# Re: The Core

Lois Shawver (rathbone@crl.com)
Mon, 18 Mar 1996 12:30:58 -0800 (PST)

Bill,

Okay, I took your suggestion and went back to the message you wrote
called "The Core", and I think it helped me understand what you mean by
polarization. Permit me to try to paraphrase you so you can
evaluate my understanding.

As I now see it, your ideas of polarization rest on the concept of a
simple equation so that two independent variables add to equal a dependent
variable (X1+X2=Y). For simplicity, let's restrict ourselves to
independent variables rated (or ranked I think you prefer?) from 1-9. It
is clear on reflection that the highest possible sums (9+9=18) and the
losest sums (1+1=2) will have perfectly correlated independent variables.
The only way a person can get a Y score of "18" is to get an X1 score of 9
and a X2 score of 9. If we step down from the very highest Y scores and
include the next highest (17), we see we can get that by having nearly
perfectly correlated variables (9+8=17 or 8+9=17). If we work on this way
for both the high sums and the low sums, we will see that the more
midrange the sum is the more variation we find in the difference between
the two independent variables. We can see that the most midrange score
(9) has, I believe, 8 different permutations possibilities. I believe
you are calling this the polarization effect.

Do I understand? Can you tell me how you translate this into your
version of construct theory? Why do you speak of "formal causes" here
and, even more, why do you use the term "formal effect"? And what does
all this suggest to you about the way a person's constructs are "woven"
together?

..Lois Shawver

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%