This is a stimulating point. testing correlation with autocorrelated data.<br>The reply by Roger is great! You could test additionally bootstrap methods (rely on Brian Cade's papers). I guess that it does not solve the autocorrelation problem actually, but ....<br>
<br>Further, Ingolf Kuhn has developed a number of measures of autocorrelation.<br>They are available in R. Just take a look at his home page.<br><br>Best!<br>Duccio<br><br><div class="gmail_quote"><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;">
<br>
> Great, this worked, thank you for your quick reply Roger. The missing values<br>
> are intended - I didn't realise that I had to specify such to R.<br>
><br>
><br>
> This probably going to show my vast ignorance of R in comparison to SPSS, but<br>
> is there a way to tell / set the significance level for the analysis?<br>
<br>
?cor.test<br>
<br>
but beware that your observations are not independent, and will have inflated<br>
significance levels, so the test output really should not be used unless you<br>
can first show that there is no spatial autocorrelation in either variable. You<br>
could look at ?Geary in the raster package as a way of checking<br>
autocorrelation. If you are an ecologist, you could look at Fortin & Dale for<br>
background, and maybe the Numerical ecology with R useR! book from Springer.<br>
The Moran() and Geary() functions in raster don't give you a significance test<br>
for autocorrelation, though, you may need other approaches in spdep for that.<br>
<br>
Hope this helps,<br>
<br>
Roger<br>
<br>
><br>
> cheers,<br>
> Rebecca<br>
><br>
><br>
><br>
><br clear="all"></blockquote></div><br> <br>