Ricochet is the best place on the internet to discuss the issues of the day, either through commenting on posts or writing your own for our active and dynamic community in a fully moderated environment. In addition, the Ricochet Audio Network offers over 50 original podcasts with new episodes released every day.
Reporter Goes to Mensa conference, Misses the Point
New York magazine recently published an article about Mensa, “My Week with America’s *Smartest People,” by Eve Peyser. She attended Mensa’s Annual Gathering in July, wrote 2,800 words about what a nice time she had, and concluded, “But if my time at the Mensa Annual Gathering taught me anything, it’s that being ‘smart’ and doing well on tests have virtually nothing to do with each other.”
How did this well-meaning writer miss so much about what made the AG enjoyable for her? I’ve been a Mensa member for a long time. Rereading, I could see how her expectations colored how she interpreted what she saw and heard.
Actually, being intelligent – whatever she meant by “smart” – and doing well on tests specifically designed to measure intelligence have virtually everything to do with each other. That is, the tests work. That wouldn’t mean much, it’s nearly a tautology, but the significance of the tests is that measured intelligence is highly correlated with major life outcomes – for better or worse. All else being equal – though it seldom is – higher IQ is associated with better outcomes.
For one thing, I think Peyser was unfamiliar with the real-world importance of IQ. She uncritically quoted from an interview with Dr. Adrian Owen, lead researcher on a 2012 study of IQ. “When we looked at the data, the bottom line is the whole concept of IQ — or of you having a higher IQ than me — is a myth,” Owen said. “There is no such thing as a single measure of IQ or a measure of general intelligence.” I think she misunderstood what Owen meant. That study identified three major components of intelligence – short-term memory, reasoning and verbal facility. Yes, those are all distinguishable. But they’re also all highly correlated, and all are reflected in an IQ test score.
She goes on to say, “Mensa’s members seemed to be, on average, as dumb as the general populace,” She based that on several sessions she attended where someone offered an opinion that Peyser thought was dumb. Smart people commonly believe some dumb things, that’s not news. I won’t belabor the point here, but Ricochet had a two-part podcast in January, hosted by Steve Hayward with guests Charles Murray and Steve Sailer (https://ricochet.com/podcast/powerline/a-conversation-with-charles-murray-and-steve-sailer-pt-1/) which I recommend if you haven’t heard it.
But my Mensa experience also offers some clues. I joined in 1977 when my son started kindergarten, because I had read somewhere that it was sort of a support group for parents of very smart kids, and I thought someone was going to need a support group when he and the schools encountered each other. I didn’t, as it turned out, and I wasn’t an active member, but I kept my membership because I thought Mensa was a Good Idea in principle.
I moved to Los Angeles in 1992 – new job, in a state where I’d never lived and knew no one, just divorced – and decided to check out the local Mensa group by attending an open house in a nearby suburb. The person who answered the door welcomed me, and said, “You’re new, aren’t you?” I asked him how he knew. “You knocked,” he said.
Then he asked me a few questions, and having learned I taught for a year in Shanghai, introduced me to someone who had been a Peace Corps volunteer in rural Thailand. I didn’t know anything about Chinese agriculture, but as a conversation starter it was worlds apart from the weather and TV football. I had a great time, and kept going back. I attended half-a-dozen Regional Gatherings while I lived in California, and have been to four AGs. Where I live now, in Minnesota, I edit the local newsletter, and I’m looking forward to our first RG since pre-covid. (Many regional gatherings are open to non-members, if you’re interested.)
Peyser’s introductory anecdote is about meeting Kimberly Bakke, “basically Mensa royalty,” she writes. I never met Kimberly, except maybe as a baby, since I did know her mother Cookie Bakke from several of those California RGs. How did Peyser end up talking with Kimberly, among 1,100 people attending? Well, she went to a session on polyamory, met someone there who’s a friend of Kimberly’s, and introduced her.
She notices that a common thread among her conversations at the AG is that Mensa is a place where they fit in, where they found “their people,” where the crowd gets their jokes. She thinks that is a good thing, but she doesn’t quite get why people would need it. She writes, “I have never in my life struggled to find smart friends who get my jokes, … but high IQ is not in the top ten or 20 or 100 qualities I look for in a friend or community.”
Many intelligent people are already in social communities where they mostly know other intelligent people – college faculty, for example. It’s a safe bet that pretty nearly anyone with a Ph.D. could qualify for membership, but they rarely do because they already know lots of Mensa-eligible people, so why bother?
That’s related to another significant fact: Nothing about Mensa is representative. The qualification is scoring in the top 2% on any of a large number of IQ tests used professionally by psychologists. SAT and GRE scores no longer count, but when they did, before the scoring was “recentered” in the 1990s, the qualifying score was 1250.
Of the 6 million Americans who could join Mensa if they wanted to, less than 1% are current members. Of the current members, less than 2% attended the AG. Generalizing about “Mensa members” from such a small sample is a fallacy, on the same scale as generalizing about Americans based only on Mensa members. Danish IQ researcher Emil Kierkegaard has a recent Substack post, at https://kirkegaard.substack.com/p/the-mensa-fallacy, on exactly this error, a study conducted among about 600 Mensa members in Europe, which found higher rates of mental illness among the study subjects and presented it as information about highly intelligent people.
Kierkegaard’s post reviews the extensive history of this error.
He also links (https://www.medrxiv.org/content/10.1101/2022.05.26.22275621v1) to a paper presented at the 2022 conference of the International Society for Intelligence Research (ISIR), in Vienna, Austria, in July 2022. That paper used UK Biobank data to examine the difference in the prevalence of mental health disorders between individuals with high (7,266 people) and average (252,249) general intelligence. Those in the higher group (2 standard deviations above the population mean) were less likely to suffer from general anxiety and PTSD, and no more likely to suffer from other mental disorders.
I knew psychologist Arthur Jensen when I lived in the Bay Area, and interviewed him for Mensa’s magazine. (The interview was spiked by a nervous editor.) Jensen said he did recruit Mensa members for some of his research, because they were easy to find and often willing to join a study. But he was far too careful a researcher to claim they were representative of all high-IQ people.
(Peyser’s article is https://nymag.com/intelligencer/2022/08/my-week-with-americas-smartest-people.html but it’s behind the paywall now.)
Published in General
I think 90% or more of consultants are bad and give the other 10% (or less) a bad name. But I’m still a consultant. Intellectuals aren’t just those other guys. (They may be soi-disant intellectuals, but that is another story altogether.) Thomas Sowell is one of our best public intellectuals.
Music is an interesting memory function. I can hear a song I haven’t heard in 30 or more years, and still recall the lyrics instantly.
But I can’t remember what I had for lunch yesterday.
Nope. This would be saying that doctors tend to the dumb end of their own profession. The low-end “cut-off” is still subject to a lot of variation, and not abrupt. How rare is a +2 sd super-genius doctor? All things being equal, I would expect about as rare as a -2 sd normie doc who somehow, amazingly made it. (These are sd’s of the doctor distro, not the gen pop).
Gardner’s writing (I understand through other reviews) doesn’t really bear on, and certainly does not refute, the well-supported link between IQ, g, and outcomes.
I think it’s really complicated, and I also think we will begin to understand, perhaps a hundred years from now, how to make our own mind and brain work for us the way we make our computer browsers work for us.
I have a problem imagining that a person’s reasoning ability is fixed in the womb, never to be increased again.
That’s not to say that I don’t acknowledge the varieties of mental brilliance that exist. I certainly do. And talent as well. And it is my belief that one does not exist without the other.
When my daughter was eight years old, she took our husky in her room with a drawing pad and two hours later came out with an astonishing portrait of our dog. I had to laugh. I think it was the moment I understood intelligence and talent because you could have put me in that room for a decade and I would never have been able to draw that picture. :-)
I think with brain imaging getting better, that at some point, we’ll see that in some people, areas of the brain light up that are forever dark in others. Perhaps that will be the next step in intelligence testing.
While not “abrupt”, there’s certainly a minimal intelligence level to make it into/succeed in law/med school, but no maximum. So I’d still expect a more lop-sided left-side compared to the right, rather than a bell curve.
I haven’t heard anybody say that, except in the same manner as height. A human in the womb has a yet-to-be-observed-or-known maximum potential height, given the strains of living more or less upright under earth gravity. The eventual height of that person can be affected in many way, prim,arily nutrition and exercise, but under ideal conditions, that person will only be so tall. Therefore, most effects upon height will be seen as negative influences, tending to make a person shorter — malnutrition, injury, abuse of the body.
Likewise, a person has a maximum potential IQ, which will be attained under ideal conditions. Just like height, some exercise is required to bring out this potential, but the genetic potential cannot be increased (let us leave drugs, hormone treatments, or mechanical interventions out of this for both height and intellect). It takes parental and individual effort to achieve this maximum, but some kids are just going to be short when they grow up, and some will likewise be a bit slow. That’s just how it goes.
I would be interested to hear how genetics somehow does not apply to the brain if there are any arguments along those lines. It is just another structure furnished by our parent’s traits, the machinery of growth and development, and a smidgen of random variation.
For me, one of the more convincing observations is that we see reversion to the mean in both height and intelligence.
And gamblers expect a win after a loss.
Do you have the full IQ distribution of physicians from a valid study?
I should have replied more directly to your comment. This excerpt is from the interview I linked to earlier. It’s Gardner’s own words:
I don’t know if he’s right about this. At some point in the near future, we’ll be talking about lit-up neurons and how and why we connect them and use what they contain. I imagine at some point we’ll think we can measure this capacity with some kind of imaging software. That would put us back in the fixed-intelligence theory, that a person’s height and hair color and reasoning and remembering capacity are fixed.
I don’t have an opinion on whether it is fixed or not. :-)
What do “morals and ethics” have to do with “intelligence”?
I don’t know. I agree it is a digression.
That said, I just read a great book on ethics and virtue (which comes from the Greek work arete, which can be translated as “excellence”) that made a very convincing case that intelligence without virtue isn’t clear thinking and reasoning. I think it is an interesting point of view.
I need to clarify my own opinion: I think we have an accurate idea of intelligence on the math side. I’d be willing to bet that we’ve nailed that potential in terms of detecting it and testing for it. But I would guess that we are not even in the ballpark on the verbal side. That’s just my working theory. :-)
About what social justice has to do with justice. He’s criticizing IQ tests for being IQ tests. His real problem is that some people treat them as something beyond what they were created to do. He wants to address those other areas, but isn’t stating it well.
Nope. I am relying on mechanics, not anecdotes. Absent extraordinary intervening factors, a normal distribution is the sum of it components, each of which is also a normal distribution.
Counter: if it’s valid to assume that high IQ professions skew positive just because they are high-IQ, then all sub-population distributions should skew away from the whole population mean. If these were all added up, the whole population would no longer be normal. Now, it may be that this can be forcibly re-normalized, but that math would probably also re-normalize the sub-populations.
So normal+normal = bigger normal, and the rest follows.
“Moral reasoning.” This is not to say that being smart makes you good. It is very much to say that every form of reasoning requires sheer mental horsepower. If we are to avoid calling ourselves meat puppets, then let us say that morality is the acted-out product of decisions made using that mental horsepower.
I don’t want either. I want hard data. As I am not finding it, I am not trusting judgments made on speculations on mechanics or other suppositions.
This is like saying you don’t know that 37+43=80 because you haven’t physically counted objects to prove it. Do you evaluate every perpetual motion machine individually, or do you reject them based on the claim?
There used to be a belief that being moral was being intelligent in the sense that it was smart to be good. :-) And Gardner is very old. :-)
I do see your point, that it has evolved to some sort of purer measure of one’s ability to solve the problems developed by the test writer.
I love you, man. I am laughing so hard here. And, yes, what you say here is true.
I may be confused here — I am not saying that about IQ tests (quite the opposite).
I have to admit that I don’t put a lot of stock in IQ tests. In fact, I’m not sure what the point is in measuring IQ — whatever it is. Or trying to get any kind of measurement of intelligence.
To what use would such data be put?
Okay, I’m sorry. I misunderstood.
For 90 days, or 90 minutes? :-)
I believe Charles Krauthammer used to say something like “The best thing about having a Ph.D from Harvard is never having to be impressed by someone with a Ph.D from Harvard.”
Because I have seen non-normal population distributions, I am simply a little more hesitant to jump on the bandwagon than you are without data. (And I’ve never known a normal physician, but that’s a different issue from population distributions.) And your examples are not germane. Had I ever found where 37+43 did not equal 80, I would be more leery of declaring that it always does. Like when it equals 7A, for instance. Or when it equals 102. Of course, if you translate those all to decimal, it could mean 80 (decimal), 94 (duodecimal), 122 (hexadecimal), or 66 (octal). Alright, maybe that is germane in a way you weren’t considering.
Now, I might assign probabilities to my guesses of whether the distribution of IQs among physicians were normal or some form of abnormal distribution. And the probability I would assign to your view is over fifty percent. I’m just not going to commit to the view until I see data that meets normal statistical standards of randomness.
But testing to get into medical school etc, would have an abby-normal effect on the otherwise-normal distribution.
Betting on chess matches?
It can be very good for schools and employers in predicting likelihood of success (or probable need for remedial training).