I'm not saying "one" could not participate. I'm saying AA is not a good candidate to do the study. I'm also not sure how a "double blind" protocol would work.
The AA organization certainly could be the funder for a study such as this, as I already explained. As to a double-blind study, you have a point there.
OK. AA picks 2,000 of those people. Why are they "lining up around the block"? Do you not see the possibility of selection bias? Are half of them going to be told to quit coming? AA isn't going to tell them that, or track the progress of those who leave. And if x percent say it worked for them, maybe they're wrong, it was just a coincidence.
Sure, there's always the potential for bias; a proper study takes as much of that into consideration as possible. It's not like sociology is breaking new ground here when it comes to studying the effects of alcoholism.
With a large enough sample size, many of these issues are also controlled for.
With a survey-type study, you have the participants take a survey every month asking specific questions about what their treatment is like or not like; what their attendance is, what their other activities are, whether or not they had a drink, and similar. A study doesn't need to have to tell participants to not attend. And they most certainly can track people who stop attending AA for whatever reason. They are still a part of the study after all. If a study group chooses to use inducements (like paying the participants for filling out the surveys) then they are still eligible as long as they truthfully report what's going on.
Studies also take into consideration a certain amount of people dropping out of the survey over time. The larger the number of participants, the more accurate the results are which can be more accurately extrapolated to the larger group.
Has anyone said such inducements are unethical?
No; just anticipating a possible objection.
So you have 2,000 people already in AA, before the study began; and a baseline group of non-alcoholics - say, 2,000 of them. That's $1 million a year, without other expenses added. What's a reasonable amount for AA to spend for a survey that doesn't prove anything? And wouldn't it be better to have 2,000 alcoholics who promise they won't go to AA? Understand, AA won't take their names and won't turn them away.
Doesn't prove anything? Maybe to you it won't or doesn't. To people who want to study
if AA works and
how it works, then it's of a great deal of importance.
Anyone in AA can join any research program.
Yup. That's exactly my point and has been for quite a while now, 11th tradition (and now the sixth apparently) notwithstanding.
You've gone from saying AA could conduct the study to saying AA members could participate in the study. Those aren't trivial differences.
The original complaint as I've understood it is that AA as an organization couldn't possibly perform a study. I've shown evidence that it certainly can. I've also shown that AAers themselves can participate in alcoholism studies and remain safely anonymous without disemboweling the 12 Traditions in the process.
Folks: Google "Does AA work" for a Scientific American link that has been posted at least twice, and for opinions of people who don't like AA. Re: to see what aspects of AA are useful and which are not. Show me what that would look like. Some people could attend but not work the steps? Some people would not be allowed to socialize? Some people would not be allowed to pray and others would be required to pray?
Sure. All that can be included in a survey questionnaire, and more. Again, depending on the type of research you're doing, there doesn't have to be a "you can pray, but you over there, you can't" thing going on. You said you read the studies I linked to earlier about the court-ordered guys, yes? In that study, they make it quite clear that they measured four groups -- one group went to AA, another group went to group therapy, a third group went to professional therapists and a fourth group didn't go to any of those. It's part of the scientific process Formulate an hypothesis, measure the data and see if the hypothesis has any predictive value.
I don't want to keep repeating myself or beat this into the ground ... and I can't retrieve the last post to find out if I made the point.
AA is decentralized to a degree that outsiders may not realize. It's not an efficient top-down research organization dedicated to funding studies, discovering the best cutting-edge methods of treating alcoholism then regularly handing out format directives from on high. That's not part of its DNA as the cliche goes.
And some other key point, which I've forgotten.
I would argue that AA is centralized to a larger degree that most people -- especially in AA -- aren't aware of.
Until TinyAl (I apologize if it wasn't him... this has been a very long thread!) posted the reported salaries of the top people in the AA World Group, I had no idea. And I've been around AA for the past twenty-odd years with my active participation for several years. I was frankly surprised to find out that the meetings that passed a few dollars in here and there, bowl out every half-time during a meeting that
I was familiar with, could translate to hundred thousand dollars plus salaries for multiple people.
When 90% of treatment agencies in the US are 12-step based, for example, and with courts across the country mandating AA treatment, why is it such a hard thing to ask for some actual proof of it's success?
So.
Any evidence towards disproving the null hypothesis yet?