Unequal group size

Question

[David]

Can you compare 5 males in one group versus 10 females in another group by either a t-test or a wilcoxin test ?

I heard that the group size difference would be a problem. If you can't then is there a way to test them in a way that accounts for the difference in sample group size.

Thank you in advance

[rpolicastro]

Can you summarize what the data looks like, and what question you want to answer with the data?

[David]

Hypothetically speaking could be height, weight, white blood cell counts etc. If the group size was 10 females versus 5 males would that invalidate a t-test or a wilcoxin test for example ? How would you address the difference in group size assuming it was a problem?

Thanks for your help in advance

[h.mon]

This is a purely statistics question, you should ask it at a more appropriate venue, such as Cross Validated.

P.S.: I also did a search for you, here are some results:

Wilcoxon Rank Sum test unequal sample sizes

t test with drastically different sample sizes

[David]

Thanks h.mon for the links.

It looks like a Welch's t-test or a Mann.Whitney (assuming the sample size is not too small) is the way to go here.

Thanks again