The question:
Are there any philosophers or sociologists who have written about the process in which the scientific community decides that a body of evidence is enough to consider a particular theory true (or at least highly likely, almost indubitable)?
Some background to my question (I'm afraid it got a bit long):
I am a researcher in psychiatry and I've been thinking about how a certain theory or hypothesis comes to be considered as true or proven among scientists. I have read Thomas Kuhn's Structure of Scientific Revolutions and his view seems to be that what hypothesis is considered the truth is not entirely based on empirical data, and it is to a considerable degree a social process in the community of scientists.
This is a good basis which aligns with my experiences, but I'd like to read some works that focus on this particular question. From my own readings of the scientific literature, psychiatry is riddled with hypotheses that the community of psychiatric researchers more or less accept as true (however, there are always some dissidents) even though the actual data do not give a black and white answer. Some degree of interpretation is thus necessary, but researchers seldom seem to be aware of this. Some such questions in psychiatry are: do SSRI's (a class of antidepressant drugs) actually alleviate depressive symptoms (mainstream view: yes)? Does cannabis cause psychosis (mainstream view: yes)? Does treatment with antipsychotic drugs reduce the progression of the psychotic illness over long time (mainstream view: yes)? Does the environment in the home, i.e. parental behaviors etc, have any role in the development of psychiatric illness (mainstream view: no)?
In all of these cases, the empirical data alone are in my opinion insufficient to give an answer to the question. We have to rely on assumptions, inferences and interpretations to arrive at an answer. There are of course many more such questions in psychiatry, and I suspect that this is common in most other scientific areas as well. I will give a concrete example below, though not from psychiatry.
Do homeopathic treatments work? ("work" in this context means work better than placebo). The scientific community at large is confident in saying no, which is a reasonable position. There are in fact a number of randomized controlled studies showing effects of homeopathic treatments in various conditions, though the more well-designed the studies are, the less likely they are to find an effect (this indicates possible bias). Furthermore, there is no plausible way in which the homeopathic treatments could work, given what is currently known about physics, chemistry and biology. This means that if we are to accept the positive RCT's, we also have to accept that there is a yet to be discovered type of energy transmission or something similar, and this seems less likely. The reasonable conclusion is that the positive RCT's are probably due to various biases (including fraud) and that homeopathic treatments don't work.
However, this is clearly not based on a Popperian notion of falsification. That some RCT's fail to find an effect does not falsify the hypothesis, or else we would have to accept that, for instance, SSRI's don't work because there are many trials with negative results. And just like for homeopathic drugs, the incentive for publication bias and some kinds of fraud is enormous with such drugs, sold by pharmaceutical companies and marketed as effective for conditions that affect large numbers of people (depression, anxiety), there is no real difference in the possibility of bias. And the history of medicine is full of discoveries of treatments and preventive measures that have been found to work even though there is no known mechanism that could explain how they work, so that is not an absolute argument against a hypothesis.
So this seems like a scenario that Thomas Kuhn described; the data is insufficient to determine what is true, and instead the community of researchers, one way or another, agrees upon what we should consider as true. In this particular case, I think it's correct to conclude that homeopathic treatments don't work and that the positive RCT's are a result of bias or fraud (and thus that their use shouldn't be supported by state-funded health care, which is an example of why it matters), but I cannot say that this conclusion is the only possible interpretation of the data.
I suppose that the moral of all of this is that we cannot be absolutely certain about anything in science, a view that I think I support, but to avoid total relativism and to use science constructively, we nevertheless need to agree upon what counts as the leading scientific hypothesis and how strongly we should believe it. In my mind, this is what actually happens all the time, but how does this process work in detail? Are there any philosophers, or perhaps sociologists, that have studied this process in greater detail than Kuhn, who mostly hinted about this process.