=========================================================================
Date:         Sat, 29 Jul 2006 09:46:03 -0600
Reply-To:     Stevan Nielsen <stevan_nielsen@byu.edu>
Sender:       "SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU>
From:         Stevan Nielsen <stevan_nielsen@byu.edu>
Subject:      Re: Listserv for AMOS?
Comments: To: "Feinstein, Zachary" <ZFeinstein@HarrisInteractive.com>
In-Reply-To:  <47009A583B6592479DBC0C6DCD40DF950227D8DE@mission.harris.harrisinteractive.com>
Content-Type: text/plain; charset="us-ascii"

I am not an AMOS expert, but obtaining correlations between latent
variables is an option in the graphics interface if you have modeled
covariance between the latent variables of interest.  It sounds as if
you have done this, because you speak of a covariance estimate.
Modeling covariance between the latent variables requires, in the
graphic interface (the only way I use AMOS) drawing a double-arrow arc
between any two latent variables of interest.  The default output is for
unstandardized coefficients.  Once a model is estimated the default
output shows unstandardized covariance and regression coefficients in
both the graphic of the model and in the printout (key F10).

To obtain standardized estimates select "Analysis properties" (the
abacus icon with six colored buttons or "Ctrl" key and "A"
simultaneously), then select the "Output" tab, then select by check mark
"Standardized estimates" from among the 15 options for output.  If you
have modeled the covariance between the latent constructs of interest,
you will now have correlation coefficients in the printout (key F10) and
an option for standardized estimates for the graphic of your estimated
model.

Stevan Lars Nielsen, Ph.D.
Clinical Professor
Clinical Psychologist
2518 WSC, BYU
Provo, UT 84602

801-422-3035; fax 801-422-0175

-----Original Message-----
From: SPSSX(r) Discussion [mailto:SPSSX-L@LISTSERV.UGA.EDU] On Behalf Of
Feinstein, Zachary
Sent: Friday, July 28, 2006 12:19 PM
To: SPSSX-L@LISTSERV.UGA.EDU
Subject: Listserv for AMOS?

Is there a listserv for AMOS?  Specifically I am hoping to make a post
about how to obtain the correlation between two latent constructs.  I
know I can get the covariance but am a little baffled by the correlation
in AMOS.  My hunch is that it is not possible as long as the latent
variable is not standardized.

Zachary