[R] Help joincount.test

CD camille.dezecache at gmail.com
Sun Jan 29 10:37:22 CET 2012


Hello,
I'm trying to analyse the spatial organization of different fields planted
with different varieties (each field has only one variety), but I have
problems trying to understand the results of the test I did.
To do this, I created different neighbourhood matrix. For example, for the
first matrix, fields are considered as neighbours if they are distant from 0
to 1000 meters from each others (independently on the variety). In the
second matrix, I consider fields as neighbours if they are distant from 0 to
3000 m.
Then I test the spatial organization with the joincount.test function in R.
The help page says "The BB join count test for spatial autocorrelation using
a spatial weights matrix in weights list form for testing whether
same-colour joins occur more frequently than would be expected if the zones
were labelled in a spatially random way."

When I take 0 < d < 1000 m, the test gives (for variety A) : same colour
statistic = 4.5, with expectation = 2.6 (highly significative). I thought
this meant that one field of variety A had an average of 4.5 neighbours of
the same variety A, while the expected number of neighbours of the same
variety was 2.6. This means that for this variety and this range of
distances, the fields are strongly organized.
But, when 0 < d < 3000 m, the test gives : same colour statistic = 2.3, with
expectation = 2.5. Because of this result, I'm not sure any more that the
same colour statistic gives a number of neighbours of the same variety,
because if it was the case, this number could have increased with increasing
d, but it could not have decreased with increasing d as I found here.

Can somebody help me understanding what really means this "same colour
statistic"?
Thanks a lot,
Camille

--
View this message in context: http://r.789695.n4.nabble.com/Help-joincount-test-tp4337880p4337880.html
Sent from the R help mailing list archive at Nabble.com.



More information about the R-help mailing list