Date:         Fri, 9 Nov 2001 09:48:24 -0500
Reply-To:     Lisa Montiel <montiell@ROCKINST.ORG>
Sender:       "SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU>
From:         Lisa Montiel <montiell@ROCKINST.ORG>
Subject:      Re: factor analysis: polarity of items
In-Reply-To:  <Pine.A41.4.21L1.0111081715580.27780-100000@login6.isis.unc.edu>
Content-Type: text/plain; charset="iso-8859-1"

I am working on a factor analysis that includes questionnaire items that are
positively and negatively worded regarding a person's sense of identity with
their neighborhood and the negative loadings are actually easier to
interpret (than if I reversed the polarity) and to explain to a
nonstatistical audience.

-Lisa

> -----Original Message-----
> From: SPSSX(r) Discussion [mailto:SPSSX-L@LISTSERV.UGA.EDU]On Behalf Of
> John Painter
> Sent: Thursday, November 08, 2001 5:20 PM
> To: SPSSX-L@LISTSERV.UGA.EDU
> Subject: Re: factor analysis: polarity of items
>
>
> Strictly speaking there is nothing wrong with negative factor loadings,
> they can lead to interesting interpretations. For example a factor
> representing depression may indicate that questions regarding sadness are
> negatively related to depression--perhaps evidence that sadness is not the
> same as depression?
>
> On the other hand, if your items address the same conceptbut some items
> are negatively worded then it may be useful to reverse the polarity of
> those items to simplify the interpretation (but again, there is nothing
> wrong with those items having negative factor loadings).
>
> hth,
> John
>
> On Thu, 8 Nov 2001, Christian F.G. Schendera wrote:
>
> > when performing a factor analysis, do I have to "turn" the
> > polarity of the items into one direction?! I know I should at
> reliability
> > analysis, but regarding FA I am not sure right now.
> > TIA, Chris
> >