[Jdm-society] Suggestions for data analysis approach

[Norm Constantine]

>>>My colleague and I have some data to analyze, and are looking for advice on appropriate approach. 

I've been hearing this sort of question from students (and colleagues) for more than 25 years, I'm amazed at the cross-disciplinary, cross-generational, and timeless consistency of this fundamental misunderstanding of the research process. My answer has not changed since 1982 - tell me first your research question and then we can talk about strengths and weaknesses of potential analyses to best address that question within the limitations of the data and resources available. Over the years I've come to elaborate a bit further, and make the point whenever I can (preferably before this question is popped) that "analysis methods exist to serve your research question, never let them drive your research." The same can be said about data, and applies to exploratory as well as explanatory research. 
-Norm

--
Norman A. Constantine, Ph.D.
Clinical Professor, School of Public Health 
University of California, Berkeley
Director, Center for Research on Adolescent Health and Development 
Public Health Institute
Phone: 925.284.8118  Fax: 925.284.8107
http://sph.berkeley.edu/faculty/constantine.html
http://crahd.phi.org/

 
----- Original Message ----- 
From: Hamm, Robert M. (HSC) 
To: jdm-society at mail.sjdm.org 
Cc: Showers, Carolin J. ; brunswik at yahoogroups.com 
Sent: Friday, May 25, 2007 12:48 PM
Subject: [Jdm-society] Suggestions for data analysis approach


My colleague and I have some data to analyze, and are looking for advice on appropriate approach. 

 

160 students in groups of 20. 

 

In each group, each student evaluated him or her self ("Self") and each of the 19 fellow students ("Peer"). 

 

Evaluation was done on 7 dimensions (Altruism, Leadership, and other virtues), on a 1 to 4 categorical scale. 

 

How analyze this? I can easily construct comparisons of average peer versus self judgments on each of the 7 dimensions. But I have the sense that this sort of data has been well explored, and there are lots of other interesting features. Where should I look? 

 

Thanks

 

Rob

 

Robert M. Hamm, PhD

Director, Clinical Decision Making Program

Professor, Department of Family and Preventive Medicine

University of Oklahoma Health Sciences Center

900 NE 10th Street, Oklahoma City OK 73104

405 271 5362 ext 32306

robert-hamm at ouhsc.edu

http://www.fammed.ouhsc.edu/robhamm/index.htm 

 



--------------------------------------------------------------------------------


_______________________________________________
Jdm-society mailing list
Jdm-society at mail.sjdm.org
http://www.sjdm.org/mailman/listinfo/jdm-society
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.sjdm.org/mail-archive/jdm-society/attachments/20070529/b45fd3cd/attachment.htm