NOTE: Statistics and specific facts are not my strong suit. I am much better with ideas than facts, with the abstract than with the concrete, with the big picture than with the nitty-griitty details. I tend to miss the trees for the forest. So generally I would prefer to play to my strengths and avoid a post like this. But because I want to raise these concerns, I'll risk humiliation and bring them up anyway, and hopefully learn from whatever correction is offered to me.
I usually enjoy the Ontario Consultants' on Religious Tolerance site. Although they have a noticeable bias (who doesn't, anyway?), I find them useful for picking up links and tidbits of info. So I was somewhat surprised and disappointed to follow a link from an online discussion to their (in my opinion) misleading attempt to come up with estimates of orientation change from existing studies.
Please understand that I don't doubt that attraction change currently may be rare. I'm not convinced that the common approaches are very successful in general. So for me this is not about defending reparative therapy or a particular "success rate." I just think that honesty, accuracy, and clarity are of value.
There's more I could say about this essay, but I want to focus on the numbers, because they're what jumped out at me. I understand that the author (B.A. Robinson) acknowledges that the figures he comes up with are "crude" estimates, blaming this on the deficiencies in the studies themselves, but I think they are more problematic than that.
1. Exodus International (1978)
Before discussing studies and success rates, Robinson initially defines conversion as being a change from homosexual orientation to heterosexual or bisexual orientation. I was pleased to see this, as it seems to me to be the most reasonable definition. (Of course, I am biased.)
But when he begins discussing the study of Exodus International in 1978, suddenly he raises the bar, and changes the definition of conversion to be a change to "exclusive heterosexuality". The psychiatrists doing the study felt that 3 of the 30 subjects studied had become exclusively heterosexual, and so Robinson says that only 3 out of 30 changed. This definition-switch seems a bit slippery to me.
Robinson then assumes that since those 30 subjects were chosen by Exodus staff to participate in the study (apparently out of 800 members at the time), that these must include all the most successful cases, and therefore these three must be the only three who changed in all Exodus ministries. Which is how he comes up with his 0.4% success rate (3/800). This, I think, is an unjustified assumption. He will make this assumption again with another study, on an even grander scale–assuming that because the subjects were selected by pro-exgay groups and not randomly chosen, that the study group must include all the "best" cases, all the cases of people who have changed. To me this seems almost as misguided as the opposite mistake of assuming that these groups of subjects are representative of all who have attempted change.
2. Shidlo and Schroeder
On to Shidlo and Schroeder, the next study for which Robinson offers his own conversion number. (I've griped a little about S&S elsewhere–here my intention is to discuss not them but OCRT's use of them.) He notes that 8 out of the 202 reported orientation change. But, seven of these eight were exgay counselors or leaders whose statements "may have been false" (OCRT). So Robinson interprets this as showing that only one out of 202 had changed, and therefore gives the stat of 0.5% (on another page he says the failure rate is 99.5%)
This strikes me as odd. Just because he thinks their statements "may have been false" because they are involved in the exgay movement, he counts them as failures? Perhaps their claims to success deserve a question mark next to them, but it seems bizarre to label them as failures. Not being confident that X is telling the truth is nowhere near the same thing as being confident that X is lying (or is self-deceived). So I think perhaps it would be more honest to say something like: "We would need to research further to ascertain the degree to which the other seven may have changed."
Again, in his discussion of Spitzer, just as in his discussion of the Exodus 1978 study, Robinson subtly switches his standards from his more modest initial definition of conversion being from homo to hetero or bisexual in one's attractions, to the much stricter standard of conversion requiring a change to exclusive heterosexuality, only counting as "successes" those 37 who had no more homosexual attractions, fantasies, etc. This seems dishonest to me. The impression the casual reader would get, given Robinson's initial definition, is that the success rates that Robinson prints in bold are the percentages of those who experience any significant attraction shift whatsoever. Robinson gets the smaller numbers (that he is hoping for) by switching to a more demanding standard of success.
And again we see the number game from before repeated:
"The 46 subjects from NARTH might have been chosen as the most successful patients from as many as 250,000 individuals who entered therapy. Unfortunately, no data has been reported about the total number of persons from whom the 200 carefully selected patients were provided. Assuming that only 100,000 subjects were involved — a VERY conservative figure, then 37 "success stories" represents a conversion rate of 0.04%"
No thinking person in her right mind would honestly say that Spitzer's group was representative of all those who pursue an exgay path. But to suggest that it must include all of the "success stories" of NARTH and Exodus, as Robinson does, seems even more ridiculous. It would mean assuming that NARTH and Exodus were omniscient regarding the changes in the lives and attractions of everyone who passed through their ministry or practice. It would mean assuming that NARTH and Exodus were able to successfully contact all these people. (Which seems doubtful because if they've changed, they've probably been out of contact from the ministry or therapist for quite some time, as they wouldn't need them any more.) And it would mean assuming that all who changed attractions would want to take part in such a study. All of these assumptions are hard to accept.
Later, Robinson discusses a statement by Jack Drescher:
"Jack Drescher…:'There are probably a small number of people with some flexibility in their sexual identity who can change. Out of the hundreds of gay men I've treated, I've had one.' If we assume that his term "sexual identity" is a synonym for "sexual orientation," and that Dr. Drescher has treated 200 gay men, then he would seem to estimate that about 99.5% of gay men have a fixed sexual orientation, and that only about 0.5% can change their orientation. "
But this is a very different thing from a study of those who attempt to change. Presumably we are talking about an ordinary population of gay men, many of whom have probably never made a concerted effort to try to change in any way. So I don't see why one should think that these numbers meaningfully apply to those people who are trying to change, as Robinson does, suggesting that Drescher's statement supports his above-mentioned success rates in the neighborhood of 0.5% or less. Regardless of the degree to which change is possible, I doubt we can generally tell whether or not a person is capable of change if he or she isn't trying to change.
Again, as I stated at the outset, I do not profess to know how common attraction change is or what the success rates are. (Just like my assurance of salvation, on good days I'm pretty sure about one person at most. 🙂 ) For all I know, the figures OCRT came up with could be correct. So my objection is not to the numbers themselves, but to the way in which they were arrived at and the way in which they are being used, both of which seem quite dubious to me.