Date:         Wed, 19 Mar 1997 00:00:21 GMT
Reply-To:     Bob Gore <BGore@PSG.UCSF.EDU>
Sender:       "SAS(r) Discussion" <SAS-L@UGA.CC.UGA.EDU>
From:         Bob Gore <BGore@PSG.UCSF.EDU>
Organization: UCSF, ITS
Subject:      Point-Biserial Correlation Coefficient

Does anybody know how to ask SAS to calculate a point-biserial correlation
coefficient?
 
 I have a project in which I am trying create a correlation matrix.  Some
of my variables are continuous and some of them are dichotomous.  For
relationships between two continuous variables, I am entering the Pearson
r.  For the relationships between two dichotomous variables, I am entering
the tetrachoric correlation coefficient.  That leaves the
dichotomous-continuous variable pairings.  Will SAS compute a
point-biserial correlation for these pairs?  If so, how?
 
Bob Gore, M.A., ABD
UCSF
bgore@psg.ucsf.edu