RE: Islam IS the true religion of humanity.
December 17, 2013 at 11:03 am
(This post was last modified: December 17, 2013 at 11:12 am by Ksa.)
(December 17, 2013 at 10:28 am)Duck Wrote:(December 17, 2013 at 10:19 am)Jacob(smooth) Wrote: Indeed. (although I'm not convinced that one can determine causality from a distribution). However I'm thinking that when 19 says natural she does not mean it in the mathmatical sense. I certainly did not start to learn about gaussian distribution until I was at post grad level so I'm not holding that against her. However it does mean I don't really know what normal means to her.
That's why I'm interested.
I think we are wandering onto slightly dangerous ground here. I also feel people may start confusing normal with a normally-distributed data set. Hardly synonymous and the word normal doesn't mean the same in the two contexts.
I think Nineteen needs to define what she means by 'normal' before we can continue in even a vaguely meaningful fashion.
Nineteen? Can you do so please?
Oh, and I think Ksa is abusing statistics slightly too; chance isn't really the right word to use, it is more accurate to talk about the variation in the data being random or not.
Ok let's do it simply for the 16 homosexual male sample. There is no complication here, grey matter was measured in straight men and homosexual men. Straight men scored an 856 figure and the homosexual males scored a 863 figure.
Question is, what is the probability for the variation to be due to baseline and not to a naturally occurring phenomenon. So first we average all uncertainties:
(59 + 68 + 45 + 52 + 53 x 2 + 67 + 59 + 64) = 57.78
Assuming that homosexuals represent 5% of the world population, we calculate the average grey matter in humans:
0.05 x 863 + 0.95 x 856 = 856.35+- 57.78
On second thoughts, those are big fucking uncertainties, we have a problem here I thought it was 0.57...hmmm well, lets see what it gives:
Standard deviation = (1/40[(856-856.35)^2+(863-856.35)^2+(863-856)^2/ 57.78 ])^1/2
Standard deviation =1.06
So according to the normal distribution the probability is under 30%...this study is shitty I thought the uncertainties were 0.57...if the uncertainty is 57 might as well have varied from 863 to 856, does someone else have another study?