Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

IQ tests are also unreliable by design. They might work once, but you get different results depending on the test subject was challenged by questions and problems of similar design.

Highlighting that fact would be helpful and expose such tests as nothing but maybe a rough estimate. Your education is probably more important for that than your capacity to solve problems.

Additionally those that focus on these numbers seem to not be the ones that score particularly well here. And in the end "science" that asserts it can determine intellectual capacity of a group is most likely wrong.

That said, you should NEVER censor yourself because someone could perhaps misinterpret your finding for their own perverted goals. That would undermine science itself and it isn't worth it.

> This isn't fucking Orwell, its people trying to make sure that dogshit research in their journal doesnt get cited by mass shooters.

This is the wrong approach.



> IQ tests are also unreliable by design.

I doubt that they are specifically designed to be unreliable.


Perhaps unreliable is the wrong description. There are significant limitations and these are generally acknowledged. You can do different IQ-Tests and you will notice a steady increase in your score over time, because you learn about the structure of problems they ask about.

Ideally IQ-Tests are systematically different that this would not apply but it isn't possible to create them this way. Your training about similar problems in the past will have significant influence on the score. That does not mean that your intellectual capacity is higher/lower to someone with a different score.

Another weakness is the focus on speed or solved problems per time unit. This is also just an estimate and to latest discoveries about intellectual capacity a very rough one. For example there is a corridor of a certain reading speed that correlates with intelligence. But especially at the upper end which also includes more or less average reading speed this correlation vanishes. Same is true with the comparatively easy problems of IQ-Tests.

The score might be an estimate, but as it I would call it an unreliable estimate. The limitations of them are intrinsic to the way they work.


They aren’t; it’s just that we don’t even know what IQ is, so measuring it is naturally questionable.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: