0
$\begingroup$

I have two equally long time series samples, drawn from two different populations, I would like to test if the means of the two samples are statistically different given population size differences, one population is much larger than the other. Can you suggest an appropriate test to do this?

The data is measuring the adoption of X (defined as : X is chosen / availability for X to be chosen) on a daily basis for a website. The two populations present X to the user in different flows so the samples are generated differently.

I was considering Kruskal-Wallis ANOVA but I am still unsure.

Improve this question
$\endgroup$

1 Answer 1

Reset to default
0
$\begingroup$

You could draw samples with the size of the smaller one from the bigger one randomly and then calculate their mean. Then you can check in which part of these cases the mean of this freshly drawn sample has a larger difference to the mean of the large sample then the measured small sample.

Improve this answer
$\endgroup$

Your Answer

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

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