I want to proof the "Relative Age Effect" of a football team. I have a list of birth dates of the team members (about 20 numbers between 1 and 365, the day of the year). I now want to use the KS-Test to test against a uniform distribution.

this is how I did it, but it looks very wrong to me:


D = 1, p-value = 3.331e-16 alternative hypothesis: two-sided

almost all players are born in the first half of the year, yet the p-value seems far too low in my opinion.



ks.test(dates, "punif", 1, 365)

To understand why, try typing


and reading the last bit of the 'Details' section.

  • $\begingroup$ thanks. Isn't the reference distribution a continuous distribution between 1 and 365, in my case it's a discrete: it can only take values 1,..,365. $\endgroup$ – spore234 Dec 8 '11 at 19:13
  • $\begingroup$ I don't think the discreteness would matter unless you had an awful lot more than 20 data points. $\endgroup$ – onestop Dec 8 '11 at 19:32

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.