Where have I read this before?

@*@*@*@*@*@* FROM CHUCK TUCKER 940105 *@*@*@*@*@*@

         I was reviewing the posts from November (395 pages) and
         December (423 pages) of 1994 in preparation for writing
         a few comments on the net only to discovered that
         several issues are being discussed for the third time.
         The discussion of statistics and IV-->DV research is
         back again with about the same results (must be some
         sameness in the CV): aggregate statistical analysis is
         inappropriate for the analysis of a single organism's
         activities. If anyone doubts that then have them give
         an analysis showing that not to be the case and we can
         examine it carefully otherwise one can read (and accept)
         Runkel's statements in this book.

         We are continuing to discuss the question: Does the
         external environment (in what ever form) cause the
         behavior of a single organism? I noted on 940101
         Morse Peckham's cute little statement (Morse died
         last year):"The assumption that a stimulus can be
         identified and that its response can be determined in
         a neutral laboratory has simply led to an immense
         proliferation of experiments on what is apparently
         going to be an endless series of configurations.
         Little that is not trivial has been discovered, and
         little of significance has been done with the
         nontrivial discoveries. But even this distinction
         is evaluative. S--> R behaviorism has discovered little
         more that that people do respond - to something or
         other (1979:30)." Does anyone have any evidence
         that "a stimulus" causes "a response"? If so, put it
         on the net so we can examine it.

         We continually read that the writers on the net do
         not understand on another because 'words' are not
         specified or defined in a similar manner (see Bill's
         umteenth statement on this issue "Words and Models"
         written as Santa was breaking into his house). If
         we suspect that this is a problem then why do we
         continue to ignore it. We all recognize that PCT
         defines these terms different from the dictionary
         (unless your dictionary is in the back of B:TCP!)
         and from all other social and behavioral scholars:
         behavior, stimulus, control, open loop, closed loop,
         action, feedback, input, output, gain, error, positive,
         negative, system, social control (sic), environment,
         model, disturbance, reorganization, change and on
         and on and on. So why not just begin a post by
         stating "What I mean by X is..." so some there might
         be more "understanding" and less "ignorance."

         Finally, it should be recognized that not one set of
         data collected using "PCT approved procedures" has
         been put on the net or published in the last two
         years. What is constantly referred to are a few
         NON-REPLICATED (and perhaps non-reproducible?) studies
         done by Marken (see his Reader), several done by
         Bourbon and his colleagues (I have not seen Tom's
         paper given at the conference in France so I don't
         know what is new) and several papers by Bill. Most
         of these involve one type of "tracking experiment":
         compensatory tracking. No one has presented data
         on "tracking" using such variable as "sound,"
         "shape," or "orientation." No one has presented
         any data that replicates any of the earlier studies.
         I would consider this to be a remarkable achievement
         for PCT and for us who spend most of our time teaching,
         "researching," and trying to convince other that PCT
         is the most useful and empirically supported model
         for understanding the "doings" of a human being. We
         are able to convince some people (in my case some
         coercion is involved - required course and grades)
         that PCT is it!

         I call for more research using PCT and the posting of
         it on the net and less of these "just words, words,
         words." But to quote Miller "Of course, I could be
         wrong." I will post some data from a "study" with
         14 students using DEMO1 in a moment. Hope it will
         be useful to y'all.

         Regards, Chuck

Tom Bourbon [950105.1341]

          @*@*@*@*@*@* FROM CHUCK TUCKER 940105 *@*@*@*@*@*@

        I was reviewing the posts from November (395 pages) and
        December (423 pages) of 1994 in preparation for writing
        a few comments on the net only to discovered that
        several issues are being discussed for the third time.
        The discussion of statistics and IV-->DV research is
        back again with about the same results (must be some
        sameness in the CV):

Yep. We never change that good old reference setting for perceptions of
people using the right tools for the job at hand, which is to identify and
explain control. Every time a new soul joins the net and contends
that traditional IV-DV methods are adequate to the task, we go to work on our
error signals. It's a lot like teaching the same subject, semester after
semester, isn't it, Chuck? Same room, same subject, but the faces looking
back at you aren't the same. That's what it is all about.

        aggregate statistical analysis is
        inappropriate for the analysis of a single organism's
        activities. If anyone doubts that then have them give
        an analysis showing that not to be the case and we can
        examine it carefully otherwise one can read (and accept)
        Runkel's statements in this book.

Yep. Every time around, it's pretty much like the way you describe it.

        We are continuing to discuss the question: Does the
        external environment (in what ever form) cause the
        behavior of a single organism? I noted on 940101
        Morse Peckham's cute little statement (Morse died
        last year):"The assumption that a stimulus can be
        identified and that its response can be determined in
        a neutral laboratory has simply led to an immense
        proliferation of experiments on what is apparently
        going to be an endless series of configurations.
        Little that is not trivial has been discovered, and
        little of significance has been done with the
        nontrivial discoveries. But even this distinction
        is evaluative. S--> R behaviorism has discovered little
        more that that people do respond - to something or
        other (1979:30)." Does anyone have any evidence
        that "a stimulus" causes "a response"? If so, put it
        on the net so we can examine it.

I second your challenge. I liked the quote from Peckham, down to the place
where he said, "S--> R behaviorism has discovered little more that that
people do respond - to something or other." Up to there, he seemed right on
track. That's where his own closet variety of S-->R showed through.

        We continually read that the writers on the net do
        not understand on another because 'words' are not
        specified or defined in a similar manner (see Bill's
        umteenth statement on this issue "Words and Models"
        written as Santa was breaking into his house). If
        we suspect that this is a problem then why do we
        continue to ignore it.

Yes, but are there any good substitutes for words? As a way around this
problem, we urge people to do some modeling, or to run some of our models,
but in most cases that doesn't seem to help very much. Either they don't
follow our suggestions, or, if they do, they dismiss the results as not
relevant to the *real* questions they *really* care about.

        Finally, it should be recognized that not one set of
        data collected using "PCT approved procedures" has
        been put on the net or published in the last two
        years.

Hey. That hurts. Go gripe at our reviewers and editors, not at us.

        What is constantly referred to are a few
        NON-REPLICATED (and perhaps non-reproducible?) studies
        done by Marken (see his Reader), several done by
        Bourbon and his colleagues (I have not seen Tom's
        paper given at the conference in France so I don't
        know what is new) and several papers by Bill.

I do not understand what you are talking about, Chuck. Every time someone
sits down at a CSG meeting and runs one of our tasks, that is a replication.
Every one of the hundreds of times a student in one of my classes ran one of
the tasks, that was a replication. If you mean we haven't *published*
replications, go talk to our reviewers and editors. They are the ones who
say: "Of course the results turned out like they did. How could it have
been otherwise. Rejected!"

Do you remember back in 1993, at CSG, when I ran the tracking tasks where
the data were predicted five years earlier? That was a replication and
extension of a demonstration done five years in the past. In the year and
a half since then, the paper about those data has been rejected and
resubmitted twice, and it's been a long time since the latest editor got a
draft from me. As things look now, I'll run the 10-year predictions before
this latest set of data ever goes into print, especially now that
publication of Closed Loop has been shut down.

In the couple of decades since you got interested in PCT, why haven't you
published several empirical papers? I'm not trying to be nasty with you.
I'd just like to encourage you to think a little bit about what you have
noticed (few publications on PCT) and see if you can come up with some
alternatives to your present explanations.

        of these involve one type of "tracking experiment":
        compensatory tracking.

Look more closely. We also publish on pursuit tracking. No one can accuse
*us* of being narrow! :wink:

        No one has presented data
        on "tracking" using such variable as "sound,"
        "shape," or "orientation." No one has presented
        any data that replicates any of the earlier studies.
        I would consider this to be a remarkable achievement
        for PCT and for us who spend most of our time teaching,
        "researching," and trying to convince other that PCT
        is the most useful and empirically supported model
        for understanding the "doings" of a human being. We
        are able to convince some people (in my case some
        coercion is involved - required course and grades)
        that PCT is it!

Do something about it. Start submitting manuscripts about empirical
studies. Then join Rick and me at the next meeting of CSG while we all
laugh over what the editors said in our most recent stacks of rejections.
We'd love to have you join the little club. :slight_smile:

Later,

Tom