Talk:Flynn effect

Flynn's "discovery" and "increases mainly among the less able"

Flynn did NOT "discover" the increase over time. Thorndike (1975), for example, provides detailed documentation of the relative increases on sub-scales of the Binet and proffers a range of possible explanations that are still being touted today. Raven (1981), re-published in Raven (2000), demonstrated the increases on the Raven Progressive Matrices among young people in the UK. What Flynn did (using the RPM!) was demonstrate that the effect was almost worldwide, ie not confined to the US and UK, and draw attention to the implication that these scores had been subject to a massive, and previously unsuspected, environmental impact which had implications for the interpretation placed on cultural differences. What he failed to do, and still to some extent, fails to do in continuing to offer his "basket-ball" analogy, is recognise that whatever explanation is accepted has also to explain dramatic international increases in such things as height and life expectancy.

As to the increases being concentrated at the lower end, what the adult data re-published in Raven (2000) show is that, as Flynn suggested, the data, reported by many previous researchers, that had previously been interpreted as showing a decrease in many abilities with increasing age must be re-interpreted as showing that there has been a dramatic increase in these abilities with date of birth. On many tests this occcurs at all levels of ability. Thus there is now abundant evidence that it is not true that the increase has been concentrated in the lower end of the distribution (and there are, in fact, horrendous measurement problems involved in substantiating any such claim). Tall people have got still taller: the whole distribution has moved up.

Having cited this evidence, I am re-inserting the material that someone deleted at the beginning of this month. However, since it would overweight the introductory paragraph to refer to these other studies at this point I am confining the alteration to simply saying that Flynn drew general attention to the increase and its implications.

Huh. I've just noticed that there isn't even a reference to Flynn's publications in the entry! I've inserted a couple.

Raven, J. (2000). The Raven’s Progressive Matrices: Change and stability over culture and time. Cognitive Psychology, 41, 1-48.

Thorndike, R. L. (1975). Mr. Binet's Test 70 Years Later. Presidential Address to the American Educational Research Association.

Quester67 13:43, 31 October 2006 (UTC)[reply]

What's this mean?

"We do not know any more than we did 50 years ago." Obviously mankind collectively knows more than we did 50 years ago. Is this meant to say that a person is not born more cognitive than he/she would have been 50 years ago?

Cultural expansion

The reason for this is the spread of European culture. The "Flynn effect" is in effect a measurement of how rapid European culture is being spread worldwide and nothing more. It is just another form of European cultural imperialism. Set the stardards and have the power (guns and resource control) to judge everyone else by those stardards. --Shabaka Tecumseh

Except that it has also (or especially) occurred among white European populations (in Europe and the US).

The fact that the increase occurs especially in White European nations, could just be further evidence of the growth of a European cultural standard; that the unifying aspect of it becomes most readily apparent in the places where it is strongest to begin with. --Shabaka Tecumseh

I certainly appreciate your concerns, and the issue of cultural bias in IQ tests is a significant and important issue, however I don't believe it holds water here. First, as the disscussant above noted, the Flynn effect is in fact most pronounced in homogenous, white European cultures. Now you argue that this may be further evidence of the growth of a "European standard" but I think you would be hard pressed to prove that the Dutch (for example) are significantly MORE homogenous or European now than they were 40 years ago! Yet the Flynn effect is clear there....

A second and perhaps more compelling argument stems from the fact that the Flynn effect is, if anything, more pronounced when we assess people with tests that are the LEAST culturally biased. Psychologists have individually adminsitered IQ tests that are purely non-verbal and can be administered with no verbal directions to persons with no formal education. These tests apparently show the least cultural bias (e.g. differences between cultural/ethnic/racial groups are lowest on these tests) yet the Flynn effect is, if anything, stronger with them. Again, while cultural bias in testing is an important issue, it is not the core of the Flynn effect.

I've got to disagree. The rise shown by the Danes seems to have tapered off, and then started going backwards. This actually strongly supports Shabaka's point, as the cultural biases first boosted the Danish scores, then, as they became 100% homogenised and the biases in the IQ tests were slowly removed due to better understand of the subtle bias effects, they came to depress those same high scorers.

Another factor is repeated testing. If you keep doing IQ tests, your scores go up. One case in the US was decided along these lines. US law says that those who cannot understand what they did wrong may not be put to death, and it has been ruled that this is an IQ of below 70.[1] Re-normalisation effects are an issue here, as are the repeated tests, as since the first test, where the guilty party scored just 68 (and was therefore safe from state execution) on the first tests, but who now scores around 75. Of course, the possible better nutrition in prison may have assisted, but it is most likely the repeated taking of IQ tests that caused the rise.

[1] http://thescotsman.scotsman.com/international.cfm?id=675512002

Training

Could it be that more and more people are getting training in doing in doing IQ tests? I read somewhere that there is a noticeable increase in the result from the 1st test you take to the 3rd one. Probably you learn to think like the test or something.

I think that is true, people are learning the way of the tests. Once I came across the following question. Which one is the odd one out: Train, Plane, Steamboat, Car, Bus. Ok, Trains can only move on tracks, Planes can move in 3 dimensions, steamboats move in water, cars are small and buses have commercials all over them. So which one is it? In the end it was the car, but with no explanation. But after a few of these you can probably figure out what the test writers were thinking. I guess people are figuring out the mindset of the people who made the tests.

Remember, this is not an effect of the same people retaking the test, but changes in the average scores in a population across generations. I.E. If the 18 year olds averaged 100 on a test in 1948, the 18 year olds now might average 128 on the same test. Now perhaps we are all exposed to more testing, but remember, this effect has occured dramatically in even the last 20 years. Has the average persons exposure to standardized IQ tests changed that much since 1985? It's a puzzle...

On the question above, I think it'd be the car, because it is the only one that you yourself are in control of. I don't know about exposure to IQ tests (I think I've taken two or so...) but standardized testing itself has exploded, with many students taking multiple AP, SAT I/II, ACT and even graduation tests in only four years.

Exercising intelligence

Flynn himself does not believe this [populations becoming more intelligent over time] to be the case. It is conceivable that something about modern society (the greater need for abstract thinking, presence of computers, more visually-oriented culture) is responsible.

Other arguments aside for the moment, I would say that modern society's requirements could be considered a precise indication of how intelligence could increase over time. After all, if modern society required greater use of muscles, the additional exercise would automatically make a population more muscular. Or, more appropriately, our equivalent "fitness quotient" (FQ?) must be going down over time, as technology spreads throughout the world and allows us all to become more sedentary. It seems to me that Flynn's disbelief implies an underlying assumption that "intelligence" is necessarily some ill-defined genetic trait that couldn't change so quickly over just a few generations. --Jeff Q 08:17, 20 May 2004 (UTC)[reply]

Original research ?

I don't know if it has been emphasized by some scientist(s) but IMO :

- the expense in education in many third-world countries were considerable (I have no serious datas yet),

- the progress in communications is simply incredible : It's not a secret on Wikipedia that I spent some time in Burkina Faso around 1986. In 14 month I could phone my mother once! Just click on http://www.cenatrin.bf/ to verify that the volume of information you can exchange has increased. --Ericd 20:36, 25 Jan 2005 (UTC)

Only increasing the lower IQs?

"Comparison of the IQ distributions indicated that 1) the mean IQ had increased by 9.7 points (the Flynn effect), 2) the gains were concentrated in the lower half of the distribution and negligible in the top half, and 3) the gains gradually decreased from low to high IQ."

Is it correct though, that if the average IQ in developed countries is rising but the higher intelligence individuals are staying the same, than we would actually see the higher IQ scores coming down (because the IQ test is calibrated to the population average)? --Nectarflowed 21:58, 26 Jan 2005 (UTC)

Correct. For example, if you were to say, try to compare Isaac Newton's IQ with a modern day person, you would have to adjust one of the two scores, since the scores are relative to the population, and the population would have changed. You might find this intersting: [1]. --maru 16:16, 2 May 2005 (UTC)[reply]

Standard deviation?

"The average rate of rise seems to be around three IQ points per decade."

Does this apply to tests with a standard deviation of 15 or some other value? Could this be mentioned in the article text?

Thanks, nyenyec ☎ 05:13, 16 July 2005 (UTC)[reply]

Microevolution

I have read many articles about the Flynn Effect but no one seems willing to consider the possibility of natural selection as the cause. In this case it would be called Microevolution or small scale changes in gene frequency over small time scales. If we agree that higher intelligence strongly improves a person’s chance of career and financial success, why not consider the possibility that this extends to reproductive success? As evolution on this time scale would only change the distribution of existing genes and not create new ones, this would be entirely consistent with the Flynn effect being strongest at the lower end of the intelligence range. No new and improved “smart” genes are being created. Instead, existing “smart” genes are becoming more common. It is also possible that the reproductive benefits of intelligence are non-linear. I suspect that the bottom tier of the intelligence spectrum within each cultural group is significantly less successful in reproduction that every other tier. The key point here is that reproductive success would need to be examined within the context of cultural groupings.

While I have no proof that this is the primary cause of the Flynn effect, it seems almost impossible to avoid including it as at least a contributing factor. A hundred years ago the primary selective pressures on the human population were factors related to mere survival. Live was “nasty, brutish, and short”. Higher intelligence would have been of secondary importance. The measured beginnings of the Fynn effect approximately correspond to timeframe when infant mortality declined and overall life expectancy began to increase. This would have inevitably changed the mix of selective pressures on the human population. Datlas99

This has been considered, however, lower than average IQ is actually of a reproductive advantage in modern societies. The fitness cost of high IQ and/or a college education are large, and thus the expection would actually be a slightly lower average IQ over the time period. --Rikurzhen 01:35, August 24, 2005 (UTC)

Can you cite a reference for this ? I'm think the way this subject was studied was critical to the conclusions that can be drawn. Stating that a lower than average IQ has a reproductive advantage only leads me to wonder how the populations were defined and how they stratified the I.Q. ranges. Think about this, the reproductive influence of I.Q. doesn't have to be linear across the entire I.Q. spectrum to have an affect. The gene frequencies for the entire gene pool can be moved by a strong selective pressure on a narrow segment. Has anyone looked at the reproductive success of the bottom tier of the I.Q. range? Datlas99

Eh? "Fitness cost" of high IQ and education? Where are you getting that from? A college education is financially an excellent investment, and smarter people live longer happier lives. Perhaps you are referring to how richer people tend to have fewer children, and education and IQ tend to lead to more money...? --Maru 04:26, 24 August 2005 (UTC)[reply]

Longer, happier lives, yes, but does that mean more children ? That's the only aspect that counts for evolution. See dysgenics. --Flammifer 08:03, 24 August 2005 (UTC)[reply]

No it doesn't, Flammifer. The assertion is dubious to begin with, as it contradicts the Flynn effect. Furthermore, the "only" effect that counts for evolution is whether or not alleles spread within the gene pool. Children numbers are almost irrelevant. It has yet to substaniated that being smart or going to college makes one reproduce less, indeed intuition would lead one to suspect that given that college is an excellent financial investment, it would lead to greater reproductive fitness, not less, and that such an effect could be significant, and not irrelevant due to other considerations, and that this is in accord with the very well supported/observed Flynn effect. --Maru 12:58, 24 August 2005 (UTC)[reply]

I'm not sure what you're reffering to with "The assertion is dubious to begin with" :-P

As for the number of children being irrelevant, mm, I'd have to question that. Seems to me that's the most common way of spreading alleles in the gene pool, though there are other strategies (helping your relatives spread their genes, making sure your children have kids too, etc.). Flammifer 15:15, 25 August 2005 (UTC)[reply]

What I meant was the whole Flynn efect is about the average populace getting smarter, no? So you're arguing that the stupider people should be outbreeding the smarter implies the populace average intelligence going down, not up. --Maru 21:51, 26 August 2005 (UTC)[reply]

That wouldn't be surprising if the Flynn effect did not represent a real change in average ability, which is a possible interpretation of the effect. It is also possible that both a positive effect and a dysgenic effect is occuring simulateously and the overall trend is upward. --Rikurzhen 22:27, August 26, 2005 (UTC)

Whoa there guys. Here's a paper along the lines of my claim about negative fitness from increased IQ: PMID 12267266. Feel free to do more literature searches. --Rikurzhen 22:09, August 24, 2005 (UTC)

Also, relative fitness is--of course--all we are able to measure. --Rikurzhen 22:12, August 24, 2005 (UTC)

Facts?

I think you are missing the point here: - Were are the facts that show that the IQ levels are really increasing? - Have the same IQ tests been used to proof this point? I have my doubts.

Well, I don't think anyone would even suggest doing the study without using the same IQ test, not when you are looking at so many hard to control variables. Thus, when the IQ scores go up, it can be pointed to that IQ scores have increased over time. Yes, it might be better guesswork, but there is a fairly strong correlation between IQ scores on IQ tests, and actual IQ.

Look at the article on IQ - there is no such thing as 'actual IQ', IQ is defined as the scores on IQ tests, compared to the scores of the general population. It is meaningless to claim (and impossible to ascertain) correlation between scores in tests and some imagined 'actual IQ'. Cederal

Rise of IQ scores vs. rise of IQ test norms

The article is somewhat misleading. IQ scores cannot rise - they are periodically recalibrated to have an average of 100. The Flynn effect is the IQ test standards gettig higher and higher, ie. previous generations would have scored lower on today's tests and the current generation higher on earlier tests. I cannot find Flynn's article online, but its abstract is clear about this:

"Demonstrates that every Stanford-Binet Intelligence Scale, WISC, WAIS, WISC-R, WAIS-R, and WPPSI standardization sample from 1932 to 1978 established norms of a higher standard than its predecessor. The obvious interpretation of this pattern is that representative samples of Americans did better and better on IQ tests over a period of 46 yrs, the total gain amounting to a rise in mean IQ of 13.8 points."

--Tgr 22:04, 16 March 2006 (UTC)[reply]

A very misleading article which seems to ignore the fact that IQ scores are normalized. By definition, the average IQ score is 100 and doesn't "increase". The average general intelligence in a population may increase, or decrease, but the average IQ is always 100. I'm sure Flynn's papers aren't as sloppy about this as this article is. - Nunh-huh 02:47, 18 April 2006 (UTC)[reply]

The article is certainly incomplete, but your claims are incorrect. An IQ score is normalized to mean 100 based on a representative sample of maybe N=1000 people. The actual population mean will be something around 100, but not precisely 100. If some kind of relavent demongraphic shift were to occur after the normalization, then mean IQ would truly change. This change could be counter-balanced by re-normalizing against a new representative sample. However, this does not change the fact that the average scores did change. Only the renormalization complicates this. If you look at a graph of average scores per year, you'll see that scores were increasing relatively constantly between normalization years with sharp drops after re-normalization -- the Flynn effect. --Rikurzhen 03:27, 18 April 2006 (UTC)[reply]

On the contrary, my "claims" are correct. IQ is normalized so that the average score is 100, and when it varies from that value, it's time for it to be renormalized. The article confuses "Intelligence" with "IQ score". - Nunh-huh 03:54, 18 April 2006 (UTC)[reply]

The delay between renormalizations is quite a bit larger than your claim suggests, such that the effects of renormalization are highly noticable at the tails of the distribution. The article confuses "Intelligence" with "IQ score". Huh? The word intelligence occurs predominantly in contexts of questioning whether the changes in IQ associated with the Flynn effect refect a change intelligence. There are lots of things lacking in this article, but distinguishing IQ and intelligence doesn't seem to be one of them. --Rikurzhen 04:26, 18 April 2006 (UTC)[reply]

Basically, this is the issue: When it is stated that IQ test scores are rising, is it referring to the raw score or the adjusted score? Kurt Weber 13:31, 24 April 2006 (UTC)[reply]

I fail to see where anyone could get the idea that this is an issue. The Flynn effect points out that similar demographics taking the *same test* will rise over time, with the most recent groups generally scoring more highly on the test than those previously. This is not a hard concept to grasp. Similarly, more recent IQ tests have to be harder than older IQ tests in order to keep the recorded score for the same demographic at the 100 baseline on the new test, rather than giving the same rising result, which would lead to the newer test having a baseline of 105, for example.

Grantees

Jensen's work outside of race and intelligence is quite highly regarded.[2] 3rd party reprint It would be highly unorthodox to refer to the Pioneer Fund / accusations of Marxism or lysenkoism every time respected scientists like Jensen / Gould or Lewontin are mentioned.--Nectar 21:21, 20 June 2006 (UTC)[reply]

Gould and Lewontin are no longer relevent for this field. I do not oppose adding Marxist, if you so desire. The Pioneer Fund is relevant, when person connected to an advocay group is cited, then this should be mentioned.Ultramarine 22:26, 20 June 2006 (UTC)[reply]

A source is needed to connect their opinions about the Flynn effect to their receiving grants from the Pioneer fund. Without a source, it's a OR synthesis to put the two facts together to make an argument. --W.R.N. 21:02, 11 February 2007 (UTC)[reply]

My humble speculation on substantial causality of the Flynn Effect

Even as everything from soup (amino acids) to nuts (arginine) has been proposed as explanations for the Flynn Effect, I feel puzzled by what may be an overarching oversight by researchers and spectators: larger headed babies can now be born due to medical interventions---most notably and ubiquitously, the caesarian section. That procedure alone is, and has been for some time, routinely done worldwide. Homo sapiens’ pelvic capacity is generally acknowledged to be near its limit as for birthing a baby with a large cranium. Common sense dictates that many large-brained fetuses never made it into the breeding gene pool. Yet many who previously would not have made it, have been doing so within the general time frame of the Flynn Effect phenomenon. --Paul K. When one considers that a sizable proportion of caesarians result from heads too large for the birth canal, then it's easy to imagine the number of larger-headed babies who've been born from this common intervention.. The correlation between cranial capacity and I.Q. is not large, but it is a substantive and positive one. Pelvic capacity has an upper range that could preclude an increase in birthable cranial circumference; with caesarian intervention, which is quite commonplace even in a Third-World or low socioeconomic milieu, this upper limit can now easily be breached.

A problem is that larger head doesn't mean smarter. I read a report some years ago that the average IQ had actually decreased over the years. 1851, before public education in the US, was actually the high point in literacy in the US. I believe most Americans score lower on IQ tests than at any time in the past - strange ( actually not really I suppose considering the political implications ) that we are evolving so fast according to Flynn. However, rises and falls in an IQ test means that it isn't actually measuring native abililty - what is it measuring? and is the test worth a hoot if it is really measuring daddy's wealth, food intake, school quality, etc. If IQ between siblings varies so much then how important is heredity - unless they have different fathers.

Large differences between sibling IQ indicates a high heritability. You only inherit 50% of your parent's genes after all. While brain size has a weak correlation to IQ, there's a correlation non the less, so Paul K's speculation might be correct. Also keep in mind that people have been getting taller, and taller people have larger brains. This is probably one of the contributors to the Flynn effect. And mankind is indeed dumbing down, there's research in that area that is listed in the dysgenics article. --Scandum 12:47, 12 December 2006 (UTC)

My humble speculation regading the main cause of the Flynn Effect

I feel puzzled by the fact that everyone seems to assume that the Flynn Effect is caused by an increase in people's intelligence, and no one suggests people aer just becoming better familiar with the questions in a standard IQ test. In the 1950s people taking an IQ test would see that kind of question for the first time, while in the 1990 everyone (in western countries) would have seen a book titled 'how to prepare for an IQ test' on his local bookstore.

People solve the questions faster because they already know the general task to be performed, and already know what solutions to expect. This knowledge developed gradually over time, as more and more people became exposed to the test, slowly shifting the average.

I don't know if this is original research, and I find it hard to believe as this explanation seems trivial to me. Has really no researcher suggested such an explanation? Could it be that people who publish in this field are so attached to the supposition that IQ tests measure an intrinsic generic property of people rather then measuring a single skill - to get high scores in IQ tests...

Cederal 10:29, 23 December 2006 (UTC)[reply]

Well, I believe this issue has been considered, not only regarding the Flynn effect, but in the general debate about the validity of I.Q. tests and the difficulty in keeping them valid as a population becomes more sophisticated. Presumably, tests have been continually reworked and restandardized in this attempt, and the psychometric community at least feels this has been addressed . . . and therefore the Flynn effect may not be a result of what you have suggested.

Morphic Resonance

Seems to me that the Flynn effect supports Rupert Sheldrake's ideas of morphic resonance. 192.139.140.243 05:19, 6 January 2007 (UTC)[reply]

split section

the psychometric properties of the flynn effect should be distinguished in discussions from the attempted explanations of the causes to the extent that that is possible. --W.R.N. 21:04, 11 February 2007 (UTC)[reply]

problem sentences

The "Abecedarian Project", an all-day program that provided various forms of environmental enrichment to children from infancy onward showed IQ gains that did not diminish over time
This explanation is troublesome for those who argue that IQ tests, and especially Raven's, measure a general intelligence factor accurately.

reasons:

this is disputed by spitz (sp?) and others
how common is this view -- who can it be attributed to? as stated, i imagine it must have been controversial b/c RPM is a constant subject of study. --W.R.N. 21:54, 11 February 2007 (UTC

Only citing from the source, Neisser.Ultramarine 22:07, 11 February 2007 (UTC)[reply]

Decline of IQ Scores in Germany, Austria, Norway and Switzlerland

Hey, I am a guy from Germany, who has a minor in psychology and writes for the German Wikipedia. In Germany that fact is much discussed, that we have seen a decline of the iq scores since the 90s. This has been proven by professor Dr. Urs Schallberger from the University of Zürich, Siegfried Lehrl from the University of Erlangen and Jon Martin Sundet from the University of Oslo. Here are some German links about this:

I think that should be added into the article. I would do that myself if i was able to speak english well enough.

Would you agreee with me, that this facts should be added? If yes, could you do me the favour and add them? Thanks! Patrick -- Cumtempore 09:02, 19 March 2007 (UTC)[reply]

edit: i just saw, that Norway has already been mentioned. So forget, what i said about Norway. Nevertheless Germany, Switzerland and Austria should be mentioned to my mind. -- Cumtempore 09:08, 19 March 2007 (UTC)[reply]