Page 1 of 1

T-test in Eviews

Posted: Mon Feb 25, 2013 1:23 am
by Student24
Hi

i have a question regarding the t-test for comparing the means of two independent samples:
I have a time series of stock returns with 3000 observations and 2 subgroups of firms for which I'd like to test the difference in means.
If I do this in Eviews with the function "tests by classification" ->"mean", do I first need to ensure that the populations are normally distributed and have equal standard deviations (as statistics books claim for most t-test methods)?
What if this is not the case, is there still a way to perform the t-test in Eviews?

Thanks for any help!

Re: T-test in Eviews

Posted: Mon Feb 25, 2013 8:51 am
by EViews Gareth
The Mean test for equality outputs a number of different test statistics, at least one of which allows for different variances.