Date:         Mon, 19 Jul 1999 21:40:28 -0400
Reply-To:     Amanda Parriag <amandaparriag@YAHOO.COM>
Sender:       "SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU>
From:         Amanda Parriag <amandaparriag@YAHOO.COM>
Subject:      questions on non-parametric data
Content-Type: text/plain; charset=us-ascii

I have two questions.

(1)When you do the marginal homogeneity test, which
statistic (the observed MH statistic or the std. MH
statistic I am guessing are most likely) do you report
in your results section and how?  I have never used
this statistic before, nor seen it referred to in
papers.

(2)I want to find two sets of correlations between
three variables.  One variable is a frequency count,
one is a nominal variable and the other is ordinal.  I
need correlations between the frequency count and the
nominal variable and a separate correlation between
the frequency count and the ordinal variable.  I
computed regular Pearson correlations so far.  I would
like to know if I can do Pearsons (any kind of
reference to back up a yes answer would be
appreciated), and if not, should I be doing two
different types of correlation, a Spearman's or a
Kendall's tau_b?  Or even an Eta?

Thank you.

Amanda
_________________________________________________________
Do You Yahoo!?
Get your free @yahoo.com address at http://mail.yahoo.com