GuyGadeboisLeRetour wrote:izentrop wrote:Janic wrote: most of which are at the ankle with BP which sprinkles them generously ...
What you will never understand is that a well-done study has enough locks that the funder has no influence on the results of the study.
What you refuse to understand despite all the evidence that has been thrown in here is that randomized studies are no better than observational
This is wrong, there are situations where observational studies are enough, and others where they are not enough, but randomized studies cannot give worse results than observational studies. It is impossible for an effective drug to show no effect in a randomized study, while the reverse is true, one can have the impression of an effect when there is none because of a bias. of the study.
This is why a result
negative in a randomized study is much more meaningful than a result
positive in a non-randomized study.
It is, moreover, general in all experience -
an absence of effect is much more significant than an observed effect. The reason is that for a bias to remove an effect, it must have exactly the opposite effect to compensate for it, no more and no less. And that is very improbable. While a bias can cause a positive effect to appear when there isn't, and that doesn't require a special bias value, so it's much more likely.
A concrete example ; experience
Michelson and Morleyfound an ABSENCE of effect of the movement of the Earth on the speed of light - the interference fringes were not moving. This is a very powerful result which has been confirmed by all the experiments and which has led to the theory of Relativity.
Conversely, the result
on superluminal neutrinos. was not very solid, because it was not an "absence of effect", it was just measuring a distance D, a time T, and doing the division D / T to compare to vs. Any error on D or T could give a result greater than c. And indeed it was, there was a faulty component. But it would have been extremely unlikely that an error would give, for example, "exactly c" if the actual speed had not been c.
Likewise, a randomized negative result is much more unlikely to arrive "by mistake" than a non-randomized positive result.
To pass for an idiot in the eyes of a fool is a gourmet pleasure. (Georges COURTELINE)
Mééé denies nui went to parties with 200 people and was not even sick moiiiiiii (Guignol des bois)