One of the schools in my district has 30 or more. Let’s just say lots of tiger moms and intense pressure. One valedictorian admitted that she cheated, and said most of her friends did also. Created a giant uproar (though probably not better conditions.)
But we have spent a lot of time discussing a somewhat suspect case.
Studying for the SATs etc. I see. But if an IQ test is an accurate measure of intelligence, studying should not help, and if it results in a significant mislabeling, it might hurt. The halo effect exists for sure, but a kid not as smart as advertised by a significant amount is either going to wind up in a job where the expectations are going to cause stress or be labeled as not living up to his or her inherent talents.
I think I’ve had nightmares that sounded like that. And it is why ETS gives all high school students two shots at least. And only count the best.
I did feed my programs in by punch card 45 years ago. It is not all that different from typing on a terminal, except that you have to write things out in advance.
The big difference is that you had to wait hours, not seconds, for a response, and that was true even when I was able to edit a file to submit to a batch run. On the other hand back then you checked what you were submitting a lot more because the cost of a mistake was so great.
I work with lots of smart people doing complicated things who can’t program worth a damn even though they have had classes and even though it would be beneficial to them. It is not a tool they feel comfortable using.
Google does not make intelligence and knowledge unnecessary. My wife makes a good bit of money writing on-line encyclopedia entries using sources open to anyone. But most people wouldn’t understand the sites or see what is important.
It is just like writing - being able to publish an e-book does not make you a good writer.
> Silly question but isn’t that only true if human intelligence fits a normal
> distribution?
Whether human intelligence “really” fits a normal distribution (in any sense of “really”) is irrelevant. All that’s relevant to the norming of the raw scores from an I.Q. test is whether it’s possible to rank the scores of everyone who took the test on a linear scale. Since the raw scores are the numbers of questions correctly answered, they can be ranked linearly. If there should be too much bunching up of raw scores at some place on the curve, the test is rewritten to fix that. Fitting something to a normal curve, regardless of whether that something “really” does match a normal curve, is pretty common. There has been some discussion as to whether intelligence “really” fits a normal curve and what that means:
And it is why ETS gives all high school students two shots at least. And only count the best.