[R-sig-eco] Various test of significance

[Sarah Goslee]

There are lots of ways to do many of those things, and we have
no way of knowing which are most appropriate for your data and
baseline assumptions.

The very best thing you can do is learn how to find options for
performing basic tasks yourself. Within R the help facility can
be useful, especially ?? and apropos().

For web searching, I use http://www.rseek.org a lot for information
on packages, functions, R tips.

If you have specific questions after you spend some time
familiarizing yourself with the options, then we can help.

Sarah

On Fri, Dec 2, 2011 at 7:28 PM, Michel Rapinski <mrapi090 at uottawa.ca> wrote:
> Hi,
>
> I've got a response matrix (Y) composed of multiple quantitative variables
> of the same nature which I am subjecting to multivariate analysis. I was
> wondering if there was a simple way of testing the following 3 things in
> R:
>
> 1) Significance of the coefficients of correlation (from the computed
> correlation matrix of Y).
> 2) Independence of the Y variables.
> 3) Multinormal distribution test.
>
> Thank you,
>
> Michel
>

-- 
Sarah Goslee
http://www.functionaldiversity.org