What psycholgists say

[From Rick Marken (950116.0845)]

Bill Powers (950115.0600 MST) --

When we come up with demonstrations that positively prove that behavior is
the control of perceptions, we may think that gaining support is simply a
matter of showing the evidence and letting it speak for itself ... That is
not what will happen. What will happen is that some people will be
profoundly shaken, even devastated, by these demonstrations ..

I, personally, have never seen this response and I did these demonstrations
for a number of colleagues when I was teaching. The only response I ever saw
was what you describe next:

What about the people who look at the demonstrations and simply refuse to
believe them? These, I am afraid, will be in the majority. Their reactions
will range from indifference to outrage.

Indifference and outrage was what I saw, with the former _far_ more common
than the latter.

So, Bruce, when you start handing out your data to other psychologists to
analyze, I hope you will be prepared for the consequences.

I am really interested in hearing about the kind of responses Bruce gets from
his colleagues. The experience was quite an eye-opener for me when I did it
(on a very small scale -- I tested maybe five or six people). I eventually
stopped doing these tests because I felt like the guy in church demonstrating
that there is no god and (believe it or not) I don't like to step on anyone's
fantasies: or maybe it's that I don't like to see grown men cry;-).

Best

Rick

Tom Bourbon [950118.1709]

SLowly catching up on six days of mail.

[From Rick Marken (950116.0845)]

Bill Powers (950115.0600 MST) --

Bill:

When we come up with demonstrations that positively prove that behavior is
the control of perceptions, we may think that gaining support is simply a
matter of showing the evidence and letting it speak for itself ... That is
not what will happen. What will happen is that some people will be
profoundly shaken, even devastated, by these demonstrations ..

Rick:

I, personally, have never seen this response and I did these demonstrations
for a number of colleagues when I was teaching.

While I was still teaching I saw a couple of people who were shaken, but
none were devastated. The most interesting profound shaker was a young new
faculty member. He was *excellent*, by traditional standards: good
traditional institutions, good traditional courses, lots of good traditional
publications while he was still a student, and so on. Of course, when I
saw how good he was by those standards, I thought, "what a waste!" Among
others, he taught some of our courses on statistics, experimental design,
and experimental psychology.

One of my thesis students (Mark Lazare, for those who have been at CSG
meetings in recent years) asked the faculty member (let's call him Fred -
not his real name) to serve on his thesis commitee. Mark knew Fred would be
a tough customer: Fred had never talked with me about PCT, but he had heard
about it through students and he lectured in some of his classes against
what he thought it was. Mark's thesis involved pairs of people cooperating
during a variation on the theme of tracking. Fred came to the lab one day,
to see the task and to get a better sense of how we did the analyses and the
modeling.

We started Fred out with the simplest example of compensatory tracking,
using a mouse as the control device. We showed him the different
variables (mouse position, cursor position, dirturbance value, etc.) and the
descriptive statistics for them (means of differences, SDs, variances,
correlations, etc). Fred noticed rather quickly that correlations he
thought should be high were near zero; ones he thought should be low, or
hadn't thought about at all, were nearly perfect; and several that were
"highly statistically significant" were obviously unimportant for
understanding what happened during the run. Fred sat quietly for a few
moments, looked me straight in the eyes, and said, "You know, of course,
what this implies about the experimental and statistical methods we use in
psychology."

I'm sure he meant for me to say, "Oh, good heavens! I hadn't thought of
that! Of course, we must be wrong!" Instead, I savored the moment, looked
him in the eyes, and said softly, "Yes."

Eventually we resumed the demonstration, arriving at the point where,
after one run, an elemental PCT model predicted his performance on the next
run, five minutes later. He was a novice tracker, but the correlation
between predicted and actual mouse positions was in the mid .90s. I said,
"That's pretty bad. After a few more runs, it will probably be over .99."
We did a few more runs, and it was. After a thoughtful pause, Fred looked
me in the eyes again and said, "You know, of course, what this implies about
the last 300 years of ideas about behavior." Savor -- savor -- "Yes."

Fred was not happy, but he was not devastated. He was obviously intrigued
by the fact that even the briefest single movement of the mouse often
resulted in a "statisticallty significant" correlation between mouse
positions and disturbance values. There are 1800 values of each variable,
so even a .05 correlation, or less -- I forget now -- is significant.
That fact bugged the devil out of him -- he was a very bright, very
well-trained, young experimentalist. I fixed him up with a little program
that would let him play with the mouse-disturbance relationship to his
heart's content. During the next week, he literally spent hours sitting in
his office, often in the dark, for reasons I do not know, trying to find
ways to move the mouse so that the "significant," but obviously meaningless,
correlations would go away.

Eventually Fred stopped playing with the mouse and went on to become a
deeply emotional opponent of every student who mentioned PCT, or who
proposed a PCT thesis. In that regard, he had joined the phalanx of
opposition in my department. During the next couple of years, Fred never
spoke to me again about PCT, or about experimental methods, or about the
past 300 years of ideas about behavior.

Be prepared, Bruce.

Later,

Tom