Re: p value for spearman's correlation

[John Darrington]

On Sat, Dec 26, 2020 at 02:15:01PM -0600, Alan Mead wrote:
     On 12/26/2020 2:00 PM, Ben Pfaff wrote:
     > On Sat, Dec 26, 2020 at 11:57 AM Alan Mead <amead2@alanmead.org> wrote:
     >> Isn't it a bug if there is a space for p-value and it's blank?
     > Sometimes this means we haven't found or written an implementation of
     > the algorithm for calculating it yet. (In this case, I haven't looked.)
     
     Still a bug, right?

Failing to report an interesting value is certainly a bug.   Reporting an
incorrect value is surely a more serious bug.   So before adding a fix for
this I'd want to be very sure it was the correct fix.


     But in this case, if I'm not wrong, then the standard hypothesis test
     applies. So the implementation that calculates the p-value for
     correlation coefficients (which are Pearson r's) in CORR would apply to
     both these statistics (which are Pearson r's) in CROSSTABS.

I'm not sure that that is the case.  But I would have to spend some time to
research it to be sure.
     
     I think John may be saying that this isn't the best approach. He's
     probably right, but I'm unfamiliar with the issue to which he alludes.

There is a brief discussion of the issue here: 
 https://www.spss-tutorials.com/spearman-rank-correlation/

but again, to be sure, I'd want to review some of the academic literature
first.

J'