OFF: IQ and adult income

[Paul Mather]

On 5 Apr 2007, at 5:26 PM, M Holmes wrote:

> Gordon Hundley writes:
>
>> On Apr 4, 2007, at 6:16 AM, M Holmes wrote:
>>
>>> Also since the
>>> more bright someone is, the better they're paid (60% correlation or
>>> thereabouts?)
>
>> Heh, that seems unlikely
>
> Huh? I'd have thought that it seems pretty obvious. Because of a
> squaring effect though, that 60% correlation means that IQ would  
> predict
> only just over a third of income variation.

Unfortunately, because income/wealth is pareto-distributed, most of  
the variance is actually contributed by a relatively few observations  
in the extreme tail.  I know from my own experience of modelling file  
sizes (another heavy-tailed distribution) that applying classical  
statistical measures and methods yields poor results.

Your mileage may vary.

Cheers,

Paul.

e-mail: paul at gromit.dlib.vt.edu

"Without music to decorate it, time is just a bunch of boring production
  deadlines or dates by which bills must be paid."
         --- Frank Vincent Zappa