Skip to content
Prev 299393 / 398503 Next

fixed trimmed mean for group

Hello,

I haven't found errors in your code. I implemented the test in the paper 
(the first, fixed symetric mean) and it also gives me zero Type I 
errors, when alpha = 0.05. Try to see the value of min(pv) or to plot 
the histogram of 'pv', hist(pv) and you'll see that there are no 
significant p-values, at that level.

Anyway I'll continue to look at it, but my first impression is that 
there is something wrong somewhere and it doesn't seem to be in 
following the formulae.

Rui Barradas

Em 07-07-2012 04:26, Agnes Ayang escreveu: