(too) small sample size?

#1
Without going into too much detail: I have a number of people every year. At one point, something happens that makes the yearly number of people decline....at least, that's what it looks like without any statistical analysis.
Which statistical analysis should I use to prove that there is a decline? or is my sample size too small to prove anything at all?
 
Last edited:

Karabiner

TS Contributor
#2
technically, it is possible to perform a test with n1=4 and n2=2
(Mann-Whitney U-test), but indeed the sample size is too small
to reach conclusive results.

With kind regards

K.