Bell Curve, Absolute CA scale, Multiple Scales, Learning etc

H. M. Hubey (
7 Nov 1995 18:26:52 -0500

For some strange reason the first attempt did not succeed.

Some months ago I posted a short article on the absolute intelligence
scale and how it resolves the contradictions inherent in the heritability
of intelligence, and the problems associated with correlation-regression
analyis, and factor analysis and the [in]famous book Bell Curve. I haven't
had time to write more on the topic until recently.

I am now almost finished with a paper (actually Appendix VIII.A of my
upcoming book) about all of the above and more, including evolution, a
multiple-tier hierarchical memory (a functional view only), the role of
learning in IQ tests, an absolute intelligence scale, determinism and
direction of evolution, and some stuff on correlation, dimensional
analysis for numbers with unknown dimensions, and more.

I've taken a step toward integrating various views such as Spearman
(scalar number, g), Thurstone (vector not a scalar), Gardner (many
types of intelligence), AI (it's a process not a structure), potential vs
actual/real and have shown how it's possible to have vectors, and scalars
derived from them etc. I've also given examples of measures which use
products of intensive (quality) and extensive (quantity) variables to
define various properties, biological/genetic inheritance,
cultural/environmental/learning effects etc.

It's in draft form but it's good enough for now. It's been sent to a
journal, so it's a preview copy. It's about 40 pages and full of plots,
and equations. It's in compressed postscript format and can be
fetched from:



All commments and suggestions welcome. Please note that it's
compyrighted material and treat it accordingly. Thank you.


Regards, Mark