New study: Sexual behavior changes but not sexual orientation

Media have been all over the Bachmann clinic story this week with lots of related discussion about how much, if at all, sexual orientation changes. As a part of this discussion, I noted that a study I am writing up found that over all married gay people assess themselves as growing more same-sex attracted over time.
Along the way, colleague Mark Yarhouse wrote to remind me that he found something similar in a study recently reported in the new issue of the Christian journal Edification (check out the entire journal for background on the whole arena of evangelicals in the study of sexual identity).
The study Mark referred to was authored by his team at Regent University and titled, “Characteristics of Mixed Orientation Couples: An Empirical Study” (start reading at page 41). The sexual minority participants had been married an average of just over 16 years and the average age was 45. While it was not a study of efforts to change, one could reasonably assume that if a group had participants who had shifted orientation very much, then this would be the group. However, that is not what they found, at least not at the level of attraction. See Table 5 below:

You will have to click the table to see it more clearly, but the table  demonstrates that the Kinsey scores shift more toward the heterosexual side when the participants were asked about their sexual behavior but when asked about their attractions, fantasies, and emotional attachments, there was no change. The Kinsey Expanded scale included an average of participant Kinsey assessment of behavior, attractions, fantasies and emotional attachments.
The authors summarized the results:

The mean score of the Kinsey behavior scale before marriage was 3.60, which falls in between the Largely heterosexual, but more than incidental homosexual and Equal amounts of heterosexual and homosexual categories. The mean score of the Kinsey behavior scale currently was 2.80, which falls in between the Largely heterosexual, but incidental homosexual and Largely heterosexual, but more than incidental homosexual categories.
On the Kinsey Expanded version, the mean score for both before marriage and the current assessment were 4.33 and 4.57 respectively. Both of these scores fall in between the
Equal amounts of heterosexual and homosexual and Largely homosexual, but more than incidental heterosexual categories.

I need to ask Mark, and perhaps he can comment, what the Expanded Kinsey result would look like if the behavioral Kinsey score was removed from the average of all Kinsey scores. Perhaps, the attraction, fantasy, etc., scores would rise moderately.
At any rate, the results are consistent with what I am finding as well. People adapt their behavior to their beliefs and commitments but their orientation does not shift, on average.
In all of the bluster about change therapies and clinics, I think evangelicals need to face what evangelical academics are finding in research. Also, a word to the media, both Christian and mainstream, quoting advocacy groups will get you two sides for dramatic tension, but if you want to know how research informs the questions you are asking, please consult those who, despite their religious loyalties, will report accurately.