https://www.ncbi.nlm.nih.gov/pubmed/26708324

The mathematic scores on the International Mathematical Olympiad (IMO) during 1959–2015 indicate a significant and negative trend between the year of competition and efficiency score (ratio of attained score and all possible score) (n = 54, r = −0.71, p < 0.01; weighted by sample size for each year) (IMO, 2016)

This is completely irrelevant to intelligence or IQ. The science Olympiads have been getting harder over time and the (domain specific) problem-solving ability of the competitors has been rising due to longer and longer preparation time and ever improving Internet databases of practice material. Comparing IMO 1967 with 2017 makes no sense.

At the same time, more low-scoring third world countries are joining the contests. The result (at the IMO, where these trends are most apparent) is increasing problem variance: more extremely hard problems to sort out the gold medalists and winning countries, together with a very slowly rising standard of difficulty for the one or two easy problems that are put on the contest to prevent total discouragement among weaker participants.

If psych studies want to analyze time trends in IMO data they should use a fixed set of countries, otherwise the new countries drag down the average scores as a function of time.

If this is meant to explain away the declining WM findings, note that it may be that lower IQ people are simply more prone to be fat. IQ is related to a number of health measures, possibly by way of a general fitness factor, including BMI.

academic gossip: That may be (re: Math Olympiad stuff), but note that the authors are merely indicating that declining Math Olympiad performance may have some relation to declining visuospatial working memory. The Math Olympiad data aren’t part of their main analysis. Your claim that the IMO trend is “completely irrelevant to intelligence or IQ” is premature. It may in fact be that despite the changes you have identified, part of the decline in IMO scores is due to worsening VS WM. Someone should examine the data with the necessary controls applied.

I’m sorry, but, what is that opening graph supposed to be showing me? To me all it really says is that there were more, bigger studies starting in the 90s.

But if I’m supposed to be believe any conclusions drawn from a linear regression and an R^2 of, what, 10%? then I’m not buying it.

It’s a meta-analysis. The studies are weighted by sample size. Are you daft?

Abstracts don’t really say much. Here is the full paper:

http://cyber.sci-hub.bz/MTAuMTAxNy9zMTM1NTYxNzcxNTAwMTI2NQ==/10.1017%40s1355617715001265.pdf

I’ll comment on it in a bit after I read it.

No doubt useless nonsense but the details of (1) the spatial awareness and cartographic skills of Inuits described by Norman Lopez in his “Arctic Dreams” and (2) similar descriptions of the spatial awareness in Aboriginal Australian people contained in Bruce Chatwin’s “Songlines” are both fascinating. Diverse extreme environmental conditions producing human beings with similar looking (to a lay person) mental skills.

Such a shame that they never invented the lawnmower or composed classical music.

Oops! Make that BARRY Lopez.

Lol.

>Fine. How does one specify that increase in difficulty? That would be the basis for an interesting study.

One can give the old items and the new items in parallel to people and analyze the data with IRT to get the difficulty params.

Of course, a possibility that the authors do not consider is that people in developing countries are brighter, which is why their countries are better developed.

Should be “developed” I believe.

My take is that the Flynn Effect is backward looking, in that it results from measuring cumulative knowledge of humankind over time. Even the least bright among us can’t help but soak in some of the collective knowledge floating around. What you are looking at here is a measure that suggests manipulating or synthesizing ability might be in decline.

Hi, thanks for the useful comment!

Granted I don’t have any fancy statistical analysis software, but looking at the blue circles if you started your analysis at 1990 (or maybe even 1975) I don’t think you’d have a very strong case that the trendline changed at all.

I am very skeptical about claims of “dysgenic effects” supposedly demonstrated by this noisy data set.

Eating habits modulate short term memory and epigenetical regulation of brain derived neurotrophic factor in hippocampus of low-and high running capacity rats

Why fasting bolsters brain power: Mark Mattson at TEDxJohnsHopkinsUniversity

Natural selection would select for mechanisms to boost the hungry animal’s ability to find food.

I am pretty skeptical of this study, simply based on how terrible the plot is in Figure 2! I think it would be easier to see if the authors hadn’t attempted to cram quite so much information into one plot.

For example, it might be more obvious that trying to fit a linear regression to this time-series data probably isn’t justified. The complexity of the meta-analysis obscures the very fundamental question of whether this is an appropriate technique at all.

Geeeeee! Eyeballing graphs! What a sophisticated technique! No need for all of those meta-analytic methods, surely. What do those idiot statisticians bother messing around with all that for? You really need to pass your brilliant insights along to the authors of the paper. Your obvious progressivist bias against dysgenic effects ought to be included as well. Everyone knows that’s what science should turn on!

Do you know how linear regression works? Do you know what logistic regression is? Do you know when it’s appropriate to use which? Do you know what R^2 means? Do you know the difference between statistical significance and actual significance? Do you know the difference between frequentist and Bayesian hypothesis testing, and which is the one that gives you answer you care about? Do you have any plausible mechanism in mind that would lead one to believe there is an uninterrupted linear decline in memory for almost a century?

No, you just have a bunch of studies and a statistical analysis package that you learned in grad school that you put numbers into and gives you numbers. Yes, the meta-analysis probably doesn’t have as many problems in experimental design and analysis as individual studies, but at the end of the day you get ridiculous results. The headline number is that BDS performance decreases at 0.01 units/year (plus or minus 0.01 units!), but whether or not the test was administered on computer is good for 0.39 performance difference and “type of country” (developed vs. not) is good for 0.73 units of difference! Tell me, how many of these tests were administered in the developing world on computer in the 40s????

So, yes, keep reciting “meta-analytic methods” as your mantra, because when you don’t know matrix algebra statistics does seem like magic. But as someone who understands the math and has a couple firing neurons, I will continue to use the sniff test to see if the result makes sense.

The progressivist slur, though, was a bit uncalled for… I’m as much of a doom’n’gloomer as everyone else, which is why I’m down on most modern “studies.”

Find out if they consume artificial sweeteners, specifically Splenda, known to reduce memory function and to induce weight gain.

“Do you know how linear regression works? Do you know what logistic regression is? Do you know when it’s appropriate to use which?”

I’m sure your eyeballs and moronic intuitions would’ve been invaluable to the authors in determining something as basic as the model of best fit.

“Do you have any plausible mechanism in mind that would lead one to believe there is an uninterrupted linear decline in memory for almost a century?”

I know of a phenomenon that almost surely exists, and that could leave researchers with a set of data on WM performance over time for which a linear regression would be the best model of fit: http://www.pnas.org/content/114/5/E727. You seem a bit too dull to grok what a statistical model really constitutes.

“but whether or not the test was administered on computer is good for 0.39 performance difference and ‘type of country’ (developed vs. not) is good for 0.73 units of difference”

Are you slow? Do you understand what a moderator is, or how moderators are handled in meta-analysis? Well, why would you? As random unz.com clown #35665823, you have seen through the entire field of statistics. Muh matrix algebra! Just one more progressivist idiot who ends up having to go after statistics itself because he doesn’t like psychometrics.

Is anon vs. anon some kind of Spy vs. Spy or some recursive story? Are you two different people or not? Are you auto-regressive? /s

Churchill’s letter to the other Winston Churchill
London, June 7, 1899.

Mr. Winston Churchill presents his compliments to Mr. Winston Churchill, and begs to draw his attention to a matter which concerns them both. He has learnt from the Press notices that Mr. Winston Churchill proposes to bring out another novel, entitled Richard Carvel, which is certain to have a considerable sale both in England and America. Mr. Winston Churchill is also the author of a novel now being published in serial form in Macmillan’s Magazine, and for which he anticipates some sale both in England and America. He also proposes to publish on the 1st of October another military chronicle on the Soudan War. He has no doubt that Mr. Winston Churchill will recognise from this letter — if indeed by no other means — that there is grave danger of his works being mistaken for those of Mr. Winston Churchill. He feels sure that Mr. Winston Churchill desires this as little as he does himself. In future to avoid mistakes as far as possible, Mr. Winston Churchill has decided to sign all published articles, stories, or other works, ‘Winston Spencer Churchill,’ and not ‘Winston Churchill’ as formerly. He trusts that this arrangement will commend itself to Mr. Winston Churchill, and he ventures to suggest, with a view to preventing further confusion which may arise out of this extraordinary coincidence, that both Mr. Winston Churchill and Mr. Winston Churchill should insert a short note in their respective publications explaining to the public which are the works of Mr. Winston Churchill and which those of Mr. Winston Churchill. The text of this note might form a subject for future discussion if Mr. Winston Churchill agrees with Mr. Winston Churchill’s proposition. He takes this occasion of complimenting Mr. Winston Churchill upon the style and success of his works, which are always brought to his notice whether in magazine or book form, and he trusts that Mr. Winston Churchill has derived equal pleasure from any work of his that may have attracted his attention.

The long term solution, I suppose was to become Sir Winston Churchill. So I propose that Earl of Anon should argue with the Count of Anon.

(Bayesian Boy here.)

Cool reply, bro.

Do you understand what a moderator is, or how moderators are handled in meta-analysis?

I dunno, do you? Does this paper give any indication wtf they actually did? If you can point me to some legitimate sources on the dark arts of statistics that would imply that the authors did not just linearly regress on indicator variables for “type of country” or “on the computers” I’d appreciate it.

Even accepting their methodology as valid, the correlation between year and BDS is -0.02. Stop the g-d presses!

The problem is one quickly runs out of honorifics. We had King Oliver, then Duke Ellington, then Count Basie, and a couple of Sirs (Coleman Hawkins played in the band of Sir Charles Thompson). Poor Kid Rock and Kid Creole had to go the Childe Harold route to use something previously unused. Even Louis Armstrong came along too late in the game and had to settle for “Pops” which is more or less the American male jazz equivalent of “Queen Mum”. That being said, props to each of the anons (unusquisque anonymorum, for the Latinists), it is always nice to see the old “argumentum ab auctoritate” (“argumentum ad verecundum”) non-ironically deployed and responded to : you’d have thought it would have dropped off the charts, or at least appeared only in disguised form, generally for comical effect, in the internet age, but you’d have thought wrong.

Marquis, Marquess, Viscount, Baron, Squire … or that brilliant Americanism of addressing the Attorney General as ‘General’. Doge, Prince, Grand Duke, Elector, ……

Best of all, how about Pope Anon and Anti-Pope Anon?

Once i saw the reference to working memory I was gripped ànd wanted the chance to ask you JT more about it. That was because one of my favoutite nieces doing first year A level was ģiven indifferent predicted grades at her academic school and so got no provisionzl offers from Oxbridge (her acknowledged leasership and acting abilities notwithstanding), at which point Tiger Mother got to work and took her to a psychologist who allegedly (as I remember being told) said her working memory was in the bottom 10 per cent. She is untidy and disorganised at things like cooking, but so am I and I think my short term memory was exceptionally high (though with the little quirk in verbal matters that I would find I had got the sense exactly right but not ipsissima verba). She settled for a place at one of the more academic Ivies.

Now i again wonder what to make of it as her A level results have come in at A*s (and very good A*s) in Politics and History, A in one European language and the equivalent of an A* in another European language which was examined under the Pre U [???] system.

(Her shy younger sister has had 4 A*s predicted but is too young to be considering offers).

She also took some rather unpleasant pills for a while for alleged ADD which one of her other uncles simply scoffed at. But neither of us see enough of her to form a truly informed view so have to basically rely on larental report. There’s something odd somewhere isn’t there?

Still trying to get my mind round working memory and short term memory in their actual working. Being an undisciplined quick learner until partly saved by example from age 22 I was a very good crammer for exams (and no: no one ever thought of me as a potential Lewis Namier in scholarship or Fellow of All Souls if only…) but I wonder if that could be ascribed indifferently to “working memory” or “short term memory”. An argument against is that the all nighter before an exam mightn’t stop one’s short term memory regurgitating but would surely interfere with the nimble functioning of working memory….

Thank you. Very interesting, partly because of further questions it prompts. I had no idea that short term memory, as distinct from the working memory one might be using when making a calculation, could be as short as ~18 seconds.

What about those (almost) all-nighters for exams where a good crammer did it because he could (and too often went to sleep in the library when trying to be dutifully conventional in the many months leading up to the exam)? “Short term” seemed applicable because of the evident rapid loss of the crammed material over the next days and year.

A newly prompted thought about successfully crammed-for exams is that the success was dependant on being able to use the crammed material in a structure of understanding and established formats of intelligent reasoning.
Then the loss of the crammed material, though not explained by its having been in ~18 sec short term memory, is to be explained like the tossing out of any other little valued junk (if only because other temporarily held junk has to take its place) and one might therefore explain it better as truncated processes of long term memory formation which make insufficient connections for permanency.

None of that goes far to explain some of my odd memory experiences that I occasionally wonder about. For example, my attendance in the morning at the first lecture in a new (verbal) subject which I didn’t take notes on but lined up while listening to it with the legible and complete notes that a cousin had previously made of the identical lesson/lecture. With my girl friend that evening I found I could remember the whole lecture apparently verbatim (she was of course enchanted). Decades later, chairing a meeting, I shored up my right to the chair by immediately, verbatim or as good as, replacing 30 seconds of a witness’s evidence that had been lost by a stenographer who fumbled the replacement of her tape. Yet, as previously noted, I have life long evidence of my tendency to remember the exact sense of words reliably without getting the exact (prose) words right. In an English Lit exam in my teens I recall grandly quoting the famous opening sentence of Pride and Prejudice at the beginning of my answer only to find after the exam that I had got about five words wrong while retaining the exact meaning. I am conscious of an apparent inconsistency here in my claiming to remember after the exam exactly what I had written an hour or so before. My resolution is to say that I have a very precise memory of the exact words – “but you did say ‘always’ ” or “it did refer to ‘natural’ children” etc – when it matters to an argument, bùt I probably instinctively save mental energy when possible (leading, as I get older, to irritating requests to others to repeat the exact words). Thia process or habit I might label with a word I remember from MBA days, viz. satisficing.

From what I heard the IMO problems were tested with a panel of professional mathematicians to determine the difficulty of the problem. For the legendary question 6 of the IMO 1988 from the test panel nobody was able to solve that single problem in 6 hours. They set it in the IMO together with two other problems and expected the competitors to solved them in 4.5 hours, i.e. on average 90 mins per problem. Of the 68 competitors 11 managed to solve it. Even Terence Tao at 13 yo got only 1 out of 7 marks for the problem but he went on to win a gold medal. The competitor who solved it the most elegantly using only a single paragraph of proof (may be much less than the 90 mins allotted time), however he only got a silver medal for the overall results.

Video Link

dux.ie