bootstrap correlation

gianmarco

TS Contributor
#1
Hi,

I am wondering if is it correct to use bootstrap methods to get an estimate of correlation coefficient (Pearson) when data violate some of the assumptions of parametric correlation.

Thanks for any hint you will provide.

Best Regards,
Gm
 

Dason

Ambassador to the humans
#2
What assumptions aren't met? I mean you can always use bootstrap to get the estimates you want it's just a question of whether they'll actually tell you anything useful.