Date: Thu, 13 Jul 2006 18:13:33 +0200
Reply-To: Karl Koch <TheRanger@gmx.net>
Sender: "SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU>
From: Karl Koch <TheRanger@gmx.net>
Subject: About Mann-Whitney Test...
Content-Type: text/plain; charset="iso-8859-1"
The Mann-Whitney assumes independent samples. In Andy Fields book  it also says that it assumes that datapoints are measured from different people - meaning that one person cannot partipate in both groups. I am not sure, if Andy Field here assumes that a datapoint always maps directory to a person (because of his background in Psychology). I find this a bit confusing and wouild like to ask you.
My datapoint are actually measures (on a 6-point scale) (which come from people but several from each person). Each person partipates in each of the two groups - from which I want to test stat. significance if they are different or not. The order with which people partipate in this two groups is counterballanced and should therefore be independent.
What do you think? Is this a case for Mann-Whitney then?
Der GMX SmartSurfer hilft bis zu 70% Ihrer Onlinekosten zu sparen!