Date:         Wed, 14 Sep 2005 12:16:03 -0400
Reply-To:     Dan Zetu <dzetu@hotmail.com>
Sender:       "SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU>
From:         Dan Zetu <dzetu@hotmail.com>
Subject:      Re: averaged scales vs factor scores
Comments: To: plancede@yahoo.com
In-Reply-To:  <20050914143357.152.qmail@web34503.mail.mud.yahoo.com>
Content-Type: text/plain; format=flowed

Hi:

The problem with factor scores is that they are not scaled, i.e. they have
always zero mean and 1 standard deviation. Plus, the methodologies employed
to derive the scores are not as robust as the ones used in deriving the
factor loadings.

The way people typically deal with this issue is what you actually
mentioned: they construct summated scales (which in fact most of the time
are simple averages) of the original values of the variables that load into
one specific factor. Of course, the summated scales will not be orthogonal
like factors (assuming you use an orthogonal rotation procedure like
varimax) but they do work pretty well most of the time.

Dan

<BLOCKQUOTE style='PADDING-LEFT: 5px; MARGIN-LEFT: 5px; BORDER-LEFT: #A0C6E5
2px solid; MARGIN-RIGHT: 0px'><font
style='FONT-SIZE:11px;FONT-FAMILY:tahoma,sans-serif'><hr color=#A0C6E5
size=1>
From:  <i>Plance Debouver &lt;plancede@yahoo.com&gt;</i><br>Reply-To:
<i>Plance Debouver &lt;plancede@yahoo.com&gt;</i><br>To:
<i>SPSSX-L@LISTSERV.UGA.EDU</i><br>Subject:  <i>averaged scales vs factor
scores</i><br>Date:  <i>Wed, 14 Sep 2005 07:33:57 -0700</i><br>&gt;I posted
this with no lucky. How do I turn factor<br>&gt;scores based on rotated
loadings of FA procedures<br>&gt;(scores obtained using regression
method)into useful<br>&gt;values for further analysis. I would like to
compare 5<br>&gt;groups of respondents.e.g nurses, doctors,
Lab<br>&gt;technicians, students.   I am told that factor scores<br>&gt;are
not comparable across studies (they are sample<br>&gt;specific) but summated
or averaged scores are.<br>&gt;<br>&gt;Any recommendation for types of
further analyses on<br>&gt;factor scores or summed or averaged scales would
be<br>&gt;appreciated.<br>&gt;<br>&gt;plance.<br>&gt;<br>&gt;<br>&gt;<br>&gt;__________________________________<br>&gt;Yahoo!
Mail - PC Magazine Editors' Choice
2005<br>&gt;http://mail.yahoo.com<br></font></BLOCKQUOTE>