| Date: | Wed, 16 Nov 2011 07:57:20 -0500 |
|---|
| Reply-To: | Art@DrKendall.org |
|---|
| Sender: | "SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU> |
|---|
| From: | Art Kendall <Art@DrKendall.org> |
|---|
| Organization: | Social Research Consultants |
|---|
| Subject: | Re: |
|---|
|
| In-Reply-To: | <1321437763.79894.YahooMailNeo@web77904.mail.sg1.yahoo.com> |
|---|
| Content-type: | text/html; charset=ISO-8859-1 |
|---|
<html>
<head>
<meta content="text/html; charset=ISO-8859-1"
http-equiv="Content-Type">
</head>
<body text="#000000" bgcolor="#FFFFFF">
<font size="+1">What did you use as a stopping rule?<br>
Why did you use promax? Were you not interested in divergent
validity?<br>
<br>
That a set of variables be close to uncorrelated it a desirable
property when you are going to use them as predictors in a GLM or
clustering?<br>
<br>
<br>
<br>
Art Kendall<br>
Social Research Consultants<br>
</font><br>
On 11/16/2011 5:02 AM, Eins Bernardo wrote:
<blockquote
cite="mid:1321437763.79894.YahooMailNeo@web77904.mail.sg1.yahoo.com"
type="cite">
<div style="color:#000; background-color:#fff; font-family:times
new roman, new york, times, serif;font-size:12pt">
<div>Dear All,</div>
<div><br>
</div>
<div>I used Principal Axis Factoring using promax method in
conducting EFA for the 81 items that utilized six-point
ordinal scale. The sample was n=381. There is no indication
of severe skewness on the data (skewness <3, kurtusis
<10 and mardia coefficients >1000). I used
commonalities and factor loadings as criteria of dropping
items. Items with commonalities of <.40 were dropped.
Items with factor loadings of <.32 were also dropped.
Crossloadings items were also dropped. Finally, 35 items were
left which loaded to six interpretable correlated factors.
The factors have the following number of items: 10, 7, 8, 4, 3
and 3. After the factor analysis, the reliability
coefficients were computed for each factor. The Cronbach
alpha are quite high.</div>
<div><br>
</div>
<div>After the EFA, a CFA was conducted using a separate sample
of n=500 using amos. Unfortunately, the chiquare has zero
pvalue and no one of the fit indices were acceptable. I tried
to improve the model (guided by the modification indices). I
found out that the fit (at least the fit indices such as
RMSEA, SRMR, cmin/df) of the model improved when I correlated
the residuals/error terms. <span style="font-weight: bold;">Question:</span>Is
it appropriate to correlate the error terms?</div>
<div><br>
</div>
<div>Thank you in advance for your comments.<br>
</div>
<div><br>
</div>
<div>Eins</div>
</div>
</blockquote>
</body>
</html>
=====================
To manage your subscription to SPSSX-L, send a message to
LISTSERV@LISTSERV.UGA.EDU (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
|
