While researching which studies to discuss in the Effectiveness of Alcoholics Anonymous Wikipedia article, I went over to the Rational Wiki Alcoholics Anonymous article to see if I could add some peer reviewed information showing that AA is not effective from the other side.
Unfortunately, the Rational Wiki did not have any good science in it. The study they refer to with allegations that AA actually increases binge drinking has the ambiguous name "Brandsma et al. 1980" Research revealed that this supposedly scientific, albeit dated, survey showing how AA is worse than no treatment would have a DOI or PMCID, the way peer-reviewed studies do. It doesn't.
A Google search of the study showed the primary people referring it were unreliable anti-steppers (including, yes, people on the fringe of academia like Stanton Peele). Apparently, this study got put on the Orange Papers site, where the usual cadre of anti-steppers mirrored it from.
Finally, I got some solid information about this article in, of all places, an Amazon review. To summarize: This study was not peer reviewed, was published by one University Park Press, which mainly prints university textbooks -- read: not published in a journal, and its methodology is, at best, suspect.
It is referred to in journal papers, but usually negatively. For example, PMC2746426 mentions "concerns with the Brandsma trial which call its experimental results into question", and PMC3602358 says that studies like this are "significantly limited in their methods or interpretability," PMC3549307 (by the same authors) says the same thing.
Why is it that anti-steppers need to use a 35-year old non-peer reviewed study to find "science" to back up their anti-AA talking points? Because the real science shows that AA actually works.
Let me rephrase that: The objective science shows the people who go to more AA meetings are more likely to stay sober. The science has not yet concluded that it's AA's 12 step program that results in the statistically undeniable positive outcomes for AA members, but research is going in that direction.