[From Jeff Vancouver 950913.0950]
Hi everyone! It has been a long time, but thought I ought to resurface
for a comment and question or two.
First, I have been lurking, but have caught very little of the
conversation. I have been trying to get some work done. Not that reading
and interacting on this net does not improve my work - I have noticed much
of my interactions on this net creeping into my work - but that I felt
more of a need to do then to think and interact. I am sure what I do
suffers as a result, but I must do so that others can evaluate me. That
evaluation has changed from going for tenure to looking for a job. Don't
weep for me, it is not that I did not get tenure, I am not going to try,
so I need to find a job. If anyone knows of a good job (preferably in a
psychology department, but any social science department or program that
is in need of a psychologist may be of interest - may be of even more
interest than a psychology department). But I digress about me.
Let me give a brief description of what I have been doing (I guess this is
more digression about me). The principle activity which should interest
people on the net is that I have creating a simulation of a manager's job
(this was the topic of a NSF proposal I posted on the net months ago for
feedback - I asked for constructive feedback and got none except a request
for clarification of what constructive feedback meant. Either the post
never made it to the net, or my definition of constructive feedback was
too narrow.) Anyway, I designed the simulation so that I can place people
in a setting with multiple goals from multiple constituencies with
multiple sources of information about goal progress. The simulation
allows me to control and monitor actions and information more closely then
a field study would allow. For example, I can apply the test to guessed
controlled variables on the part of my participants. This is exactly one
of the types of studies I plan to do.
So my first question is for other ideas from the net. Rick (950912.0930 &
1830) claims that a scientist ought to being trying to disprove a theory
with empirical data, not trying to prove it. He also claimed it was easy.
I am puzzled by this. Not the abstract idea - his view of science is
widely held. Rather, I am interested in the particulars of applying that
method to PCT. I am not sure how to set up empirical tests that have the
potential to disconfirm PCT. Perhaps my problem is that I am working and
thinking around the program level in the hierarchy. I think it is around
this level that much of the processes that interests me and others in my
field (industrial/organization psychology) occur. Based on past
communications with this net, I am guessing that the PCT understanding of
those levels is considered not sufficiently thoughtout (modelled) that it
is ready for disconfirmation.
Given the difficulty of disconfirming tests (more on this later in the
post), I am also interesting in tests which pit PCT against another
theory. This is somewhat of a change in my past position. I have been
focusing on the commonality among psychological and systems theories
rather than the differences. This viewpoint met with much resistance on
the net, so a competitive stand may be more welcomed, although not by
everyone. Indeed, I am not sure how much knowledge of other theories
exist on the net. As Rick has pointed out, one must have a decent
representation of a theoritician's model to construct empirical tests
using it. Most of those on the net are attempting to construct a decent
representation of PCT, not other theories. Indeed, if one is to pit PCT
against other theories or against "reality", PCT must be well
represented. But, I put the question out there anyway.
My change in heart is not complete. I still suspect that most currently
popular theories in psychology are actually compatable or encompassed in
PCT. That whatever differences exist are mostly due to poorly specificied
or unspecified processes, constructs, or mechanisms in the non-PCT
theories. So "pitting" is not always possible. I think the goal on my
part is to show that PCT can produce better predictions, lead to
interventions with better chances of success, or make predictions that no
other theory would make. (I think I might have some data along those
lines right now, but that is another post).
In an attempt to forestall anticipated perceptions of Rick's reaction to
this post - perceptions that will lead to errors for me - I will make a
few more points. First, I think Rick has given up on other theories
altogether and will argue that I am wasting my time. My first counter
argument to that idea has not changed. I am interested in getting other
scientists to see the merit of PCT (or PCT-like thinking) 1) because
consensus among the scientific community is mostly what science is about,
and 2) because those scientists are the ones who will hire me and I have
to convince them I have something to contribute to their world. Pitting
theories against each other using consensually accepted methods is one way
to convince them. Of course, it is in my interest that the tests results
show PCT coming out ahead, as that is the theory I am promoting. In other
words, if I use the best (i.e., most consensually excepted) methods to
show that some stupid theory (in the other scientists minds) is in fact,
not better than the popular theory, I have not done myself any service
(although I may have done people on this net a service). Of course, I
serious doubt the study would go again PCT (ah, but that is ascientific,
excuse me.) By the way, I think that modeling studies are consensually
accepted, though it is a much tougher sell than experimental studies.
My second counter argument is the one for which I cannot predict a
reaction from those on the net, so it most interests me what you all will
think. Specifically, the reason I think PCT will not fail an empirical
test is because I cannot imagine an empirical test that would disconfirm
it. This is why I am requesting help on that. But, let me explain
further. First, I accept Rick's notion that the control process is fact.
So I guess disconfirmation is that we are not controlling perceptions.
But, we cannot measure perceptions directly, so we must make guesses. We
could always guess wrong, and this excuss can always be used to counter a
"disconfirming" experiment. Instead, I am of the mind that PCT is much
like evolutionary theory. It is not really disconfirmable, yet it is a
very useful way of looking at species and how they come about and
reproduce. This is not to say that some of the details of PCT and
evolutionary theory are not subject to disconfirmation (indeed, I think
some of the details in PCT are improperly emphasized), but that the
general model is correct (I think this is the distinction betweeen PCT and
HPCT that is often made on the net). Back to evolutionary theory for a
minute, can anyone think of a way to disprove it? You would need to find
a species that came about without having evolved - a very difficult task
indeed.
The argument over the disconfirmation of PCT is critical to my continued
logic arguments with colleaques about the merit of PCT. One of the more
popular counter-arguments of late is the idea that a meta-theory of PCT's
nature is useless because is says nothing by saying everything. (We have
visited this argument before, it is E. Locke's, but it has apparently
found a receptive audience in other scholars). So, I am interested in
either 1) showing or describing tests that can be used to disconfirm it,
or 2) showing or describing tests that show its usefulness over other
theories. As I said, I do not think the former (1) is possible, but I
think that enough of the later (2s) would be sufficient for the former
(1), and given my current beliefs, that is where I am focusing my
energies.
Enough said for now. I would appreciate any thoughts. I would
appreciate at least a confirmation that this was received, as I am not
sure my posts are getting through.
Later
Jeff
···
_________________________________________________________________________
Jeffrey B. Vancouver
Assistant Professor Phone: (212)998-7816
Department of Psychology Fax: (212)995-4018
New York University e-mail: jeffv@xp.psych.nyu.edu
6 Washington Pl., Rm 572
New York, NY 10003