1

I computed the T-score and P-value using t.test() for my data, and finally I've plotted the density of my p-value and I've got strange plot. I don't know, why I see peak around 0, which it's density is bigger than 1 ? can anybody help me with the interpretation of this plot ? I used the plot(density(mymatrix)) enter image description here

Danica
  • 21,852
  • 1
  • 59
  • 115
user2806363
  • 2,313
  • 3
  • 17
  • 27
  • @gung, thanks. how can I avoid this problem ? for the density(), one of the kernels should be used ? –  Oct 03 '13 at 14:11
  • 4
    I don't see a problem. This is just how it works. You just have to understand these facts when you view / interpret this plot. – gung - Reinstate Monica Oct 03 '13 at 14:22
  • it would also make things a little less confusing to use `plot(density(mymatrix,from=0,to=1))`, i.e. set the limits to [0,1] – Ben Bolker Oct 03 '13 at 14:35
  • 2
    @Ben That's right, but it would create a biased plot: what is shown between 0 and 1 would integrate to less than unity. See http://stats.stackexchange.com/q/65866/919 for a discussion and some solutions. – whuber Oct 03 '13 at 15:12

0 Answers0