SAT tests (and other academic tests) aren’t static- just like IQ tests, they’re periodically re-written. I don’t know if they are “re-scaled” like IQ tests are (in order to counter the Flynn effect and keep the average at 100), so I don’t know if the Flynn effect also includes academic test scores.
Also- I thought we were arguing about genetic differences between populations for cognitive ability (also called intelligence)- it’s odd that you’d consider academic test scores to be more relevant for comparing this characteristic than tests designed to measure it (though I don’t think any test, including IQ tests, do a particularly good job of measuring the nebulous concept of intelligence). Perhaps you’re arguing that it’s not a difference in intelligence, but a difference in genetics for test-taking-skills (whatever the test-taking genes are). The Flynn effect clearly shows that large differences in IQ test scores (such as the white-1945 to white-1995 gap) can be due to environmental factors exclusively- along with the no-genetic-evidence-for-the-genetic-explanation thing. Anything that can be said about average black IQ test scores can be said about average white IQ test scores from some period of time in the past. It seems pretty clear that there’s no reason to believe that the IQ test score gap is genetic in nature. You’re left with the clearly absurd argument that academic test scores (like the SAT) are a better reflection of intelligence differences than IQ test scores (note again that I don’t think there’s any reason to believe any particular test does a particularly good job of quantifying intelligence)! It’s you with the double-standard, not me.
So you’re left with the genetic hypothesis and nothing more. Essentially you’re argument (it seems to me) is that different populations have differences in their gene pools, therefore there must necessarily be different potentials for genes for high and low intelligence among different populations. You seem to be arguing that every population must be genetically smarter or dumber than every other population, on average. So you’re just using test scores to sort them- you already know, just because their genes are different, that one population must be smarter than another- and you look at academic test scores to see who should be on top and who on bottom. I think these assumptions are absurd and non-scientific; even if two populations’ gene pools have some degree of differences in their genes for intelligence (and we have no data even to suggest this), there’s no reason to believe that one population’s is superior to the other, or even that the two populations have any functional difference in their average intelligence. Not without actually finding the genes, and showing that people who have genes X, Y, and Z are smarter (or dumber) then people who don’t- and showing that different populations have different average likelihood of having these genes.
Even if this was widely accepted (it isn’t), it doesn’t address my point- not to mention the highly controversial nature of g itself.
This is pretty weak (again relying on the controversial “g”), and you have no link for further reading. I (and some other researchers) see no reason why the white-white 1945-1995 gap couldn’t have the same causes as the black-white gap.
Considering the highly controversial nature of g, and the frequent cites of poor research by the likes of Rushton and Lynn, I don’t find this convincing at all.
Howard Gardner and Robert Sternberg, among others.
Wicherts also says “each IQ gap should not be confused with real (i.e., latent) differences in intelligence”, which has been one of my main and recurring points. I think the same applies to other test score gaps.
I know you think we are arguing about differences in intelligence, because that’s what you are obsessed with, and you cannot see past those blinders. If you look above in this thread, you will see that I have repeatedly try to say that what I am talking about in this thread is the skillset underpinning the ability to perform on academic test scores. The evidence for that not being tied to the environmental influence of SES is overwhelming. Alternate putative possibilities are completely unsupported and unpersuasive. What is left is a difference for that skillset driven by genes, and a genetically-driven difference is the most parsimonious and the most supported explanation.
I believe there is a difference in average “intelligence” as measure by IQ tests as well. So do you, and we have come to an agreement that you think this difference is currently about 15 points on a standardized IQ test, and that you think this was higher in the past–about 21 points in 1972. All IQ test scores for all takers are rising. Because you accept the research by Flynn that the gap is narrowing faster than genetic explanations would allow for, it follows that IQ must be (at least partly) environmentally influenced, diminishing (in your mind) the reasonableness of a jump to a genetically-determined outcome.
Where we continue to differ is not so much the general conclusion of Flynn’s research, but the jump to a conclusion that IQ tests are an accurate reflection of a population intelligence. Perhaps they are. I don’t know how they are developed, how widely they are administered, and to whom under what conditions. They are not my area of expertise, by any means. I find Flynn’s black-white gaps, which you accept so readily, to seem kind of large.
Academic test scores such as the SAT, but also nearly every standardized test for schooling, are broadly administered, and have had substantial and focused studies looking at concomitant SES and other factors. These tests are at marked odds with the notion that intelligence itself is rising. Outcomes on SATs and similar tests are heavily correlated with intelligence in the sense that high intelligence is necessary (though not sufficient) to perform well on them. And, as Chen points out, they have been shown to be g-loaded. A kid with an IQ of 85 is not going to be an SAT star. You can make an argument that a kid with an IQ of 185 might also crap out on the SAT for a variety of reasons, but SATs nevertheless reflect some sort of rough correlation with intelligence given similar environmental conditions such as SES status.
SAT scores are not rising. I have given you the actual raw data above, as well a cite showing you when they were re-centered because the raw data showed the scores to be dropping. The idea that intelligence itself is rising is at diametric odds with academic test scores, and for this reason I do not know what to make of Flynn’s research. For it to actually mean anything at all, it would have to correlate better with some proxy other than IQ tests as currently administered. All the various academic tests are much more broadly administered, with several million students taking it every year, and of course every school in the US also has any number of standardized State and local exams and so on. Were real intelligence rising, it should be reflected on those sorts of exams. I am not aware of any such data.
Even if such data exists, it only means that “intelligence” is partly environmental. It is not some sort of grand seal on the casket of a dead genetic theory. And I don’t personally find it controversial at all that our environment does contribute to our adult intelligence. What your genes do is establish a maximum potential for a skillset given X environment. I believe the brain remains fairly plastic well into early adulthood at least. But at the level of entire populations, I do not think there have been overall rises in real intelligence to the degree Flynn proposes simply because I don’t see evidence for that rise in other proxies (such as academic test scores, but also any number of other intelligence-reflecting proxies). Still, if for some reason his data and conclusions are great, I don’t see that they contribute much to the debate here, which is around the nature of the residual, stubborn gap in outcomes; always following the same general pattern and never closing. A real rise in intelligence is not a reason to discard a genetic explanation.
It is not that SAT scores per se are a better way to measure intelligence. It’s that they correlate with intelligence, and they are very broadly administered. So if you have some sort of data on millions of test takers, and your IQ data is on thousands of test takers, any disparity should give pause as to which one better reflects a real average for a total population.
But again, the idea that the Flynn effect " clearly shows that large differences in IQ test scores (such as the white-1945 to white-1995 gap) can be due to environmental factors exclusively" is silly. If it’s real, it simply shows that intelligence is partly environmental, and I don’t disagree with that. When you eliminate the difference, you have data that says “exclusively” but right now all you have is this strange notion that dumb whites now are about the level of smart blacks X number of years ago, and that smart whites X number of years ago would be dumb whites in today’s world.
Worse, in yesterday’s world’s absolute standard of cognitive function, your assessment of average black intelligence based on Flynn’s data puts half the entire black population into a sub-80 IQ category, which is well into a very low-functioning category. This means that a typical adult black in your Flynn world in the 1940s was pretty much unable to work at anything beyond menial labor even given an opportunity to learn job skills as an adult. I don’t believe that.
The “unsupported” part is false (though they are not conclusive), and the “unpersuasive” is just your opinion, and obviously I disagree.
The “parsimonious” is your opinion (and I disagree), and the “most supported” is false.
If IQ test scores rose over a period of time but academic test scores like the SAT did not, then I think there is just one conclusion: the skills/abilities that IQ tests measure have little to do with the skills/abilities that SAT tests measure. I think it’s likely that neither test measures intelligence particularly well.
I don’t think “real intelligence” (whatever that is) is rising either. Whatever IQ tests measure seems to have risen.
The “never closing” and “same general pattern” is false- these outcome differences have only been looked at closely for a few decades, and by some measures, they have shrunk.
Most parts of this statements are very, very false about “my assessment of average black intelligence”.
The Flynn effect demonstrates that anything that can be said about black people’s IQ scores can also be said about white people’s IQ scores from a period of time in the past. Since I (like the previously cited researcher Wicherts) don’t believe that either of these IQ score gaps reflect any real difference in actual intelligence, I also don’t believe that the “dumb whites” of the past were actually any dumber than white people now. Similarly, I don’t believe a gap in SAT scores reflects any real difference in actual intelligence. It may reflect a real difference in something, but there’s no reason to believe this something is genetic without actual genetic data. I believe there are plenty of explanations just as “parsimonious” and reasonable that are environmental- many of which (like lowered teacher expectations) are known to exist and known to have an effect on a child’s educational development- that remain that might explain these gaps. I think you totally disregard (maybe you just don’t know) how significantly different it is to be a black person in America (and likely in other countries), and how pervasive and significant these experiences can be.
You can believe anything you want, as a matter of faith.
As a matter of fact, however, Flynn’s data and analysis either reflects an average rise in intelligence for a whole population, or it does not.
If intelligence itself has been rising, then at least part of what we call intelligence is environmental. I don’t find that controversial. In addition, the environmentally-related part of average black intelligence is rising faster than whites.
The dilemma you then have is that, if you accept Flynn’s data and analysis as reflective of the whole population average intelligence, you have a very large residual gap, and even larger former gap, and an average intelligence that was very low in the past.
You cannot have it both ways. You cannot use Flynn’s conclusions to trumpet that intelligence itself is rising, while simultaneously arguing that IQs don’t measure intelligence particularly well. If IQ tests are not valid measurements of anything, then nothing of substance is rising, and Flynn’s research has no significance because measurement of IQ tests have no significance.
If IQ tests do measure intelligence, and Flynn’s data is correct that IQ tests as currently administered are able to reflect intelligence averages for whole populations, then the average intelligence for adult blacks in the US is 85 in 2002. This is exactly your “assessment of average black intelligence in 2002,” because you are promoting Flynn’s research and you are saying it means something for the average IQ of black and white populations. This is the basis for why you think his research is significant.
It is a hugely different thing for Flynn to say that actual population intelligence is rising as opposed to looking at academic test scores. They are two very different things, even though a higher intelligence is a requirement to do better on an academic test, on average. One is a measure of potential; the other a measure of the application of that potential to master taught content.
It’s true that IQ intelligence conclusions apply equally to whites and blacks. If blacks were dumber in the past, so were whites. Neither seems likely to me. But the degree to which blacks were dumber in the past is so extraordinary, according to Flynn, that I find that conclusion alone very difficult to swallow. IQ tests are designed to reflect a general cognition required to function in daily life. Flynn’s numbers for blacks put them in a genuinely limited mental state 60 years ago. It doesn’t matter if you believe that’s true; that’s exactly what IQ tests mean.
And since academic scores are not rising and are more widely administered, I am not inclined to put much stock in the Flynn effect as a reflection that intelligence is rising–for any whole population–very much or very fast.
I want to know where all that rising intelligence went.
In any case, it’s clear that the skillset gap underpinning outcomes for academic test scores is not closed at a population level, and is definitely not eliminated by perturbations in SES that supposedly drive the environmentally-dependent IQ scores.
Jeez, how many times do I have to repeat myself? Like Wicherts, I believe it does not.
I don’t.
No, IQ tests can measure something “of substance” without measuring intelligence particularly well- it can measure things like problem-solving skills, spatial skills, pattern recognition, etc.
Each of these three sentences is false. You keep saying you have little expertise in IQ tests. And then, for some reason, you seem absolutely determined to demonstrate just how weak your knowledge of IQ tests are.
There is some indication that the gap is shrinking, there remain plenty of reasonable non-genetic explanations, and no genetic data to support the genetic explanation. So there remains no reason to believe that the best explanation for the test-score gap is genetic.
What non-genetic factors account for the observed differences?
In terms of genetic data until very recently you could have said the same about height. Both intelligence and height are heritable quantitative traits. It’s only in the last year or so that some of genetic variants for height have been uncovered.
The most parsimonious view would be that group differences reflect individual differences (both environmental and genetic). No uniform factor has been identified that depresses scores in a single population (see David C Rowe’s papers I cited earlier).
It could be many things- I’ve suggested some, but I’m not confident in any particular factor. I think it’s likely that there are things intrinsic to the “black experience” in America (and perhaps elsewhere) that serve as obstacles to academic and educational development and achievement. But even having no idea what the causes are, there’s no reason to elevate the genetic explanation without actual genetic evidence.
Do you have any evidence the black-white gap is still shrinking (say, over the last ten years) for either IQ or academic tests? I would like to review it.
Since you want to avoid using the term “intelligence” for what IQ tests measure, does it remain your contention that for skillsets such as problem solving, spatial skills and pattern recognition, the black aptitude at a population level corresponds with its IQ equivalent of 79 in 1972, and substantially lower in decades prior? And in 2002, the average adult black population in the US had a full standard deviation lower for those skillsets of problem solving, spatial skills and pattern recognition than whites and asians (as measured by IQ tests)?
And are you contending that an IQ of 79 or lower does not have any actual real-life correlation with a diminished ability to function in daily life? Is it the case that you think Flynn’s data for an average IQ in the 70s for the adult black population 60 years ago does not reflect a “genuinely limited mental state”?
The unique “black experience” hypothesis has already been addressed. Again see Rowe et al. (1994) who carefully investigated this:
To the extent environmental influences are causing the gap, they are influences that are common to both Black and Whites and which have the same relation in both populations.
See also, Rowe & Cleveland’s [1996 paper Academic Achievement in Blacks and Whites: Are the Developmental Processes Similar?/ also found that the best fit model indicates that differences are due to the same causal factors as individual within group variation.
Who says they’re substantially lower in decades prior? They may have been, but I haven’t seen such data. No one has put forward the notion that the Flynn effect has gone on forever- or even for more time then the data suggests. As for the rest of your question, I’m not going to bother trying to interpret the weird use of language. This is a tangent you’ve been focused on, and I feel no need to humor you.
No- 60 years ago, I don’t believe the average black people had a “genuinely limited mental state” (whatever that means). I haven’t seen IQ test score data for black people from 60 years ago- but whatever it was, like Wicherts, I believe that people, on average, were not intrinsically less intelligent (again, whatever that means) in the past as compared to now (as Wicherts puts- it an IQ test score gap does not necessarily imply “real (i.e., latent) differences in intelligence”).
This is ridiculous- that paper doesn’t even come close to getting into all of the various environmental (including media, culture, education, discrimination, etc.) factors that could go into the “black experience”.
In addition, most of the data from the cited Rowe paper is self-reported survey data from children. Self-reported data is notoriously unreliable, for many reason- self-reported data from children is hardly going to be any better.
Wouldn’t you expect the developmental processes to be somewhat different though from the factors they looked at? What evidence do you have to show that discrimination pushes down scores in one group, while say pushing it up in others (eg. Ashkenazi Jewish & East Asian high group averages?).
Then you have to wonder how those factors would affect things like cranial capacity, reaction time measures, differences from age 3 upwards, larger group differences on backward digit span (more demanding) relative to forward digit span memory tests etc?