When I have heard of the Flynn effect they usually use ~1920 as the starting point to document changes in intelligence. I believe this period around the start of the 20th century is when intelligence tests were first devised and popularized.
Could the Flynn effect precede IQ tests, and is there any way to determine that?
In short, no. The Flynn effect is very real but its very definition depends on the existence of widespread, standardized tests like IQ tests. For those that aren’t aware of what the Flynn Effect represents, it is simply a statistical acknowledgement that newer generations seem to be getting much smarter based on objective measures like IQ tests.
However, the idea itself seems to have some flaws just based on common observations. The cumulative effect is so strong that an average person born in say 1930 would be considered borderline retarded to retarded today. The problem is that some of those people are still around and still quite intelligent even in their advanced age like some of my family. I am 41 myself and have known lots of people born all the way back to the 19th century and they certainly weren’t notably dumb either (well, a few of them were but most were at least as knowledgeable and competent in the areas that were relevant to their lives as people are today).
I have no problem believing that students today are better at academic work and standardized tests than any generation before them but the Flynn Effect has a major problem because it doesn’t map well at all retroactively. If it did, you would have to conclude that your grandparents and great-grandparents would have been too stupid to hold a job or manage simple life tasks based on their IQ scores adjusted downward because of the Flynn Effect and that obviously was not true for most of them.
An IQ of 70 should be enough for independence and manual labor, which is what most people worked at the turn of the century.
Is the Flynn effect just that generations are becoming smarter over time, or just that people are getting better at doing standardized tests due to exposure?
Is this really saying that most people in the past would be considered mentally retarded by modern standards? This just seems hard to believe.
Except a genius can be engaged in manual labor if there are no other options, and there were jobs which demanded mental acuity which were filled back then.
The Flynn Effect is an interesting conundrum because, one, IQ does measure something which correlates to what you’d expect intelligence to correlate to: Academic achievement, job performance and post-school achievement, and a negative correlation with being convicted of a crime, and, two, the Flynn Effect is remarkably consistent across multiple measures of IQ.
However, people were doing all the things we correlate with high IQ at times when a naïve projection of the Flynn Effect would predict the average person was rather slow. So, if we accept the idea that intelligence has always been normally distributed around a mean, we’d predict that at a certain point in the past there’d have been too few intelligent people to fill all of the intelligence-requiring jobs; that is, we’d notice a shortage of MDs, for example, which was more acute in the past and gradually eased as the Flynn Effect moved the mean IQ up enough for the upper fringe above the “smart enough to be a doctor line” was fat enough to fill the need. I doubt we observe that.
That having been said, it’s hard to compare like-for-like if you go back far enough. What a neurosurgeon in 1910 knew would amount to a fraction of what a current neurosurgeon would need to know; this is edging into the old fallacy of “people in the past were dumb because we’ve had more time to do research and think through implications” but if someone effectively only had to take a portion of a course, that doesn’t refute the idea that they’re less intelligent.
Interesting insights on the Flynn Effect from a dude called James Flynn.
Jim Flynn talked me through my course selection on my first day at university. What a lovely patient man. I kind of wish I had listened to him and majored in political studies rather than history but never
mind.
I think he views his effect in a much subtler way than simply saying that overall intelligence increases over time.
Derleth writes:
> However, people were doing all the things we correlate with high IQ at times when
> a naïve projection of the Flynn Effect would predict the average person was rather
> slow. So, if we accept the idea that intelligence has always been normally
> distributed around a mean, we’d predict that at a certain point in the past there’d
> have been too few intelligent people to fill all of the intelligence-requiring jobs;
> that is, we’d notice a shortage of MDs, for example, which was more acute in the
> past and gradually eased as the Flynn Effect moved the mean IQ up enough for
> the upper fringe above the “smart enough to be a doctor line” was fat enough to
> fill the need. I doubt we observe that.
As recently as the mid-nineteenth century, being a doctor wasn’t considered much of a profession and even the best doctors knew much less than the ones today who barely scrape through medical school:
IANAP, but it seams fairly obvious to me that you will do better on an IQ test if you are familiar to some extent with the type of tasks they give you, either by having taken such a test before or by the fact that you like solving brain teasers in magazines (or at least have a general understanding of how they work).
Oh, I’m absolutely certain he does. After all, if you project the Flynn Effect back very far at all, you end up with people who would be too intellectually handicapped to live on their own being the majority of the population in the fairly recent past, and that can’t possibly be true.
Getting at what the Flynn Effect really is inevitably takes us into a fairly deep discussion of what it is an IQ test is actually measuring. Given all the things an IQ score correlates with, both on the high and on the low side, the tests are obviously measuring something interesting and important, but precisely how much of this something is due to this or that factor is certainly an area of active debate.
I mentioned this in my last paragraph a little bit. It’s honestly a bit hard to come up with professions that are really like-for-like. Authors might be the best thing to compare: How many good authors were there at a given point in the past? Language use correlates with IQ to some extent, in that we’d be able to use it to diagnose gross intellectual deficiencies, and we certainly notice that every decade or so after the invention of the novel (to pick a form invented prior to the Flynn Effect’s period of applicability) has had a certain number of good novels written in it. That alone is evidence that intelligence can’t have risen that far from the past few decades or so: If it had, the people at the beginning wouldn’t have had enough language skills to put together something as complex as a good solid novel.
So that’s not evidence that there’s been no rise, and maybe people as smart as Hemingway were rarer then than they are now, but it is evidence that the Flynn Effect must be recent, relatively shallow, and perhaps not connected to every kind of intelligence.
It’s complicated.
So your saying back at the turn of the century most people were less intelligent because of the scarcity of jobs that required (in your opinion) intelligence ?
Totally incorrect, IMHO. A lot of those old timers in those manual jobs were quite clever and intelligent.
How does the Flynn effect apply to the SAT, which has seen declining scores over the years?
By some measures, average people, and people in certain walks, would appear to have been smarter in the past.
I keep thinking of Neil Postman’s example of 19th-century political oratory, in which it was considered normal for citizens to listen to (literally) hours of reasoned debate at a sitting. Long-form storytelling and lecturing were similarly common.
That’s a trickier one – you have to account for the fact that the SAT is not taken by a random subset of the population and control for the selection bias. A lot more people, from a much broader range of backgrounds, take the SAT today than took it 40 years ago.
Under the heading Possible end of progression on the Wiki page entitled Flynn effect, tests in Norway and Denmark and in the U.K are discussed. In all three, scores are falling.
Does anyone have a table or chart of what SAT scores a random sample of high school seniors in the U.S. would receive each year over the history of the test? If not, then there’s no way to know if the scores have changed on average. If the test scores are just for those students going on to college, the proportion of students in college is much greater now that it was early in the history of the SAT. If the test scores are just for those students applying to the colleges that required the SAT, the earlier in the history of the SAT, the fewer and more selective the colleges that required applicants to take it. The SAT has also had its way of assigning scores changed several times. And of course SAT scores are about what you learn in school at least as much as about your intelligence. SAT scores are useless for examining the change in I.Q. over time.
It’s fairly hopeless to use the SAT scores as a measure of anything over time. There is at least a pretense that I.Q. scores are given to everyone of a given age in a given country, so it’s possible to look at scores for everyone of that age in that country at that point. Furthermore, consider why nobody before Flynn in the early 1980’s ever even considered the question of why average I.Q. scores are increasing. It’s because no one ever noticed it or thought it possible until then. Remember how I.Q. tests are normed. The test is given to a large number of people (say, 100,000) in a specific place who are a specific age. The answers are scored and the scores are ranked in a single linear scale. It’s assumed that the scores will form a normal curve. In a I.Q. test where the mean is 100 and the standard deviation is 15, there will thus be about three people scoring at or above four standard deviations over the mean, since one in 31,000 will be at that level. It’s thus impossible to assign an I.Q. over 160 for a test normed with around 100,000 test subjects.
Before Flynn made his observation, nobody even tried to compare scores over history. Each time a new I.Q. test was created (say, once every ten years), the old test was usually just forgotten. There were generally no comparisons with old tests. It was only when Flynn began studying this issue in the early 1980’s (and only by accident, really) that he was able to compare enough scores of people on old and new I.Q. tests. He was the first one to notice that there were consistent increases in whatever I.Q. tests measured.
As to how people could have survived in the past despite not being able to score nearly as high as us on modern I.Q. tests, Flynn says that the habits of mind today are different in several ways. We now do much more clothing the concrete world with classification, introducing abstractions that we try to make logically consistent, and taking the hypothetical seriously by wondering about what might have been rather than what is. He discusses this carefully in the TED talk that friedo links to. The average person uses much different habits of mind these days than they did even a hundred years ago. When you constantly use these habits of mind much of the time, you’ll score better on I.Q. tests that require these habits. You need to use the habits for many jobs today. Seriously, listen to the TED talk before you comment further.
The argument that society couldn’t function if everyone had an IQ of 70 is what I disagree with. There were lots of children employed in agriculture and industry in the 19th century, and they had lower IQs than the adults but the factories and farms still ran.
If IQ were already much lower than it is now, and many jobs were filled with children back then then society can function with lower IQs.
There is probably some of that. My first-grade (1960) elementary school was down the road from the folks who made the No. 2 pencil an educational cliche, so I started on playing the multiple-choice, standardized test game years before many of my peers. Not that similar tests that were not scored electronically did not exist, but there was something intimidatingly futuristic about the new ones.
Novels are a good choice as long as we don’t fall into the trap of judging them subjectively. Bulwer-Lytton may be laughably antique and melodramatic today, but he could string words into a coherent sentence. Often a very long and convoluted sentence, but it was grammatically correct with a broad vocabulary.
Boy howdy, it is!
The results would be interesting if today’s students could write the 1900 or 1920 IQ tests meant for their age groups back then.
Would they ace the tests or be discovered to have the IQs of a hogshead of kumquats?
If the latter, test cultural bias would be more likely and Flynn would have some ‘splainin’ to do.
Also phrased as “you’re likely to test higher if your background doesn’t mean you check the edibility of the pencil instead of writing with it” - applied to non-industrial peoples as well as those of inherently low intelligence.