[R] significance testing under sampling from small finite populations

["Jens Oehlschlägel"]

Dear All,

Are there recommended functions to test for (exact) significance when the
samples are drawn from small finite populations (which sometimes are not much
bigger than the sample)?

I am looking for 

- differences in central tendency
- differences of proportions
- difference in distributions

My current application would be on unpaired comparision of discrete
(6-valued) questionaire data, but I would also appreciate pointers to functions for
continuous data or paired comparisions.


Thanks for any help
Best regards


Jens Oehlschlägel

--