I think you are confusing three quite distinct issues.
-correlate-, -pwcorr- and -spearman- all calculate results to machine precision.
Each of them has their own choice of display format, which happen to be identical at %7.4f.
If you pick up the correlation in r(rho), then what display format you use to print that out elsewhere is up to you.
I wonder what use 5th and higher decimal places can be for correlations in any case.
Nick
[email protected]
René Geppert
Still there's one question left to the code: How can I determine the
number of decimals the coefficients shall be rounded to?
(pw)corr and spearman do not seem to explicitly provide for such an option.
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
