jma
Member/Review Board Member
Posts: 5
|
Post by jma on Jun 4, 2014 4:27:39 GMT -5
I stumbled upon this artice on the "Less Wrong" blog: Parapsychology: the control group for science And this one expanded on it about Bem's last meta-analysis: The Control Group Is Out Of ControlOf course, the reasoning is obviously problematic: -> parapsychology gets positive results -> we know for a fact that psi doesn't exist -> so that means there is a problem with the scientific method as we currently practice it. But beyond that obvious reasoning problem (how can we know that the second premise is correct or not?), I must say that I do think about that issue quite a bit. I'm not certain that it's the case, like the blogpost writer seems to be. But it's a facinating epistemological and methodological question. Specifically, I do ponder (for example) how much can we trust meta-analysis or what's the point of bayesian analysis if it just includes the scientist prior belief "into the equation". Anyway, just thought I would share those finds.
|
|
|
Post by salexanderhardison on Jun 11, 2014 22:08:21 GMT -5
Hi again,
I have to confess a certain degree of ignorance regarding the recent MA. A friend has a good knowledge of it and has spoken with Michael Duggan about it. Personally, I'm not as well acquainted with the intricacies of it as I might like to be. The issue of pre-registering studies into a system, to me, seems to be one which should be implemented far more frequently in parapsychology (and, for that matter, other sciences) and that the MA failed to do that, likely, fails to grant us an ideal sample of studies -even if Bem and the co-authors contested this. I would have to do a more detailed analysis of the results of the MA (and the included and potentially non-included studies) before I could have a sufficient opinion on it.
I will say that a great many people have commented on the failures within the social sciences since the Feeling the Future study was originally published. That it took a successful study on precognition to bring those issues to wider public awareness immediately gives your point(s) some degree of weight -that people are now questioning scientific methodology, in general, is certainly fascinating.
P.S.- I'm also not convinced of the absolute relevance of Bayesian statistics in parapsychology, or even in experiments from other fields.
|
|