How much confidence can we place in clinical evidence?
Rebecca Aris interviews Robert Matthews
Aston University
Robert Matthews clears up the misconception around terms such as ‘2 sigma’ and ‘5 sigma’ and explains how significance can impact of the credibility of clinical evidence.
Robert Matthews is Visiting Reader in Science at Aston University, Birmingham. He currently focuses on using techniques known as Bayesian methods to try and make better sense of new research findings in drugs trials and epidemiological studies, allowing new results to be put into a better context.
Robert is concerned that the misconception of sigma levels can impact on the credibility of clinical evidence.
He speaks with us here on why he would expect more replicable results in medicine and why there is confusion around terms such as ‘2 sigma’ and ‘5 sigma’ and he explains what they really mean.
To listen to the full interview, please click on the play button below.
Quicklinks
00:15 – The definition of sigma.
01:28 – Why Robert thinks there is confusion around concepts such as 2 sigma and 5 sigma.
06:01 – Why it is worrying that scientists regard 2 sigma levels of evidence as representing a publishable result.
06:51 – In medicine, why you would expect the failure rate in replicating results in clinical data to be higher and impact this has on the credibility of clinical evidence.
09:36 – What can be done to put right the misconception of significance.
11:06 – How we should change our approach to assessing evidence to ensure that it is more credible?
Subscribe to pharmaphorum on YouTube to be the first to hear about new videos - http://www.youtube.com/pharmaphorum.
About the interviewee:
Robert A. J. Matthews is Visiting Reader in Science at Aston University, Birmingham, UK. He read physics at Oxford in 1978.
He has written a few books including, ‘25 BIG IDEAS: The Science that's Changing our World’ and ‘Q&,A: Cosmic Conundrums and Everyday Mysteries of Science’. He was recently interviewed by radio 4.
He can be reached at rajm@physics.org.
What can be done to put right the misconception of sigma?
