How can I perform a two sample t-test if one sample is small (n<30) and one sample is large (n>30). The usual method is giving wrong results. What should I do?
Asked
Active
Viewed 67 times
0
-
welcome to the site. Indicate results/output. – Dec 31 '16 at 11:41
-
This should work perfectly well so there may be a slip in your calculations. Please edit in how you know it is giving the wrong results. – mdewey Dec 31 '16 at 13:31
-
1I don't see a problem either but what do you mean by "giving the wrong results"? – Michael R. Chernick Dec 31 '16 at 13:53
1 Answers
0
When you perform t test for unequal sample size, your test loses in strength. Maybe your p value doesn't express the "right" answer because of it. Please check other people's questions about it, as here.
Adelson Araújo
- 224