Monday, May 28, 2012

Reparative Therapy Study Author Renounces His Work

Those committed to reparative therapy frequently point to the one published study that tried to suggest such therapy may have an effect.  This study, published by Dr. Robert Spitzer in 2003, interviewed 200 gay men and women who reported some "minimal change" in their sexual orientation lasting 5 years since reparative therapy. Now, the New York Times is reporting that Dr. Spitzer recently renounced the study (see here or here) .

They study's methodology was so problematic, it is stunning it was published.  First, the sample was entirely self-selected (i.e. people who had undergone reparative therapy and wanted to participate could volunteer), including ex-gay political advocates. This means the people who participated may have simply been those particularly motivated to convince themselves or others they were not gay, or those particularly motivated to believe reparative therapy works, or those of more ambiguous sexual orientation to begin with, etc.  (Even among this population, the changes reported were limited!) There was no random assignment to a therapy group and a control group, which is the hallmark of a proper experiment.

Worse, the study relied entirely on retrospective self-report: participants were asked, for example, how often they had desired someone of the same sex in the year before therapy--which was years prior at the time of the interview--and how much they had desired it in the past year. There is no guarantee whatsoever about the accuracy of their memories, or even that they were reporting the truth, which is particularly problematic amongst those so motivated to report change.

So, it is interesting to hear  that Dr. Spitzer himself has recently admitted to these problems and renounced the study.  Hopefully this study can be laid to rest in public discourse.

Wednesday, May 16, 2012

Study Finds Analytic Thinking Promotes Religious Disbelief

A research article published in Science recently examined the relationship between analytic thinking and religious belief. The authors tested the following idea: an intuitive cognitive system can often give rise to religious belief (see my series on this topic for some background). An analytic thinking system can override the intuitive system.  Therefore, analytic thinking may be associated with religious disbelief.

To test this hypothesis, the authors had study participants complete an analytic thinking task that requires overriding intuitions to reach the correct response, e.g. "If it takes 5 machines 5 min to make 5 widgets, how longs would it take 100 machines to make 100 widgets?"  (The intuitive response is 100; the correct answer, if you reason it out, is 5).  Participants then reported on their religiosity.  As expected, analytic thinking was negatively associated with religious belief--i.e., the individual tendency to override intuitions with analytic thinking and get the right answers was associated with religious disbelief.

Of course, this is only a correlational result. In a series of four follow-up experiments, the authors directly induced analytic mindsets in randomly assigned participants by using subtle, previously validated manipulations. For example, prior work has found that when people read a piece of text written in a difficult-to-read font, the extra effort and engagement leads to greater analytic thinking (for example, people show increased performance on logic puzzles). In the present study, participants were randomly assigned to fill out the same religiosity questionnaire in an easy-to-read font or a difficult-to-read font.  As predicted, completing the questionnaire with difficult-to-read font led participants to report lower levels of religious belief.

Or, for example, prior tests had found that simply viewing the statue The Thinker primes an analytic mindset in people, relative to viewing a control statue (again, they show increased performance on logic puzzles). When randomly assigned participants were primed with viewing The Thinker, they again reported lower religious belief than a control group. And so on with other methods used to induce an analytic mindset.

These results suggest that analytic thinking tends to override religious intuitions. The authors carefully caveat that their results do not comment on the inherent truths involved or what one should believe, and that this is just a descriptive account of cognitive processes involved in religious belief/disbelief. But, given what we know about how misleading intuitions can be, I would say what the authors do not (and cannot in a scientific paper): the results fit perfectly with the idea that religious belief is often founded on intuitions, and these intuitions are often found to be flawed when examined critically.

For those who followed my series on the psychology of religious intuitions, this finding shouldn't be too surprising!

Further Reading: Gervais, W.M, & Norenzayan, A. (2012). Analytic thinking promotes religious disbelief. Science, 336.