Can we get back to talking about PCT now?

In a message dated 97-12-18 23:55:57 EST,

<< Bill Powers (971218.1913 MST) --

>>Can we get back to talking about PCT now? Sounds good. For starters,
let's review all the important discoveries about mind and behavior that have
been made through documented empirical PCT research over the past 50 years.
Regards, Bruce >>

I (Mark) would like to say "it would be much quicker to review all the
important discoveries about mind and behavior that have been made through
documented empirical Psychological research over the past 2500 years."

For Example list all the Psychological constructs and theories that contain
the words - Always, Never, Will and Does -- when referring to a principle,
law, cannon, or any other objective measure, that will apply to all living
creatures on this planet. (Please, exclude Bio-chemistry, and stay in the
realm of Psychology the study of mind and behavior)

Let me start the list
1) PCT
2) _____________

(I can't seem to think of another)

Can we get back to talking about PCT now?

Happy Holidays

Mark :slight_smile:

As a side note, during my graduate studies me and few others decided to hunt
through the Psych Journals to try to find corr. Data higher than .7. The
hours turned into days there were a few but when looking at them further I
don't recall any corr.s that were above +/- .7 that weren't biologically
based., meaning I don't recall a traditionally based Psych construct having a
corr above +/- .7 with any other variables.

I know there is more to life than a corr. But in regards to predictive
validity how much variance do you have to account for with any statistic
before you can say your right 95 out of 100 times or should you push for 100
out of 100, after all they are just numbers. Would it be so bad to say "Well
Mrs. Smith, I am sorry your son committed suicide, but look 95 others did not
when I used this variable to predict

[From Richard Kennaway (971219.1825 GMT)]

Mark Lazare, Fri, 19 Dec 1997 13:07:42 EST:

I know there is more to life than a corr. But in regards to predictive
validity how much variance do you have to account for with any statistic
before you can say your right 95 out of 100 times or should you push for 100
out of 100, after all they are just numbers. Would it be so bad to say "Well
Mrs. Smith, I am sorry your son committed suicide, but look 95 others did not
when I used this variable to predict

From the same paper I just mentioned, for the bivariate normal

distribution, to predict the sign of one variable (relative to its mean)
from the value of the other with 95% accuracy requires a correlation of at
least 0.988. 99% accuracy requires c >= 0.9995. The general formula is c
= cos(pi(1 - p)).

-- Richard Kennaway, jrk@sys.uea.ac.uk, http://www.sys.uea.ac.uk/~jrk/
   School of Information Systems, Univ. of East Anglia, Norwich, U.K.