How does SPSS calculating P-values for Correlations?

#1
Hi,

I'm wondering how SPSS calculates the significance value that it outputs with its bivariate correlations (Pearson's and Spearman's). For example, are
they using Fisher transformations on the coefficients, or permutation
tests on the observed coefficients, or parametric bootstrap, etc? Thank you for your help.
 

Dragan

Super Moderator
#2
Hi,

I'm wondering how SPSS calculates the significance value that it outputs with its bivariate correlations (Pearson's and Spearman's). Thank you for your help.
The computed p-values (or sig. values) are based on Student's t-distribution with df=N-2.

t = r*Sqrt[N-2] / Sqrt[1-r^2]

where r is the bivariate correlation.