0

I've been trying to run a t-test on familiarity data, in which I'm trying to see whether an ape reacts more strongly if he is familiar with the sound broadcasted. I have only one observation for x and y in my t-test, which gives me the following error message:

Error in t.test.default(ubt, uat) : not enough 'x' observations

I've seen a similar problem somewhere else (Error using t.test() in R - not enough 'y' observations), but here there was only 1 y observation and more x observations, whereas I'm dealing with one observation in both x and y.

I'm a bit stuck here - does anybody know a way around my problem? This is the data I've entered so far. It works with time (hours minutes seconds) - I hope that won't be an issue...

unfbeforet<-c("00:01:02")
unfaftert<-c("00:22:20")

ubt<-times(unfbeforet)
uat<-times(unfaftert)

t.test(ubt,uat)

Thanks in advance!

Community
  • 1
  • 1
Lundill
  • 71
  • 1
  • 1
  • 2
  • 2
    Please consult a statistician. You can't do statistics without replicate measurements and there are specific assumptions that need to be fulfilled if you want to use a t-test. – Roland Apr 14 '14 at 11:33
  • Agree with @Roland, but the first clause -- **you can't do statistics without replicate measurements** -- is the most important. – Ben Bolker Apr 14 '14 at 12:02

0 Answers0