From: Brian Phillips (deepbluehalo@earthlink.net)
Date: Wed Sep 11 2002 - 05:14:48 MDT
Date: Tue, 10 Sep 2002 11:47:32 -0400
From: Harvey Newstrom <mail@HarveyNewstrom.com>
Subject: REVIEWS: The Bell Curve
In response to numerous requests that I offer any evidence against The
Bell Curve, I provide the following links. These were quickly googled
up because I don't have the time to do a lot of research right now.>>
We eagerly await your more comprehensive review.
BE forewarned (and so forearmed) that your reviews
will elicit
<The biggest complaints against this book seems to be:
- - the authors were not skilled in the topic they tried to measure>
The standard reply to this is that one of the authors was
a political scientist, the other was a behavioral scientist.
The Bell Curve (which is the "pop science" version)
was primarily a sociological and political critique of
characteristics and hazards in the schema of American society.
The G Factor is what you read if you need the psychometric
data in a form that is not watered down. I warn you, it's not
exactly material for the layman though.
<- - the statistical method they used is invalid for the type of data the
tried to measure (it was for single-factor traits not multi-factor
traits I think?)>
The statistical methods are fairly rigourous. For a more exacting
analysis (including more multi-factor methodologies than you
can shake a stick at) you have to go into the psychometric literature.
<- - they extrapolate data from single-race studies showing correlations
within a group and apply them outside the groups to all races>
This is colloquially called "the Jensen error". It is true that in
strict formalist terms the within group correlations are not
strictly extendable to out-group correlations, but as Jensen
has persuasively argued a biological mechanism which
acts in such a profound fashion, but works so differently
on cognitive capabilities on the basis of apparent geographical
origin in deep time? It's not real likely... no one has managed
to suggest an even marginal alternative to "Jenson's error"
other than envirnomental arguements which are contradicted by
the evidence.
<- - they assume cause-and-effect from statistics which is never
derivable from statistics. If blacks have lower income and lower IQ
scores it may be due to the same lower standard of education, and
cannot be assumed to be a cause-and-effect as this book assumes>
Again you need to read deeper. Children resemble their
biological parents cognitively far far more than their adoptive ones.
The only environmental effects which can perturb this estimation of
heritability factors is prenatal nutrition and possibly neonatal medical
interventions. (Which all the "academic racists"(as they have been so
viciously slandered) agree holds great promise for bridging the gap,
look into the literature on rationals for the Flynn effect and it's
disparate
effects on the lower n-tiles of the normal IQ curve)
<- - there are methods to try to break apart multi-factor traits, such as
comparing well-educated whites and well-educated blacks or blacks with
lower grades and whites with lower grades, which shows education and
not race is the determinant factor. This book does not try to break
apart multi-factors but just assumes the factors they desire to be the
cause. Studies that do try to break apart multi-factors show that
education is a better predictor.>>
Of course they show education is a better predictor. A certain degree
of education can only be attained by a given level of cognitive ability.
Jenson and others have analyzed this multiple factor issue very carefully.
Education is a *better* predictor, because it can subsume effects
from both "g" AND factors such as motivation, personality, helpful
socioeconomic boosts from parents etc. Nobody is suggesting the
other factors aren't important.. they are a Gordian knot in point of fact.
<- - they did not publish this through normal scientific channels and it
was not peer reviewed, and most peer reviewers reject their methods as
statistically invalid>
Of course not. It was a public policy text aimed at the lay audience.
The psychometricists whose evidence they are (comewhat ineptly) using
are among the world leaders in citations in peer-reviewed journals.
Jenson alone is a minor deity in the number of "citation classics: he's
produced. This wasn't a science text.. it was a POLICY text.
<<- - other experts who review it usually reject it or find the
methodology invalid>>
True. Experts also reject Mises as being a heartless idiot too.
Tit for tat.
<- - their representation of sources does not match what the original
sources said
- - there is evidence that some data was cleansed or manipulated to make
it come out right>>
This sort of thing does happen in professional journals also right?
By individuals whose views are not controversial or relevant to
larger political debates?
<<- - many researchers say their data was misrepresented in the book and
did not say what they represented>>
You hear this whenever anyone gets in the hotseat. They may even be
truthful.
<- - specifically the twin studies results seem to have been manufactured
from studies that never finished or made conclusions>
Ah. Wondered if you'd toss this in. The Twin studies issue has been
a triumph of ideology and self-aggrandizement over good scholarship
since the first shots at Pearman and Burt.
<<- - specifically the Scarr-Weinberg study found 1.5% correlation between
racial bloodgroup factors and IQ, but this book misrepresents it as
supporting their claim>>
Will look this one up. First thought is how difficult it is to get
isolating
blood factors even for non-behavioral correlation. Later later.
<- - most IQ research shows that IQ does not fall in a normalized
bell-curve>
You will please support this incredible statement. Thank you.
<- - peer reviewers cannot make the data referenced in the book match the
bell-curve they claim. Different references come up with different
curves but do not show consistent predictability as the book implies.>
You are familiar with the statistical concept of sampling error?
Does the notion of different g-loadings of the various measures of
cognitive ability (SES,education,grades etc.) mean anything to you?
<<- - and/or the curve can take on any shape depending on the scheme used
to measure it>>
Well yes it can. All theories are equally good yes? Including the crystal
speheres
and geocentrism? The normal curve is a fixture of psychological statistics,
and this is unlikely to change. The only minor issues are at the
far far right end (that's people like us Harvey) and the truly severely
retarded (who despite their doubtless vast human worth are not posting
to ExI-L).
<<- - the authors themselves claimed that IQ only accounted for 5-10-20%
of differences>>
References please.
<<- - many of the studies they cite to support racial IQ differences were
funded by conservative groups specifically trying to prove race
differences, such as the Pioneer Fund <http://www.pioneerfund.org/>.
These "studies" are rejected by most scientists as political treatises
and not scientifically valid.>>
References please. Bringing up the pioneer funding issue is guilt by
association and you simply won't go uncalled on it in a forum
of this caliber.
<<- - the book suggests getting rid of welfare and controlling immigration
by race for the good of America. These are clearly political views
and not scientifically measured data.>>
It's a political text for the layman.
<<We should firmly define what this book claims before we can argue for
or against it. This book claims that IQ is the predominant predictor
of success and social class in life.>>
It is . It's not everything but it influences many many other factors
that in turn influence success and social class. This is beyond debate
in the *hard* social sciences.
<< It argues that socio-economic
status has little if anything to do with success and status. >>
Not at all. It argues that socio-econmic status is based on
factors which are argely heritable. Don't confuse a surface
manifestation with a deep trait.
< It argues that these are inborn traits that cannot be changed. >
To the extant that they are heritable no they can't. Short
of extropian ideas creating the "Final Solution" to racism...
i.e. making it a historical footnote.
< People born into one class are there because of genetic racial factors
relating to IQ which cannot be changed or altered. People cannot rise
above their current station in life and better themselves to another
class or status, because these are pre-determined by genes at birth.>
NOT SO! People who DO rise above their parental SES (people
like my parents and myself) do so because of intrinsic endowments
operating in tandem with fortuitous circumstance. People who are
not so fortunate remain at old levels or fall further down.
<<More specifically, the book claims that IQ is genetic and cannot be
enhanced or developed by any means. People's status and class are
determined by genes and cannot be altered after birth.>>
IQ is not genetic (is it a gene?), it is *heritable*! Not heriditary
like a title mind you, merely it has a strong tendancy to be rather
similar in parent and offspring.
<<Specifically,
Murray claims that blacks have a lower average income because they
have lower IQs>>
Yes. This is factually true to the extant of scientific knowledge at
the present time.
< and that social and economic factors have little to do
with it.>>
No no. He is aruging that the IQ CAUSES the social and
economic factors which in turn cause much else noxious.
You are not fallowing the causal chain he is proposing here Harvey.
< He also personally argues against social programs such as
welfare because he believes these people cannot be helped or educated
to do better.>>
Not by telling them they are victims of the evil white man they can't be.
< They will always make less money and hold lower class
jobs because as a race he thinks they are not capable as doing what
other races do.>
Only insofar as their success is correlated to their IQs. Athletes,
craftsmen whose attainments focus on personality factors such as consistency
and the like.. horses for courses!
<I do not believe that these statements are true. They seem to go
against everything that transhumanists believe. They seem to run
counter to many people's successes in improving their lot in life. In
real life, scores on IQ tests vary greatly and are not good predictors
of success or status as this book claims.>>
Earth to Harvey, please stop saying the world is flat ! IQ is an
excellent
measure of cognitive ability. It is largely heritable (probably about.3 to
.7).
Ethnic groups differ in IQ scores, in groups, on average. They have
differing normal curves. These differences lead to performance differences.
IQ is highly relevant. It is not a shibboleth, it is not pseudoscience,
it is scientific fact. It is not a reason for assholes in white sheets to
become more noxious. It is just a fact of life, like the fact that
whites are not exactly dominating the running events at the olympics
right now. :)
It doesn't justify bigotry or hatred, but it is reality. We all need to
live with it till we can change it (at the chromosomal level).
regards,
Brian
This archive was generated by hypermail 2.1.5 : Sat Nov 02 2002 - 09:16:57 MST