Re: Nonparametric in SPSS 19
Posted by
Martha Hewett on
Dec 07, 2011; 12:56am
URL: http://spssx-discussion.165.s1.nabble.com/Nonparametric-in-SPSS-19-tp5054005p5054093.html
I don't think most (any?) of the tests
would make sense with nominal data. A lot of non-par procedures are
based on the assumption that the variable is continuous, though typically
not normal. A lot of the tests create ordinal ranks, but the original
variable is continuous. People do use them for ordinal data and there
might even be some where that doesn't violate (or saliently violate) the
test's assumptions. I guess you would have to declare your variable
as scale. It doesn't make sense you can run a t test on an ordinal
var in SPSS but not a non-par test. Go figure.
| Martha Hewett
|
| Director of
Research | 612.335.5865
|
| Center for Energy
and Environment
|
| 212 Third Avenue
North, Suite 560 | Minneapolis, MN 55401
|
| (cell) 612.839.2358
| (fax) 612.335.5888 | www.mncee.org
|
From:
ericgraig <[hidden email]>
To:
[hidden email]
Date:
12/06/2011 06:46 PM
Subject:
Nonparametric
in SPSS 19
Sent by:
"SPSSX(r)
Discussion" <[hidden email]>
My understanding is that nonparametric tests are to
be used with nominal and
ordinal data. Yet SPSS's new Nonparametric choice on the "Analyze"
menu
seems to only accept scale data. Anyone know what's up with that?
Are we
supposed to just redefine our ordinal data to scale level or am I missing
something. Using the 'Legacy' choice I can play with ordinal data
without
any problems.
Advice?
Thanks.
Eric
[hidden email]
--
View this message in context: http://spssx-discussion.1045642.n5.nabble.com/Nonparametric-in-SPSS-19-tp5054005p5054005.html
Sent from the SPSSX Discussion mailing list archive at Nabble.com.
=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD