[ExI] The Catholic Impact (was Re: Origin of ethics and morals)

[Anders Sandberg]

Kelly Anderson wrote:
>> Now, intelligence is due to a lot of small gene components with overall
>> heritability h^2=0.5 or so. In selection experiments h^2=(R-M)/(S-M), where
>> R is the response in the population trait, S is the value it had in the
>> selected parent population and M is the population mean. As parents for the
>> next generation we have a population with mean intelligence S = 0.5*(100-12)
>> + 0.5*(100+12)*(1-1/40) = 98.6 (the first term is the under iq 100 people,
>> the second the slightly decreased 100+ population, and 12 is the expectation
>> of a half normal distribution). So we should expect an IQ in the next
>> generation of (98.6-100)*0.5 + 100 = 99.3. Not much.
>>     
>
> Trying to follow along here... where does the 12 come from? Is that
> the presumed difference in IQ between the priestly class and the
> normal class?
>   
No. I split the normal distribution into two symmetric halves, and 
calculated their contributions separately (since the below average halve 
would not have any priests, while the above average halve would have 
1/40 less people), then adding them together again. The average value of 
the positive half of a normal distribution is sigma*sqrt(2/pi) (See 
http://en.wikipedia.org/wiki/Half-normal_distribution ) so in this case 
it was 12.

If I had time I would likely have done a gene model instead, since I 
honestly don't think this analysis works that great except close to IQ 100.

-- 
Anders Sandberg,
Future of Humanity Institute
Philosophy Faculty of Oxford University