RolandRB
Rest in Peace
You still aren't addressing my point. The fact that there IS a placebo effect, at all, is boggling. The fact that it is expected by statisticians is irrelevant: it just shows they are used to the placebo being effective. You seem to be inferring that the placebo effect is irrelevant, that it is essentially just unaccounted for "other variables" or some such. Perhaps it is. Perhaps it isn't: perhaps it is the person's belief that they will get better that accounts for the placebo effect (and they do actually get better). Perhaps this is also a component of the effectiveness of the drug, as well, and perhaps this is why most drugs' effectiveness tapers off over the years, as the hype surrounding it wears off.
I did address your point. That there IS a placebo effect is due to random chance and the probability distribution. It is not "boggling". It is a perfectly natural and expected outcome due to the nature of statistics. Overall, people will not get "better" on placebo. Some will -- most won't. To further explain, if a clinical trial is held across many sites, as it usually is, then drugs have succesfully gone to market where at some of these sites the placebo arm has done better that the trial drug arm. This happens at about one in six sites. Nothing "boggling" about it.
Statistical variations confuse many people, especially those who "want to believe" something. I saw an article once that a medic should not have published, but did, entitled "Is Laura cured of HIV?". While making no claims he pointed out that this person had been given anti-retroviral drugs followed by what was either an innoculation against HIV or a placebo and then this person had had no sign of HIV in their system for six months after being taken off all medication. Was this due to an active innoculation or did this person have placebo. They had been given placebo. I was the programmer on this trial and I saw the results after unblinding. Did the placebo somehow "cure" this person? No, this is just a normal outcome due to statistical chance and of no significance.