Non-parametric test for null mean=0
Posted: Mon Mar 17, 2014 7:45 am
Hi everyone,
I have a dataset, that is definitely not normally distributed. Here is the workfile containing the series xyz:
I would like to test the null that the mean is equal to zero. Unfortunately I am not a statistician and not too confident when it comes to chosing the correct test. Since the data are far from being normally distributed, I'd say that I can't use the regular simple hypothesis test, because it relies on a t-distribution. If I take the log of the series, the data will be more normally distributed, but then I'd have to test the null that the mean is equal to log(0) and that's minus infinity, right? Therefore that options seems to be ruled out as well. I'd be really grateful, if someone could help me out with a nonparametric test, that is capable of testing the null that the mean of my series is equal to zero.
Thanks a lot.
I have a dataset, that is definitely not normally distributed. Here is the workfile containing the series xyz:
I would like to test the null that the mean is equal to zero. Unfortunately I am not a statistician and not too confident when it comes to chosing the correct test. Since the data are far from being normally distributed, I'd say that I can't use the regular simple hypothesis test, because it relies on a t-distribution. If I take the log of the series, the data will be more normally distributed, but then I'd have to test the null that the mean is equal to log(0) and that's minus infinity, right? Therefore that options seems to be ruled out as well. I'd be really grateful, if someone could help me out with a nonparametric test, that is capable of testing the null that the mean of my series is equal to zero.
Thanks a lot.