Quote: (01-03-2019 10:17 PM)edlefou Wrote:
The ugly truth is that IQ has some predictive power, but Taleb gives statistical proof that its predictive power is embarrassingly low and what it predicts is not meaningful.
Anecdotal evidence about 80 IQ children and gameless 5'6 guys doesn't disprove his statistical analysis.
Imagine this:
There's an amazing "measure" for predicting the area of a rectangle by taking the length and the width and multiplying the two together.
Many build lucrative consultancies for the military and large corporations selling this measurement methodology.
Racists/Eugenists show that certain groups of rectangles are inferior because the multiples of their widths by lengths have values that aren't as good as the multiples of the widths by lengths of certain other groups of rectangles.
However, Taleb comes along and proves that multiplying the length by the width only gives you the correct area of a rectangle only 13% to 50% of the time, and the rest of the time it gives you garbage.
Taleb's point would be that your rectangle measuring formula is not scientific and is poorly thought out.
Taleb would call you a psychometrics peddler looking for suckers and that your rectangular formula matters far less than you think, as he's shown in his analysis.
How would you prove he's wrong? Are his graphs wrong? Has he made a mathematical mistake?
He's not wrong, he's talking about a different thing as I said above.
As for racists/eugenists, that's your problem :
You are the one letting extremists depriving you of a useful concept because of their twisted world view.
You need to reject their bad framing of the discussion - can you ?