Dear R-pros
I have a problem, for which usually I would apply a chisq.test or a
fisher.test:
40 objects, each given either a 0 or 1, regarding if this object
later on will be remembered by a subject or not.
7 subjects investigated
means: we have a 2x40 matrix, each cell the number of
Probably, to check homogenity you only need to consider a vector of
dimension 8 consisting in the number of times you get any of the
configurations (0,7), (1,6), (2,5),...(7,0). This is most likely a
sufficient statistics. Under homogeneity, it should be distributed
according to a multinomial
On 19 Aug 2003 at 10:15, Christoph Lehmann wrote:
Dear R-pros
I have a problem, for which usually I would apply a chisq.test or a
fisher.test:
try
?chisq.test
which tells you you can say
chisq.test(..., sim=TRUE, B=2)
and the simulation will be very fast!
Kjetil Halvorsen
