View Single Post
  #51 (permalink)  
Old 14th January 2008, 12:55 AM
BillyJoe BillyJoe is offline
Senior Member
 
Join Date: Jan 2008
Location: Mooroolbark
Posts: 308
BillyJoe is on a distinguished road
Default

Quote:
Originally Posted by Acleron View Post
BillyJoe, your protocol should work with one proviso.

The two groups, treated and placebo, will have to be matched. For example, it would not be a fair test if the treatment group contained a smaller proportion of self-limiting conditions. Another problem with ill-matched groups would be the duration of treatment and end-point of the test for success or failure.

I have answered this before, but here is a view from that blog referenced by colmcq. It supports my original response to you:

Quote:
Given a large enough sample of the population, splitting this group in two, in a random manner, will produce two groups that are indistinguishable by any factor you choose to measure.
Do you disagree with this?
In any case, I reiterate that the first priority with any trial of homoeopathy is that the homoeopath/patient interaction must not be interfered with.
Reply With Quote