[R] wilcox.test; data type conversion?

Par Leijonhufvud par at hunter-gatherer.org
Sat Oct 30 19:57:29 CEST 2010


Peter Dalgaard <pdalgd at gmail.com> [2010.10.30] wrote:
> I wouldn't bother with that. The p-value is based on the correct
> covariance matrix of the rank sums, tie-breaking just adds noise to the
> analysis. 

Good to know.

> If you really want an exact p-value, package exactRankTests is

In this case I suspect that for upper secondary students we won't get
that fancy. This is the first time they deal with statistics beyond
mean, median and possibly standard deviation (and simple graphs, of
course).  Thus the request for simple and elegant solutions; arcane
incantations won't make them leave the comforts of Excel.

> the ticket. (Or, if there is really only 3 females in 9 students, you
> can get ambitious and set up the permutation distribution by enumerating
> the choose(9,3)=84 possible outcomes.)

Well, this being totally fake data I created for a student exercise,
there can be as many male and female students in the class as I think
they can be bothered typing in... :-)

/Par



More information about the R-help mailing list