52 STATISTICAL METHODS. 



I. -FORMULAS. 



M= ^f) =v ^_^ P.E. X = 0.6745-^. x=V-M. 



yn 



./2(a; 2 ./) / , <T 



n **' p ' K<r= 4 v^r' 



^p = ^'^ - 0.7979<r. P.^. = g = 0.6745a. 



= " 2 - *!*(+ ft) = 



a ~ Vl + A) = 



A^ (for graduated variates) = l 



A^ (for integral variates) = 2^^ . 100^, where fc equals the number of classes. 



_ S(dev. a;.dev. y.f) _ 

 -- " 



Po (spurious correlation) = - 



+ V 1/tV + V 

 7i (index of heredity, uniparental inheritance) = p . 



x = p, 7i 2 + Pa^i [biparental inheritance: unassortative mating]. 

 <r a <r, 



= P ! ~ Pl ^ 3 . -/ia + ^ ~ Pl1 ? - . /i [biparental inheritance; assortative 

 1 PJ-* <r a 1 pi-* (r a 



mating]* 



