Is GC seriously trying to suggest that as soon as one study ever has found a positive correlation then it is the “correct” study and every other study is untrustworthy?
I think so. That is a ridiculously asymmetrical burden of proof. And there's the essential symmetry which belies all non-truth-seeking behaviors: I could make the same argument GC did, with just as much force behind it (if not more). Therefore, wrong statements can be supported with this argument just as easily as right statements. This means that "this argument supports my statement" isn't meaningful evidence toward the statement being true.
No, I never said the authority of 1 positive study was not rather low. The authority of positive 100 studies is considerably greater than that of 1 study. However the number of negative studies has no bearing on the situation, simply the relative number of positive studies.
1 Positive Study: Not much authority
100 Positive Studies: A lot more authority.
1000 Positive Studies: Even more authority
This perhaps *the* most important principle there is. Rejecting this principle will literally kill people and that is because a general claim
"Poison X kills people" reduces to a number of hypothesis as to
"How does Poison X kill people?" The number of ways that Poison X kills people are potentially vast and likely only one of these claims is true.
The methodology of the study will cover only a limited number of these hypothesis. That means if I have a lot of money to launch a lot of studies *and* I have an interest in continuing to poison everyone else for whatever reason all I have to do so, all I do is keep on testing a specific false model as to how the poison works for a very long time. The general claim (this is poison) is important, how it works precisely is less important.
This is not to say that the results of negative studies are unimportant. They are important because their methodology can be analysed to establish how the poison doesn't work. The poisonous of the substance in general however, once established by positive result studies is not disputed by the existence of negative result studies and that is to all practical considerations what matters.