[ Date Index ] [ Thread Index ] [ <= Previous by date / thread ] [ Next by date / thread => ]
On Sat, Dec 19, 2009 at 04:54:01PM +0000, Neil Williams wrote: > > Placebo is just another name for operator bias. Bias occurs whenever > anyone involved in making the judgement have any knowledge of what has > changed between the various times at which a judgement is made. Bias is > unavoidable; you can only compensate for it, never avoid it. > > Placebo != imagination > Bias != imagination > > The placebo effect and operator bias are absolutely real effects, > statistically measurable and completely unrelated to imagination. > > Science has proved again and again that if the judgement is made with > knowledge of when the change was made, the judgement will reinforce the > beliefs of those making the judgement with a statistically significant > effect. > > Double-blind randomised trials are the only rigorous method of > compensating for the ever present placebo effect, no amount of bluff > and blustering will change the reality. > See http://en.wikipedia.org/wiki/Placebo http://www.bbc.co.uk/science/horizon/2002/homeopathy.shtml Getting a truly double blind randomised result that is statistically significant takes a great deal of effort. Another example of this "placebo" effect is in finance and is why the FSA insist on a clause such as "past performance is no guide to future returns". http://www.moneymadeclear.fsa.gov.uk/products/investments/types/pooled_investments.html -- Henry Photocopies or faxes of my signature are not binding. This email has been signed with an electronic signature in accordance with subsection 7(3) of the Electronic Communications Act 2000. Digital Key Signature: GPG RSA 0xFB447AA1 Sat Dec 19 17:30:38 GMT 2009
Attachment:
signature.asc
Description: Digital signature
-- The Mailing List for the Devon & Cornwall LUG http://mailman.dclug.org.uk/listinfo/list FAQ: http://www.dcglug.org.uk/linux_adm/list-faq.html