LISTSERV at the University of Georgia
Menubar Imagemap
Home Browse Manage Request Manuals Register
Previous messageNext messagePrevious in topicNext in topicPrevious by same authorNext by same authorPrevious page (November 2006, week 4)Back to main SAS-L pageJoin or leave SAS-L (or change settings)ReplyPost a new messageSearchProportional fontNon-proportional font
Date:   Tue, 28 Nov 2006 10:06:33 -0800
Reply-To:   skoob <abhiscooby@GMAIL.COM>
Sender:   "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:   skoob <abhiscooby@GMAIL.COM>
Organization:   http://groups.google.com
Subject:   Need Help for correlation analysis
Comments:   To: sas-l@uga.edu
Content-Type:   text/plain; charset="iso-8859-1"

Hi,

I am trying to examine association (correlation) between 9 predictor variables and one dependent variable using PROC CORR. My sample size is 31. the data has been obtained from observations.

The predictor variables are discretely distributed and the dependent variable is continuously distributed (but not normal according to Shapiro Wilk Test) Because of this I can't use Pearson's Correlation Coefficient. I have 2 questions.

1. Is it ok to use Spearman's Correlation Coefficient? Is there any better method of analysis?

2. If I want to do partial correlation analysis and my code looks like this:

PROC CORR SPEARMAN; VAR ALN DESIGN_COST; PARTIAL PC REGR PTE PTT TI FF AS ASB; RUN;

My interpretation of the code is that I will get the correlation between DESIGN_COST and ALN while controlling other variables (PC REGR PTE PTT TI FF AS ASB).

Is my interpretation correct?

Thanks in Advance.

Abhijeet


Back to: Top of message | Previous page | Main SAS-L page