Date:         Thu, 29 Dec 2005 06:35:41 -0500
Reply-To:     Peter Flom <flom@NDRI.ORG>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         Peter Flom <flom@NDRI.ORG>
Subject:      Re: paramaetric and non-parametric test.
Comments: To: dokiya777@GMAIL.COM
Content-Type: text/plain; charset=US-ASCII

>>> Raj Datta <dokiya777@GMAIL.COM>  >>> wrote
<<<
I have 2 group of data with each having sample size of 75. Each group is
having the no. of stores and the tracking variable is the number of sales.
When i do the normality test, it fails. But than also i went ahead and
done t test with 2 groups and it turns out to be significant. But some
body told to me that if normality test has failed (though sample size is
>30) i should use Mann - Whiteney U test.

Please help which test to use. This is a problem pertaining to paramaetric
and non-parametric test.
>>>

Why is the tracking variable 'number of sales' as opposed to 'dollar value of sales' or something?
What question are you interested in (not what STATISTICS question, what substantive question?

It's very unclear what you are trying to model, and without knowing that, neither I nor anyone else
can give good advice.  If you write back with more details (to SAS-L, not just to me, please) then
someone here will probably be able to help you.


Peter