Freethought & Rationalism Archive

Happy Wonderer · 10-15-2002, 06:59 PM

Quote:

Originally posted by ohwilleke:

People who perform well on one IQ test have a strong tendency to perform well on just about any other test... from SATs, the grades in schools, to evaluations of relative intelligence by peers or teachers, to reading habits, etc.

Wouldn't performance on the SAT, school grades and even evaluations by teachers be negatively impacted if one was test-incompetent? I'm not sure what reading habits have to do with general intelligence, although a poor vocabulary will decrease test performance. You have no hope at "Miller's analogies" if you don't know what any of the words mean.

Quote:

(Even if a test is biased)there is no evidence that people with greater general intelligence within the group that a test is biased against are performing worse on biased IQ tests than people of lower general intelligence within the group with respect to which the test is biased.

I'm not really sure what you are trying to say here. The hypothesis is that there is a genetic/racial component to intelligence. How does the fact that members of a given group are equally penalized relevant? If the test is written in Latin, some highly verbal English-speakers may be able to pick out a correct answer or two -- they will still perform much worse than any Greek speakers. Maybe I'm not understanding your meaning here.

Quote:

Corrolation is not the same as precisely measurement in an absolute sense.

I'm confused by your meaning. In technical usage, correlation is a measure of the relatedness of two variables. I'm not sure how to apply any other meaning to the term in the context of performance on achievement tests. Apparent correlations ("African Americans tend to be good sports players because I see so many on T.V.") are highly suspect.

Quote:

The vast majority of people have an unambigious opinion about what their race is which is widely shared by the communities that they live in and work with.

That is an interesting observation, but I don't see what it has to do with the hypothesis of a racial component for intelligence. Assuming that the hypothesis were true, if I happened to consider myself Asian would it change my intelligence?

My point was that the fact that race is so difficult to identify in the first place makes it very difficult to test the hypothesis. If the race that I consider myself to be affects my "general intelligence", then any differences would be a psychological phenonoma, not a genetic one.

I think that I agree with the rest of your post, which has good hypothetical examples of the statistician's maxum that "correlation does not imply causation."

HW
(I'm trying to cut down the size of my posts, I hope my ellipses don't change your meaning.)

echidna · 10-15-2002, 11:26 PM

HW, presumably you are familiar with Ravens Matrices. As Doubtingt pointed out earlier, these are supposedly culture-free.

On the subject of race itself, although impossible to objectivify absolutely as ohwilleke says, microbiology for instance certainly makes great use of the existence of regionally grouped biological differences when addressing immunology for instance and why certain peoples are more or less subject to certain diseases. I think sickle cell anaemia is the traditional disease used to illustrate clear the clear importance of recognising biological differences between people.

It seems inevitable then that psychology will also use such groupings to understand behaviour.

Happy Wonderer · 10-16-2002, 08:14 PM

Quote:

Originally posted by echidna:
HW, presumably you are familiar with Ravens Matrices. As Doubtingt pointed out earlier, these are supposedly culture-free.

Raven's Progressive Matrices are (supposedly) a culture fair method of assesssment. This isn't exactly the same as culture free, since difference in a given's culture's attitude towards tests, the attitude of the test administrator towards the subjects, and other biases make it difficult to use the results to compare across cultures. They are designed to fairly compare within cultures, ie "is this kid a good match for an accelerated education program." AFIK it makes no claims to defeat test awareness or other more subtle biases.

The test is "fair" in the sense that it avoids the obvious causes of culture bias -- it uses no letters, numerals, or pictures of objects that may be unfamiliar. (I think I recall that the puzzles used color, which discriminates against the color-blind, but I digress...) Obviously this avoids vocabulary bias, but at what expense? In other words, but we are measuring something, but is it the fabled I.Q.?

For example, assuming we believe in the left brain/right brain hypothesis, Raven's is a very much "left brain" task. It could under-represent the "general intelligence" of people who are stronger at "right brain" tasks.

Examiner bias is an extremely difficult factor to get rid of. You can't do a double blind study here -- the examiner is present can unconsiously bias the results. If you see an Asian kid and think Asian kids are smart, by facial expressions you can encourage them to continue on a problem that you might discourage a different minority kid from completing. Of course, there are protocols and training that attempt to minimize this, but humans being what they are.

(There may be automated versions of Raven's, although that right away presupposes that the subject is comfortable around computers.)

Just so we know what we are talking about, here is a general description that I came across:

Quote:

The Ravens instruments are administered according to LPAD procedures, using a "test-teach-retest" approach. The Set Variations instruments are constructed and administered on principles similar to those of Ravens, with a sample problem for each set of variations which receives intensive mediation, and then independent performance is observed on a series of problems similar to--but also becoming progressively more difficult than--the mediational example. The tasks require the learner to look at a series of designs, and complete the series by selecting a correct alternative from a number of choices. To choose the correct alternative the subject must understand the relationship among the variables. The tasks progressively add variables and change the dimensions used to establish the relationships. What is assessed on these tasks is the subject's ability to think using analogies presented as figural (visual/perceptual) information, and their response to the teaching of strategies to solve the problem. The operations involved are those of perceptual closure and discrimination; the generation of new information through synthesis, permutations and seriation; inferential thinking; analogical thinking, deductive reasoning, and relational thinking.

<a href="http://www.icelp.org/Pages/LPAD_Description.htm" target="_blank">From here</a>

As far as identifying races and understanding fundamental differences, fair enough if they exist. Currently, though, I'm not aware of large-scale careful studies using genetic markers to separate the subjects into "races" (is there a standard for that in the genetics world?) To be meaningful you would need thousands of subjects, and that sort of study would be too expensive for for any social sciences department that I know of to afford...

HW

Edited to add: to say nothing of the campus-wide firestorm that even suggesting such a study would cause...

[ October 16, 2002: Message edited by: Happy Wonderer ]

doubtingt · 10-17-2002, 07:18 AM

Quote:

Originally posted by Happy Wonderer:

The test is "fair" in the sense that it avoids the obvious causes of culture bias -- it uses no letters, numerals, or pictures of objects that may be unfamiliar. (I think I recall that the puzzles used color, which discriminates against the color-blind, but I digress...) Obviously this avoids vocabulary bias, but at what expense? In other words, but we are measuring something, but is it the fabled I.Q.?

For example, assuming we believe in the left brain/right brain hypothesis, Raven's is a very much "left brain" task. It could under-represent the "general intelligence" of people who are stronger at "right brain" tasks.

Examiner bias is an extremely difficult factor to get rid of. You can't do a double blind study here -- the examiner is present can unconsiously bias the results. If you see an Asian kid and think Asian kids are smart, by facial expressions you can encourage them to continue on a problem that you might discourage a different minority kid from completing. Of course, there are protocols and training that attempt to minimize this, but humans being what they are.

(There may be automated versions of Raven's, although that right away presupposes that the subject is comfortable around computers.)

[ October 16, 2002: Message edited by: Happy Wonderer ]

At some point, when many different types of tests that are assumed to tap general reasoning ability show similar results, the possible alternative explanations for performance on any one test becomes a rather implausible explanation for the across test differences.

As for right-left brain differences they are highly over emphasized, but even to the extent they are valid, the Ravens has both analytical properties and spatial properties that supposedly tap the left and right brain, respectively.

[ October 17, 2002: Message edited by: doubtingt ]

doubtingt · 10-17-2002, 07:30 AM

Quote:

Originally posted by MadMordigan:
A more fundamental problem with IQ tests is while they are good at preventing false positives (people can't score higher than their ability) they are useless at prevent false negatives (people can easily score lower than their ability). Without some sort of evidence that every testee is 'trying their best', we have no idea what these scores mean.

Certainly, motivation is always a factor that affects cognitive performance. However, motivation is highly unstable and fluctuating, whereas ability is relatively stable. Therefore, stability of the differences in performance across samples, ages, and SES are far more consistent with ability differences than motivational differences.

ps418 · 10-17-2002, 12:59 PM

Echidna,

Have you read this paper? I read it at work today. It does not discuss between-group differences, but it does provide very good evidence that a substantial portion of the individual variation in IQ is genetic. Do you agree with the authors' conclusions?

<a href="http://www.mugu.com/cgi-bin/Upstream/Issues/psychology/IQ/bouchard-twins.html" target="_blank">Sources of human psychological differences: the Minnesota study of twins reared apart. Science, Oct 12, 1990, v250, p223-266. </a>

Quote:

Since 1979, a continuing study of monozygotic and dizygotic twins, separated in infancy and reared apart, has subjected more than 100 sets of reared-apart twins or triplets to a week of intensive psychological and physiological assessment. Like the prior, smaller studies of monozygotic twins reared apart, about 70% of the variance in IQ was found to be associated with genetic variation. On multiple measures of personality and temperament, occupational and leisure-time interests, and social attitudes, monozygotic twins reared apart are about as similar as are monozygotic twins reared together. These findings extend and support those from numerous other twin, family, and adoption studies. It is a plausible hypothesis that genetic differences affect psychological differences largely indirectly, by influencing the effective environment of the developing child. This evidence for the strong heritability of most psychological traits, sensibly construed, does not detract from the value or importance of parenting, education, and other propaedeutic interventions.

Quote:

The study of these reared-apart twins has led to two general and seemingly remarkable conclusions concerning the sources of the psychological differences - behavioral variation - between people: (i) generic factors exert a pronounced and pervasive influence on behavioral variability, and (ii) the effect of being reared in the same home is negligible for many psychological traits. These conclusions will not come as revelations to the many behavioral geneticists who have observed similar results and drawn similar conclusions [5].

Quote:

Moreover, these findings do not imply that traits like IQ cannot be enhanced. Flynn [35], in a survey covering 14 countries, has shown that the average IQ test score has significantly increased in recent years. This increase may be limited to that part of the population with low IQs [36]. The present findings, therefore, do not define or limit what might be conceivably achieved in an optimal environment. They do indicate that, in the current environments of the broad middle-class, in industrialized societies, two-thirds of the observed variance of IQ can be traced to genetic variation.

Patrick

echidna · 10-17-2002, 02:42 PM

Patrick, I questioned Bouchard�s study on the first page of the thread. In such light, no I don�t think I can agree with his findings.

Quote:

Murray and Herrnstein also rely heavily on Thomas Bouchard, whose study of separated-at-birth twins has "proved" that not only is intelligence largely genetically determined, but so are religiosity, political orientation and leisure-time interests. The Bell Curve uses Bouchard to rehabilitate Sir Cyril Burt, whose twin-based evidence for inherited intelligence is now believed to be fraudulent. Their logic is that Burt's research must have been sound, because Burt's findings closely resemble Bouchard's, and Bouchard's research is "accepted by most scholars as a model of its kind."

That illustrates the sort of scholars Murray and Herrnstein associate with. More reputable researchers have raised many questions about Bouchard's work: While other twin researchers estimate that 50 percent of the average variation in intelligence can be attributed to heredity, Bouchard comes up with 70 percent. Even the twin studies that came up with more conservative estimates of intelligence's "heritability" (itself a highly dubious concept) are flawed because the supposedly "separated-at-birth" twins usually turn out to have been raised in close proximity; Bouchard refuses to let skeptics examine the case histories of the twins he studied, essentially rendering his research into so many "Believe It or Not!" anecdotes (Scientific American, 6/93; The Nation, 11/28/94).

Bouchard, of course, is also a major recipient of Pioneer money-- "We couldn't have done this project without the Pioneer Fund," he told GQ (11/94). And he's a colleague and mentor of (and has some peculiar views in common with) perhaps the crankiest of all of Pioneer's beneficiaries, J. Philippe Rushton.

echidna · 10-17-2002, 03:19 PM

Quote:

Originally posted by Happy Wonderer:
Raven's Progressive Matrices are (supposedly) a culture fair method of assesssment. This isn't exactly the same as culture free, since difference in a given's culture's attitude towards tests, the attitude of the test administrator towards the subjects, and other biases make it difficult to use the results to compare across cultures. They are designed to fairly compare within cultures, ie "is this kid a good match for an accelerated education program." AFIK it makes no claims to defeat test awareness or other more subtle biases.

Interestingly and further to MM�s post, I recall reading of a study around1925 where even the race of the tester was evaluated. When tested by someone of another race, both blacks and whites scored noticeably lower.

Of course those test methods would be outdated now, but more recently there have been studies testing heart rate & so forth, which have found that physical contact (a touch on the arm) with a member of a clearly different race produces marked physiologically different responses to when touched by someone of the same race.

ps418 · 10-17-2002, 03:21 PM

Quote:

Originally posted by echidna:
Patrick, I questioned Bouchard�s study on the first page of the thread. In such light, no I don�t think I can agree with his findings.

On what basis specifically do you disagree with Bouchard et al's conclusions? Most of the objections in the text you posted are merely implications of guilt-by-association. The one substantive objection appears to be directed at pre-Bouchard et al twin studies:

Quote:

Even the twin studies that came up with more conservative estimates of intelligence's "heritability" (itself a highly dubious concept) are flawed because the supposedly "separated-at-birth" twins usually turn out to have been raised in close proximity;

Contact between the raised-apart monozygotic twins is fully discussed in the Bouchard et al paper (some had considerable contact, some had none prior to assessment). They found that contact between monozygotic twins had very little effect on IQ scores. And, for what its worth, the article <a href="http://www.mugu.com/cgi-bin/Upstream/Issues/bell-curve/support-bell-curve.html" target="_blank">Mainstream Science on Intelligence, </a>, the one with 53 signatories, all of which are professionals in relevant fields, states that:

Quote:

Individuals differ in intelligence due to differences in both their environments and genetic heritage. Heritability estimates range from 0.4 to 0.8 (on a scale from 0 to 1), most thereby indicating that genetics plays a bigger role than does environment in creating IQ differences among individuals. (Heritability is the squared correlation of phenotype with genotype.)

Of course the concensus could be wrong! I dont deny that. But I would need to see some better reasons before I can conclude that they are wrong.

Patrick

echidna · 10-17-2002, 04:10 PM

HW, I�ve argued previously about the objective existence of race (even although via arbitrary construct), you may even be somewhat entertained by the early harsh misunderstanding which is ultimately resolved by the end of the thread. Although it doesn�t address the specifics of the immunological significance of race in this case, it goes some way to explaining my perspective on the existence of race. I�ll see if I can find some more specific responses later.

Why is social darwinism wrong? Why is racism wrong? (Page 3)
<a href="http://iidb.org/cgi-bin/ultimatebb.cgi?ubb=get_topic&f=52&t=000276&p=3" target="_blank">http://iidb.org/cgi-bin/ultimatebb.cgi?ubb=get_topic&f=52&t=000276&p=3</a>

My $%!& OSX is only sporadically compatible with the UBB code.

Thread Tools	Search this Thread
Show Printable Version	Search this Thread: Advanced Search

10-15-2002, 11:26 PM	#52
echidna Veteran Member Join Date: May 2001 Location: Melbourne, Australia Posts: 2,832	HW, presumably you are familiar with Ravens Matrices. As Doubtingt pointed out earlier, these are supposedly culture-free. On the subject of race itself, although impossible to objectivify absolutely as ohwilleke says, microbiology for instance certainly makes great use of the existence of regionally grouped biological differences when addressing immunology for instance and why certain peoples are more or less subject to certain diseases. I think sickle cell anaemia is the traditional disease used to illustrate clear the clear importance of recognising biological differences between people. It seems inevitable then that psychology will also use such groupings to understand behaviour.

10-17-2002, 04:10 PM	#60
echidna Veteran Member Join Date: May 2001 Location: Melbourne, Australia Posts: 2,832	HW, I�ve argued previously about the objective existence of race (even although via arbitrary construct), you may even be somewhat entertained by the early harsh misunderstanding which is ultimately resolved by the end of the thread. Although it doesn�t address the specifics of the immunological significance of race in this case, it goes some way to explaining my perspective on the existence of race. I�ll see if I can find some more specific responses later. Why is social darwinism wrong? Why is racism wrong? (Page 3) <a href="http://iidb.org/cgi-bin/ultimatebb.cgi?ubb=get_topic&f=52&t=000276&p=3" target="_blank">http://iidb.org/cgi-bin/ultimatebb.cgi?ubb=get_topic&f=52&t=000276&p=3</a> My $%!& OSX is only sporadically compatible with the UBB code.

Freethought & Rationalism Archive

The archives are read only.