I have a small data set:

Small   Large    

11.834  7.7951
12.23   8.7618
11.972  8.8966
12.11   9.7235

I would like to perform a t-test. The data represents the lower development threshold "Temperature" for a fly species grown on small and large carcasses.

Am I correct in thinking a paired, two tail t-test should be used in this situation?

Also, will the unbalanced samples affect the result?


2 Answers 2


You certainly can't do a paired t-test, as you don't have pairs. Whether you should do a one-tailed or two-tailed test depends on your hypothesis. The unbalanced samples should not be a problem.


Unbalanced samples cause some statistic issues, but it shouldn't matter here. Your sample size is differ by only one.

You only have paired-data if you repeat the measurement on the same subject, which you aren't. What t-test you use depends on how you assume the variance in each group. Do you want to assume equal variance?

Whether you want to do one-tailed or two-tailed depends on you what you want to test. Do you want to just test for equality? Do you want to test if mean in one group is larger than the other group?


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

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