A recent study by Marica Ferri and her co-workers at the Italian Agency for Public Health (hey, my compatriots actually manage to do some newsworthy research, occasionally!) have compared several so-called 12-step programs for recovery from alcoholism, including the famous (or infamous, depending on who you ask) “Alcoholics Anonymous,” which mixes group therapy with Christian prayer. The results, published in the journal The Cochrane Library, are clear: it makes no difference which program one enrolls in, one has about the same (not very high) chance of returning to sobriety for good.
The news is controversial, of course, because of AA’s religious underpinning. If Ferri et al. had simply compared the effectiveness of several drugs to treat, say, the common headache, everyone would be happy: it doesn’t matter if you take Tylenol or aspirin, but it does help if you take either one. Let the slickest TV ad make the difference between the two brands. But with AA the implicit comparison (though Ferri and colleagues wisely stayed away from it in print) is between Jesus and rehab secular therapy. Apparently, a draw ain’t good enough for Jesus.
Indeed, this is a particular example of the whole “power of prayer” scientific and theological idiocy that has occasionally surfaced even in professional medical journals (and has been thoroughly debunked). Whenever effects of prayer or other religious mumbo-jumbo have been alleged, they have easily been explained either by placebo effects or by small random statistical fluctuations (in which case they will not be repeatable in a follow-up study). It is funny, in a sad way, to see religious people haggling over issues of marginal statistical significance, because they don’t seem to realize that by engaging in such exercises they are implicitly admitting to severe limits to the power of their allegedly all-powerful god. I mean, if finding Jesus could really rescue you from alcoholism, wouldn’t the outcome be clear and indisputable? Since when miracles have a probability level so close to that of random events that a professional statistician can’t tell the difference?