Comments on: Send Mathematics to the Rescue http://orbitchange.com/blog/2007/02/27/math/ Sharad Sharma examines the transformation challenges facing the software industry Tue, 14 Oct 2008 05:46:36 +0000 http://wordpress.org/?v=2.0.2 by: Sharad Sharma http://orbitchange.com/blog/2007/02/27/math/#comment-954 Tue, 06 Mar 2007 17:23:27 +0000 http://orbitchange.com/blog/2007/02/27/math/#comment-954 Good point, Ravi. I agree that the real impact of mathematics will be in the early drug discovery stage. Good point, Ravi. I agree that the real impact of mathematics will be in the early drug discovery stage.

]]> by: Ravi Aranke http://orbitchange.com/blog/2007/02/27/math/#comment-947 Tue, 06 Mar 2007 03:25:08 +0000 http://orbitchange.com/blog/2007/02/27/math/#comment-947 Sharad, I do not have high hopes for Mathematics (of the statistical analysis variety) improving state of practise in Medicine or Social sciences by an order of magnitude. There are just way too many variables - some known, many unknown - to have a controlled study. The techniques are around for decades and we have very little confidence in the stability of these studies. What is determined as statistically significant for one set of population, under one set of circumstances, is difficult to extrapolate to population at large. Co-relations are notoriously fickle and tell more about the experimenters' bias. You find what you seek. What is needed is understanding of the cause-and-effect relationships. This is where the DNA sequencing, gene analysis etc. comes in and this is where, IMO, mathematics is going to make huge impact. Regards, Ravi Sharad,

I do not have high hopes for Mathematics (of the statistical analysis variety) improving state of practise in Medicine or Social sciences by an order of magnitude. There are just way too many variables - some known, many unknown - to have a controlled study. The techniques are around for decades and we have very little confidence in the stability of these studies. What is determined as statistically significant for one set of population, under one set of circumstances, is difficult to extrapolate to population at large. Co-relations are notoriously fickle and tell more about the experimenters’ bias. You find what you seek.

What is needed is understanding of the cause-and-effect relationships. This is where the DNA sequencing, gene analysis etc. comes in and this is where, IMO, mathematics is going to make huge impact.

Regards,
Ravi

]]>