I recently ran across an essay which apparently has had currency for some time among those fighting food faddists. It is about the prevalence of false findings in the majority of studies. Or, as it puts it "Moreover, for many current scientific fields, claimed research findings may often be simply accurate measures of the prevailing bias." This fits well with the rule I told my son when he was watch some food fad videos on YouTube. I told him to be careful whenever a video said "a study by..." as it is easy to find single studies proving almost anything. Single studies prove cold fusion works, cell phones cause cancer (and are completely safe), transfats are poisons (and safe), sugar causes hyperactivity (despite countless studies debunking this) and so on.
I do not have the time right now to write a full essay on this topic, but I thought I would bring it to the attention of any readers happening by, as it is interesting to consider in light of the claim 95% of scientists support AGW. If that is true, it still may be, as suggested here, that individual studies may not be so much accurate as reflections of the prevailing orthodoxy. I have suggested this before, though blamed it more on publication bias, funding bias and personal bias of those entering the field. This essay supports the fact study design and sampling may perform the same task. So, I will be revisiting this when time allows.
UPDATE: It is especially interesting to consider this essay in combination with another on the many societal reasons for orthodoxy predominating in the climate field.