Date:         Thu, 13 Jul 2006 07:55:31 -0400
Reply-To:     Peter Flom <Flom@NDRI.ORG>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         Peter Flom <Flom@NDRI.ORG>
Subject:      Re: running ANOVA to test for differences
Comments: To: duklaa <duklaa@GMAIL.COM>
In-Reply-To:  <1152787354.352963.11360@i42g2000cwa.googlegroups.com>
Content-Type: text/plain; charset=US-ASCII

>>> duklaa <duklaa@GMAIL.COM> 7/13/2006 6:42 am >>> wrote
<<<
I have responses to a survey which asked a bunch of questions with
Likert scale responses. I want to make sure that the responses are all
different (i.e. the respondents didn't just click uniformly for all
questions to answer quickly).

How do I run ANOVA to test for this? I don't believe that the
responses
are normally distributed across the scale. Should I use GLM instead?

SNIP
>>>

Short answer, no not ANOVA.  No not GLM (not with any code).  Those
can't possibly answer your question.  You aren't testing for
differences

Longer answer:  If you want to see if a person answered the same for
every question, the best approach probably
depends on number of people and number of questions.  If neither is
large, then you can just use PROC PRINT and eyeball the data.
If either is large, then you probably want something in the data step.
I could cobble something together, but it's probably better to a) First
read
my even longer answer and b) Re-post your question to SAS-L with a
better subject, so the data step mavens can have a shot at it, if you
decide
that this is really what you want to do.  I would guess that some sort
of array would work.

Even longer answer:  You haven't really defined your question well. OK,
*some* people who answer quickly without thinking may choose the same
answer every time.  Some may not.  Some may click randomly.

What sorts of questions were these?  Were they ability questions, or
personality questions, or opinions, or what?
What was your sample size?  How many questions were there?

What you really want to look for is individuals who answered in ways
that don't make sense.  But how to best do this depends on the answers
to the above questions.
It won't be easy.  It may not be possible.  But you certainly do not
have a problem for ANOVA or GLM


HTH

Peter





Peter L. Flom, PhD
Assistant Director, Statistics and Data Analysis Core
Center for Drug Use and HIV Research
National Development and Research Institutes
71 W. 23rd St
http://cduhr.ndri.org
www.peterflom.com
New York, NY 10010
(212) 845-4485 (voice)
(917) 438-0894 (fax)