Rubin said:
>
>As someone who has worked on the foundations, I suggest you
>look at the real problem. In principle, you start out by
>considering every possible theoretical model, and you use
>the data to combine with your outlook to produce results.
No. I do not look at every possible model. I prefer hypothesis testing, But
if someone wants to use CR to examine every possible model, they will come
up with one single result,,, the causal model that is implicit IN THE DATA.
They should cross validate their findings, however, because even CR is
subject to chance variation.
>In practice, you cannot do this exactly, as it would
>require an infinitely large and infinitely fast computer
>operating with zero cost. But it does tell you that much
>of the current statistical religion is wrong.
This warning should be sent to the traditional structural equation modeling
folks using LISREL etc. They are the ones who have the problem with
ambiguity. With CR the data support only one model, unlike the case with
LISREL. If you are simply suggesting that we have to know everything to
know anything then you are saying that only an omniscient God can have
anything to say in science, I do not think things are that bad, There is a
reasonable faith that keeps science from grinding to a halt because we do
not understand every tick in the Laplace clockwork universe,
>
> The polarization effect is there. Have you paid any
>>attention to the posts on corresponding correlations/regressions?
>
>WHY should one look at correlations or regressions?
Because we can not do experiments on many of the variables of interest in
the social sciences, Manipulation of these variables may be impossible or
unethical, So we have to make do with measurement and correlation/regression
methods, I cover something of the history of this problem in my latest
paper, with a bit about the place of astronomy as a nonexperimental but
mathematical science,
>Are
>these linear relations even approximately correct? Using
>linear approximations is reasonable for SMALL effects, but
>those using correlations and regressions usually have large
>ranges for their variables.
First, larger variances are desirable in correlational research, It is why
use factor analysis, The eigenvector/value solutions maximize variances,
Such maximization is ideal for our causes while minimal variation is ideal
for our effects, I discuss this in my upcoming paper in Structural Equation
Modeling, where I discuss the history of factor analysis and how Thurstone
almost discovered corresponding regressions,
What you are saying, however, is that it is pointless to look for linear
relationships because they do not exist in nature, I do not think things are
that bad, And certainly 90% or more of the statistical analyses done assume
the possibility of linearity, Some would even say that nonlinear
relationships can be broken down into a set of discrete linear stages or
subsequences, Such slicing may be better than just giving up on linearity,,
especially since most people cannot understand linear theories, much less
nonlinear ones,
>While least squares was heavily
>used in the physical sciences for improving accuracy, it was
>used with theoretical models, which were either simple or
>derived from simple assumptions by theoretical reasoning.
>In fact, many of the early theories, supposedly obtained
>from data, would have been rejected by the usual statistical
>procedures; fortunately, they had not yet come into use.
>When one has simple theories, of the type which someone
>doing logical reasoning would formulate, where the error
>in the data is small compared to that in the theory, they
>can be found by "data analysis". Apart from things like
>Mendelian and similar genetics, biology does not work that
>way, and the social sciences are even worse.
As an experimental psychologist, I do not think you are correct, True human
behavior is complex, but we can systematically isolate and control factors
and find linear effects,
> Stop
>>being so stuffy and condescending and try it yourself, Tell me, are any
>>statisticians trainned to think for themselves, Most of you guys have a
way
>>of ignoring data and logic.
>
>No, it is you who ignore logic by not formulating your
>models carefully, and expecting to get them from data.
Dr Rubin, you have not apparently read my latest papers, Would you like to
read them? They go pretty deep into assumptions, I have been discussing
things here on this newslist with certain industry, attempting to explain my
assumptions but having to waste a lot of time responding to people's snotty
comments, I would prefer straight forward questions about assumptions, I
have made my assumptions pretty clear in the way I formulate the causal
model, The confusion has come from my having to differentiate my
assumptions from those that people who do not bother to read me carefully
assume that I make,
>As a statistician, I MUST NOT tell you what assumptions
>about the model to make, but I also must tell you that
>assumptions are needed to analyze the data, and you should
>be aware that your assumptions can be wrong.
I have spoken at length about logic and assumptions in early posts to these
newlists, I know we all make assumptions, Which of mine do you not
understand?
>
> It reminds me of religious fundamentalists.
>>"The received view (from the masters) is that it can not be done and that
is
>>that."
>
>It is in mathematics that one can show it cannot be done.
How does mathematics show that the polarization effect does not reflect
logical dependencies in the causal model y=x1+x2, where x1 and x2 are
combined roughly in all possible combinations? Show it to me, I have been
begging you mathematicians to disprove this simply simulation since 1985 and
not one single one of you has done anything but claim, without logic or
demonstration, that I must be wrong, This
all appears to come down to the fact that I am not famous and that you have
all invested in a discipline that thinks it knows so much that all it has to
do now is pump out canned statisticians who can not think for themselves, I
think it is irresponsible, even unethical for universities to be trainning
such people to be so closeminded and unscholarly. It leads them to think
they do not make assumptions,
>I suggest you read about self-consistent behavior under
>uncertainty; my paper in _Statistics and Decisions_, 1987,
>has only quoted mathematics beyond what everyone should be
>able to handle.
I would like to read the paper, Are you familiar with my repertory grid
measure of logical inconsistency? I developed it back in 1980 and have
published a number of articles on the method, It is very powerful and useful
in clinical and developmental contexts, But, unfortunately, because I also
published research that disproved the life work of some powerful old men in
personal construct theory, I was ignored... even black balled. I even wrote
a computer program to perform the analysis,,, and everybody else's methods,
hoping people would compare the different methods, I was shunned because I
gave the program away. The old boys were selling their methods for hard cash
and I cost them money by trying to advance the science instead of my wallet,
Yes, I would be interested in reading your paper... only I am an unemployed
PhD, stuck in a small town in Georgia without access to a good library,
Would you mind mailing me the paper?
>There are other publications on this;
>one written for social scientists with little mathematics
>is the book by Clemen, _Making Hard Decisions_. There is
>also Raiffa's book, _Decision Analysis_. All of these
>look at the problem from the point of behavior of a
>"rational" person, using only a limited self-consistency
>aspect of rationality to identify the problems.
I am happy for them, I have developed another mathematical measure that I
call integrative complexity, Together with logical consistency and
corresponding regressions, they make of a coherent system for analysing the
structure of constructions, If I have just gone along with the criminals, I
would be famous by now,
By the way, the integrative complexity measure turns Tversky's feature
analysis inside out and goes way beyond him,
>
> It is as though you have never read a history of math, philosophy or
>>science, Don't you have any idea at all about what it is to discover new
>>things? Why in the world do so many departments hire such dim wits and
>>cowards?
>
>My record on discovery of new things is open to inspection.
>
So exactly what have you discovered about the polarization effect in the
causal model y=x1+x2, as I have defined it?
Bill Chambers
===========================================================================
This list is open to everyone. Occasionally, less thoughtful
people send inappropriate messages. Please DO NOT COMPLAIN TO
THE POSTMASTER about these messages because the postmaster has no
way of controlling them, and excessive complaints will result in
termination of the list.
For information about this list, including information about the
problem of inappropriate messages and information about how to
unsubscribe, please see the web page at
http://jse.stat.ncsu.edu/
===========================================================================
