Source: TW
There has been some talk among American academics to do away with a gatekeeping exam called the GRE for admission to American grad-school.
In our days that exam was in part IQ test, i
.e. the verbal+quantitative sections and partly claptrap (a 3rd part I forget).
The 1st 2 parts was indeed reasonably correlated with people IQ
but not the best measure of g
as it could be gamed by those of above average
but not spectacular IQ.
Indeed as a package it was gamed to a large extent
by people from say India and the chIna-s in our days.
It is true that it was reasonably effective in filtering people who intended to attend American univs for professional degrees –
among Indians that mainly included people wanting to credential themselves for tech, business or finance.
However, it was a rather poor predictor of whether someone would be able to do decent science.
Hence, having it as gatekeeping exam for grad school was indeed pretty useless &
doing away with GRE proper for research programs might be a good idea.+++(5)+++
However, this academic movement also wants to do away with something called the AGRE
which was specific to subject you wanted to specialize in.
In my day I recall exploring the biology, biochem/molbio (which I took finally) & chemistry AGREs.
This I felt was actually a good test for your knowledge in the specialty you were choosing.
It was probably even better designed than the entrance exam for an Indian PhD program,
which in our days was called the CSIR (an easy test for any serious student).
Hence, doing away with AGRE seems quite stupid.
The best would be to dump GRE
and simply have AGRE for a research program.
Some Indian PhD programs had their own additional gatekeeping tests.
I took 2 of these decades ago:
the integrated PhD entrance for the Indian institute of Science & that for TIFR/NCBS (unfortunately now aping occidental sjw).
I must say both exams were well-designed and followed up with a good in person interview.
In the end for a person serious about an academic career
there is no substitute for a detailed knowledge of the field
and a “divya-chakShus” for making discoveries.+++(5)+++
The latter cannot be easily tested before hand;
hence the former should be examined.
It should have an element of surprise that levels the field.
For example, in my days many of our acquaintances were learning by rote the structure of ATP
which was one of the common questions in the exam.
If they had been taken by surprise
they would have failed this basic probe weakening the “gaming”.
In the past 30 years
in fields like molbio/biochem
for most part the American univs have seen grad students as a labor market
who will perform leg work in the trenches.
This really did not need a serious probe and only a general IQ check.
As a result it has created an academic ponzi scheme
that Indian programs are forced to emulate.
As a result there are tons of doctors of philosophy
who should instead be termed certified technicians.+++(5)+++
Source: TW
I’ve seen some twts for exams like SAT/GRE and those against it. Going against the grain I don’t think they are good exams at all. While unlike Taleb I do think IQ and g are real, I don’ think those American IQ tests are the best. Instead I think the Indian style exams of our days with multiple days of exams on many different subjects are a better mode of examination for college entrance provided the assessors are good and the exams are set without repetition of the identical problems every other year or so.
Source: TW
I do think the Indian examination system from my days (longtime ago) with multi-day long written tests is better than American style IQ test for college entrance. While the American tests measure ball-park IQ, I they are not really good a clearing chaff from grain. Even great Am scientists’ Feynman&Wilson were incorrectly IQ-estimated. A long multi-pronged exam is a much better test of character than a short IQ test.
The bigger issue is the deshi educational attitude where production of (substandard) engineers&physicians is seen as the sole end resulting in an arms race to game the system. This is accompanied, it seems these days, by inflation of marks on exam making people look better than what they are. In my days in undergrad which I did in desh 1 person would be above 70%, ~5-6 in 60s and rest below that. Perhaps the best thing is to institute a system that separates people who are genuinely good at science early and make the rest go through a longer basic scientific training before studying to be doctors/engs.
Source: TW
Those who understand these matters will realize the nature of the malaise affecting the science&some interesting facets of the sociology of science that perpetuates it; the western model of science primarily focused on getting grad students with high IQ to do science @ the behest of powerful PIs;in the selection process what was not usually checked was whether the high IQ individuals really had the personality profile to do effective science. In many cases they were like the foolish V1s of viShNusharman who resurrected the lion - the prefer to do clever things without checking the logicality or usefulness of such things to generate new knowledge. This combined with a mad rush for magazine publications rather than actually making new discoveries, along with the cartel style boosterism via recommendations, resulted in artful fraudsters generating tons of irreproducible &/or insignificant research; this also helped define what was hot in science often making inconsequential research directions look hot.
Finally there is an interesting ethnic angle regarding fraud: while it is committed by deshIs, chInas &mlechChas the mlechChas have a very well oiled system in place to exonerate themselves & cover up the matter as though they have not really committed any crime in the first place parallels of the cover of crimes of the anglosphere, who are the biggest criminals of all, cannot be missed by the discerning