Are the Pearson and Spearman rank correlation coefficients related in a specific way for uniform random variables? Specifically, is the relationship $\rho_{spearman} = 2*\sin(\frac{\pi}{6}\rho_{pearson})$? If so, why?
3
-
$\begingroup$ This is a very specific formula that you are asking about, so I don't quite understand the purpose of the question. Is this a formula that you have derived and would like checked, or a formula that you have seen and whose derivation you find unclear? A little more detail on what you know and don't know thus far would be helpful. $\endgroup$– Yemon ChoiCommented Feb 7, 2010 at 19:50
-
1$\begingroup$ I came across an unsupported statement of the claim by an unreputable source as part of a method to obtain a specified correlation coefficient between two U(0,1) RVs. $\endgroup$– user3875Commented Feb 7, 2010 at 21:17
-
$\begingroup$ wilmott.com/… $\endgroup$– user3875Commented Feb 8, 2010 at 1:43
Add a comment
|
2 Answers
$\begingroup$
$\endgroup$
0
This formula is from Pearson 1907, see e.g.
Rank Correlation and Product-Moment Correlation Author(s): P. A. P. Moran Source: Biometrika, Vol. 35, No. 1/2 (May, 1948), pp. 203-206
Johan
$\begingroup$
$\endgroup$
This is only valid for normal distribution.
