Statistics and scientific journal articles ;Let us congratulate Nature

[ksparth@vsnl.com]

The magazine Nature deserves congratulations for the approach they followed in tackling statistics related issues. Improper use of statistical methods in medical journals may have disastrous results.

 

Several years ago the British Medical Journal(BMJ) did examine the statistical  infirmities of the papers published in that journal.The results were revealing. In some countries Biostatistics as a discipline does not exist.Many papers published in journals from such countries may contain wrong conclusions as they did not use valid statistical analysis.Bogus claims of cure by medications when none existed is an instance in point.



After reading the editorial in BMJ, I wrote an note in the Medical Physics Bulletin of the Association of Medical Physicists of India (It is currently a Journal)requesting medical physicists to learn statistical methods to assist their medical colleagues while drafting papers.Medical physicists may be readily available in many centres. 



In many journals we see extremes.In the first category,there will not be any hint that they used stistically valid procedures to analyse the data. The second category will have too much references to the statistical methods employed. In the latter, a biostatistician is likely to be a co-author!



Yet another problem is the tendency by lay press to run stories based on articles containing bogus data. Not that they do it intentionally.With the spread of electronic media incorrect message is brought to every drawing room.



I hope that other journals will follow the steps which Nature took



WISHING ALL THE MEMBERS OF THE GROUP A HAPPY NEW YEAR



K.S.Parthasarathy





************************************************************************

You are currently subscribed to the Radsafe mailing list. To

unsubscribe, send an e-mail to Majordomo@list.vanderbilt.edu  Put the

text "unsubscribe radsafe" (no quote marks) in the body of the e-mail,

with no subject line. You can view the Radsafe archives at

http://www.vanderbilt.edu/radsafe/