Reader Comments

Post a new comment on this article

Non-responders and self selection bias

Posted by plosmedicine on 31 Mar 2009 at 00:19 GMT

Author: Paul T Francis
Position: Associate Professor, Community Medicine
Institution: Amrita Institute of Medical Sciences, Cochin, India
Submitted Date: January 09, 2008
Published Date: January 10, 2008
This comment was originally posted as a “Reader Response” on the publication date indicated above. All Reader Responses are now available as comments.

This is in response to the posting by Amanda Cross (09 January 2008) on Non-responders. Since only 1 out of 7 eligible participants have responded the researchers should have found out the important characteristics of the non-responders to ensure that no major bias has crept into the data analysis. A few important characteristics of non-responders that could have affected the results are: 1. they could be younger 2. they might have lesser number of other risk factors for cancer 3. they might be having healthier life style and 4. for all what we know they might be eating more red meat than those who responded! The responders might have been more worried of a higher risk of cancer and hence resulted in self-selection. A random sample could have resulted in different conclusion. This is not to belittle their great effort at all. I feel that if the authors contact a sample of non-responders and collect data on the relevant variables they could strengthen the conclusions of this very important study.

No competing interests declared.