Date:         Wed, 7 Dec 2005 11:29:49 +0100
Reply-To:     Marta García-Granero
              <biostatistics@terra.es>
Sender:       "SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU>
From:         Marta García-Granero
              <biostatistics@terra.es>
Organization: Asesoría Bioestadística
Subject:      Re: Significance testing between R-square from linear and
              R-square              from curvilinear regression (Revised).
In-Reply-To:  <6250203B042D8349A920AA610383093692E4F3@UTHEVS3.mail.uthouston.edu>
Content-Type: text/plain; charset=ISO-8859-15

Hi:

Steiger's homepage:
http://www.unav.es/genetica/new/soft.html#misc

And Multicorr page:
http://www.interchange.ubc.ca/steiger/multi.htm

Anyway, the links in this page are wrong (they point to a
nonexisting ftp site). I found out that the correct links for
multicorr program & documentation are:
- Program: http://www.interchange.ubc.ca/steiger/multicor.zip
- Manual:  http://www.interchange.ubc.ca/steiger/multicor.pdf

HTH, Marta

SPR> Steiger's multicor program will do that.

SPR> "Want to compare multiple, partial, or canonical correlations for
SPR> equality?
SPR> Steiger and Browne (1984) show how you can do this by reprasing the
SPR> hypothesis as a pattern hypothesis testable by MULTICORR. For a recent
SPR> review of some related correlational tests, see Olkin and Finn (1995).
SPR> References
SPR> Olkin, I., and Finn, J.D. (1995). Correlations redux. Psychological
SPR> Bulletin, 118, 155-164.
SPR> Steiger, J.H. (1979). Multicorr: a computer program for fast, accurate,
SPR> small-sample tests of correlational pattern hypotheses. Educational and
SPR> Psychological Measurement, 39, 677-680.
SPR> Steiger, J.H. (1980). Tests for comparing elements of a correlation
SPR> matrix. Psychological Bulletin, 87, 245-251.
SPR> Steiger, J.H., & Browne, M.W. (1984). The comparison of interdependent
SPR> correlations between optimal linear composites. Psychometrika, 49,
SPR> 11-21. "


SPR> Paul R. Swank, Ph.D.
SPR> Professor, Developmental Pediatrics
SPR> Director of Research, Center for Improving the Readiness of Children for
SPR> Learning and Education (C.I.R.C.L.E.)
SPR> Medical School
SPR> UT Health Science Center at Houston

SPR> -----Original Message-----
SPR> From: SPSSX(r) Discussion [mailto:SPSSX-L@LISTSERV.UGA.EDU] On Behalf Of
SPR> Sungeun Chung
SPR> Sent: Tuesday, December 06, 2005 11:18 AM
SPR> To: SPSSX-L@LISTSERV.UGA.EDU
SPR> Subject: Significance testing between R-square from linear and R-square
SPR> from curvilinear regression (Revised).

SPR>  Hi, I want to clarify my question in previous question.

SPR> I have two variables,  X: time (11 time points) and Y: means of
SPR> evaluation (one group, N = 132).

SPR> I expect that exponential function will fit better than a linear
SPR> function to predict the trajectory.

SPR> I did a curve estimation with exponential function and found R-square
SPR> from the exponential  function is greater than R-square from a linear
SPR> function.
SPR> But how can I test if the difference in R-squares is signifcant?

SPR> I found a formular in Blalock (1972). But it requires correlation
SPR> between Y and transformed Y.

SPR> Should I do time series analysis?

SPR> Any help will be greately appreciated.  Thanks in advance,

SPR> Sungeun

SPR> -
SPR> Sungeun Chung, Ph.D.

SPR> Assistant Professor
SPR> Department of Communication
SPR> Memorial Hall 303G
SPR> Western Illinois University
SPR> Macomb, IL 61455-1390
SPR> Tel. (309) 298-2219
SPR> E-mail: S-Chung@wiu.edu; chseun21@yahoo.com

SPR> __________ Información de NOD32 1.1312 (20051205) __________

SPR> Este mensaje ha sido analizado con  NOD32 Antivirus System
SPR> http://www.nod32.com




--
Saludos,
 Marta                            mailto:biostatistics@terra.es