/sci/ - Science & Math

https://jordanbpeterson.com/psychology/on-the-so-called-jewish-question/

>>9824051
>cutoff
What he does is arbitrarily define "high IQ" at some level and then take a ratio of the cumulative distributions above that point, before multiplying it by the size of the populations. It might give a rough idea but it's hardly rigorous and with functions as non-linear as the normal distribution I doubt it is accurate if the selection is somewhere near the curvy bits in the middle.

Thing is, I'm looking to find a way to map the overall population distribution onto the selected distribution and find out if the expected ratio of, say, Jewish or Chinese physicists, matches up with the real ratio. The expected ratio is not as simple as taking one value because it will be different as you move up the scale. I'm interested in whether or not there is some neat way to check this.

>>9824074
Someone should do it.

>>9824130
I think I figured it out myself.
>Take general population IQ distribution P1(x)
>Take Jewish IQ distribution P2(x)
>Take proportion of Jews in population Pjew (constant)
>Multiply P2*Pjew=P3(x) (conditional probability)
>Create ratio function R(x)=P3/P1 (probability that individual of x IQ is also a Jew)
>Take known sample distribution (occupation, etc) P4(x)
>multiply R(x)*P4(x)=R2(x) (conditional probability that member of sample with IQ of x is also a Jew, given the overall ratio)
>for total ratio integrate R2(x). Since it is a probability density function multiplied by a constant it will converge to a figure less than 1

Should be computable with MS Excel so any brainlet can do it. Is this a valid method?

>>9824148
Tried that out in excel with some haphazard data. The OP chart suggests an IQ distribution with a mean of ~120 and a standard deviation of ~13.56. Assuming a Jewish average IQ of 110 and an arbitrary standard deviation of 15, same as the general population, put all that into the algorithm and get something like pic related.

According to Lynn in "The Chosen People" Jews were 4.8% of medical practitioners in 2000. This seems fairly consistent with the algorithm's result of a convergence slightly below 4.7%. The data is haphazard and likely inaccurate but if it's close to the truth then Jewish overrepresentation is adequately explained by IQ differences, at least in this particular field.

>>9824263
now try other fields

>>9824074
The jewish iq question is a simple one.
They support each other within the tribe, leading to superior environmental factors compared to the competition, including better education (jewish schools) and have much more money statistically

>>9824279
Well, the problem is finding reliable data. I could extract the sd for the medical professionals from the pic, but I haven't been able to find any figures on the sd for Jewish IQ yet. Look what happens when you change it to 10, for example. The shape of the distribution has an effect, not just the mean. Not to mention, both the overrepresentation ratios and the IQs change over time so one needs data that comes from the same timeframe.

>>9824283
I have read MacDonald, and I've read plenty about ethnic conflicts in historical terms. I have also heard the arguments from neocons like Murray, that it's just IQ. I have seen no analysis of this type, perhaps I am just looking in the wrong places.

I am curious where the truth lies.

>>9824286
IQ varies between the ethnical groups because our ancestors' environments required different developments of their cognitive functions.

The genetical aspect decides our overall weaknesses/strenghts, but the environment determines how we, as individuals, develop our cognitive functions.

Let's say 100 people could get an IQ of 170. Parenthood, socializing, education, adaptability and many more factors decrease the chance to reach your full potential. Even more importantly, there are more rewarding abilities than 'iq', like 'eq' and therefore, very few consciously strive to invest as much effort into increasing their iq as possible

>>9824334
The point is that it's not just 1 or 0.

Maximum potential is determined by genetics, but our whole life influences our probabilities and this complexity is hardly comprehensible. The main issue for us is ideologists using statistics to rationalize their biased world views, but those are the ones financing these kind of studies in the first place

>>9824334
Is EQ also capped by genetics?

>>9824423
>Is EQ also capped by genetics?
EQ doesn't exist

>>9824562
They why are 110 IQ chads managers and CEOs while 140 IQ engineers and scientists work for them?

>>9824591
>They why are 110 IQ chads managers and CEOs while 140 IQ engineers and scientists work for them?
What does this have to do with EQ not existing?

>>9824595
In this scenario IQ is not a predictor of success. Something must exist where IQ fails.

>>9824611
>In this scenario IQ is not a predictor of success.
What do you mean?

>>9824621
Leaders are the richest and if they do badly with high IQ(over 128) then IQ isn't everything

>>9824635
Of course IQ isn't everything, that's why Chris Langan was a bouncer and writes his nonsensical CTMU, but that doesn't mean IQ isn't a predictor of success

>>9824635
That's like saying a chimp is superior to you because it has fucked more bitches in its life

To high IQ people, we regulars (even assuming /sci/ has an above average IQ, which is needlessly generous) are like dumb gorillas. Like mindless beasts. They don't care about the trinkets that we pursue. Nothing we ever do will ever matter and we will be forgotten in just a decade or few after we die.

>>9824637
>>9824671
Success is money and power because that actually matters

>>9824677
define "matters"

>>9824021
>>9824051
https://ideasanddata.wordpress.com/2018/04/24/the-jewish-question-an-empirical-examiniation/

>>9824021
>iq
Not science, sorry. iq is pseudoscience.

>>9825637
>Not science, sorry. iq is pseudoscience.
What do you mean?

>>9825637
Isn't it one of the most valid and predictive constructs in all of psychology?

>>9825648
Literally all middle-ages "methods" were the most "valid" ones back then. That doesn't make them valid.
>>9825642
Yet iq cannot even answer the most fundamental questions:
>doesn't know what is intelligence
>doesn't define intelligence influences
>doesn't explain intelligence physical mechanism
Iq is pseudoscience.

>>9825710
It doesn't need to be able to answer questions like that to be scientific.
Those are important fundamental questions, but they can be separated from psychometrics.

>>9824051
Is Jordy saying that there's a perfect correlation between wealth and iq? That doesn't seem right lol

>>9824021
>there are college professors with a lower than average IQ

Good god.

>>9825812
The main problem is that he is deliberately misrepresenting the math by implying you can compare distributions that way. It's almost like he has already made up his mind on the issue and is picking the method to affirm his existing views.
>>9825710
Define "valid"

>>9824021
Try using a chi-squared test

>>9826008
>chi-squared test
Thanks, this looks promising, will look into it.

>>9826008
>>9826686
ok I think I remember what it was now, and that's good for testing whether or not a sample is likely to have come from a particular greater distribution but it doesn't really do anything to predict the ratios.

>>9826703
I guess you could try to use linear regression.

What data are you using?

>>9825730
The definition you are spouting is applied in social sciences where fenomena isn't even described but implied elements based on the psychology theory of the decade, are used for statistics and thrown in the archive without any kind of physical or direct fenomena explanation.

Iq is included in the social "" science "" topics and research.

>>9827662
Anything I can find. My problem is not finding correlation between distributions. How would you even apply something like that to this problem?

Problem restated:
>the distribution of parameter x in general population is known
>the distribution of parameter x in subpopulation y is known
>the representation of subpopulation y in sample z is known
>the distribution of parameter x in sample z is known
>question: is the representation of y in z adequately explained by differences in parameter x

I first need to get a hypothetical distribution to apply any sort of regression test. How does that even apply to this problem? I can arrive at an expected ratio as I've explained above, but how do I put it through a Chi squared or linear regression test?
See
>>9824148
>>9824263
>>9824285

>>9825637
>Mom, look, I posted it again!