On the Coefficient of Racial Likeness

Karl Pearson
1926 Biometrika  
I hare known cases in which an anthropologist boa measured 00 to 100 characters in the 30 to 80 indlTidualJ of one tribe and sex, who were accessible to him, and he has pressed me to tell him whether this groop vu distinguishable from a «im<lar small sample of a second tribe. Thus the problem is a very real one. * The correlation coefficient of tha mean of two Tftrittei is th* tune u the correction coefficient of the two TuUtet themaelTM. * I noted that oat of the 680 coefficients tabled by me
more » ... ients tabled by me nearly a. sixth, 04, wen for series with lets than 16 crania meunred.
doi:10.2307/2332498 fatcat:bgz5nmgahrh4zffpn5fkaqwzwq