This is a test version of Biostars. For the public version, visit https://www.biostars.org.
Effect of unequal sample sizes on p-value

Hello, I have performed the following test to test if the two groups are different. the p-value is very small, and its probably because of the high sample size.

I want to ask if having unequal sample sizes in the two groups (341 vs 1457) will lead to spurious p-values, as opposed to having the same sample sizes in both groups. My guess is that since both sample sizes are large, it might be fine, but I am not an expert on this. If having such unequal sample sizes is cause for concern, I could take a random subsample from group 2 to have the same sample size for both.

I had asked a similar question earlier, but I could not get a satisfactory response, so I am asking again. Hope that is OK.

enter image description here

statistics conceptual

1 answer

I want to ask if having unequal sample sizes in the two groups (341 vs 1457) will lead to spurious p-values

No, it is fine to have a different number of observation in 2 groups. https://en.wikipedia.org/wiki/Mann%E2%80%93Whitney_U_test - you can see that the test assumes n elements in one group and m elements in another.

Log in to answer this question.