[R] Point biserial correlation => Is there any specific command or could I just use cor.test?

[Michael Dewey]

At 13:15 20/01/2012, Cadu wrote:
>Hello,
>
>
>
>I found in the forum two threads about point biserial correlation. One of
>them (1) mentioned "a point-biserial correlation is just a Pearson
>correlation where one of the variables is dichotomous.  Thus,  the command
>is just the normal cor function". The other (2) mentioned "Professor Fox's
>package polycor" as a way to calculate  point biserial correlation?

polycor calculate polyserial correlations. I am confident if John Fox 
had meant point polyserial he would have written thus. You can always 
check what it is doing by downloading the source - R is open source.

>
>
>HELP: I am a little confused about what to do. I just need to obtain the
>correlation index and p-value. Could I just use cor.test? Or is necessary to
>use other command or even "Professor Fox's package polycor"?
>
>
>
>Thanks in advance,
>
>Cadu
>
>
>
>1. http://r.789695.n4.nabble.com/Point-biserial-correlation-td862060.html
>
>2.
>http://r.789695.n4.nabble.com/Correlation-dichotomous-factor-continous-numer
>ical-and-ordered-factor-td865214.html#a865215
>
>
>
>
>         [[alternative HTML version deleted]]

Michael Dewey
info at aghmed.fsnet.co.uk
http://www.aghmed.fsnet.co.uk/home.html