

Jeff

[From Rick Marken (2000.07.20.1450)]

Jeff, your site is now proudly listed on the CSG related
sites page at http://www.ed.uiuc.edu/csg/sites.html

Jeff Vancouver (000720.1615 EST) --

maybe you could read the paper and tell me if you find anything
worthwhile in the methods.

I will definitely try to find time. I probably won't get to it
until after the meeting, though; it's a pretty long paper.

One caution, your use of the term "valid" is going to be an issue.
By valid I assume you mean a method that reduces alternative
explanations.

I mean only one thing by "valid" in this context: valid research
methods are those that produce results that are not subject to
the _behavioral illusion_. The behavioral illusion occurs when
the nature of observed relationships between variables is seen
as being determined by characteristics of the organism when,
in fact, they are determined by characteristics of the
environment in which the organism is situated.

That being said, I am grateful to you for pushing me to use
the Test.

And I am grateful to you for using it and, more important,
for successfully publishing research based on the TCV.

Best

Rick

···

--
Richard S. Marken Phone or Fax: 310 474-0313
MindReadings.com mailto: marken@mindreadings.com
www.mindreadings.com

[From Bill Powers (2000.07.20.2027 MDTO)]

Rick Marken (2000.07.20.0830)--

Anyway, I am sorry if I offended you, Jeff. I think you are
doing _very good work_ on PCT and doing a great job of
bringing PCT to the attention of conventional psychologists.
It's really because of that that I am so disappointed
that you are not coming to the CSG meeting.

That bit, to my mind, is better than any amount of insisting on PCT
correctness. I appreciate your pushing for keeping on track; but there are
various ways of doing this, some of which are less alienating than others.

That's the real Rick talking.

Best,

Bill P.