Blanchard chose participants by feelings of being a woman rather than for gender dysphoria. Excluding those transsexuals who's gender incongruence focuses on not being our biological sex is only likely to over-estimate the proportion of transsexuals who have especially strong longings (including sexual fantasies) of being our target sex. By using different criteria to create the core autogynephila scale than to compare "homosexual" and "non-homosexual" transsexuals, Blanchard assumes, perhaps incorrectly, that autogynephilia is a single unified construct in the minds of transsexuals.

This is page 6 of 13.          [1]   [2]   [3]   [4]   [5]   [6]   [7]   [8]   [9]   [10]   [11]   [12]   [13]  

Transsexual? Gender Identity Conflict? And how participant selection bias might influnce Ray Blanchards's Results about Autogynephilia.

When we conduct a social science study, the way we choose participants can have a profound impact on the types of results we get. For example, gay men and lesbian women used to be viewed as disordered because those who participated in studies were those who came to therapy. That is, those gay men and lesbian women who were comfortable with who they were did not seek therapy so their experiences were not represented in therapeutic samples. Transsexuality may be somewhat different because any transsexual who would like sexual reassignment surgery needs a therapist's permission, at least if they would like their Sex Reassignment Surgery completed by a reputable surgeon. Still, we can ask, "How did Blanchard choose his participants and what could his choice of participants say about the results?"

The Clarke Institute maintains a database of questionnaires completed by their clients. Blanchard used these questionnaires to identify 302 male-bodied persons who, at least while wearing some female clothing, had ever "felt like a women." They were considered the full sample of those with gender identity conflicts. Two-hundred twelve (212) of these participants "felt like a woman" continuously for at least a year and were labeled "transsexuals."

At first glance this seems like a perfectly reasonable selection criteria. Many male-to-female transsexuals I know have said they "felt like women." Then again, many do not. This includes me! That is, if I were a client at the Clarke institute the information I would have given on their questionnaires would not have been included in this analysis. What does it mean to "feel like a woman" or to "feel like a man????" I personally feel that my transsexuality is more about feeling uncomfortable being viewed as a man (gender dysphoria) than being happy about being viewed as a woman (gender euphoria). I am not the only transsexual with this view of gender identity conflict. Yet we were excluded from Blanchard's transsexual sample and even his full sample. Excluding those transsexuals who's gender incongruence focuses on not being our biological sex is only likely to over-estimate the proportion of transsexuals who have especially strong longings (including sexual fantasies) of being our target sex.

A second way Ray Blanchard selected participants creates another more subtle bias. When Blanchard assigned transsexuals to sexual orientation groups and compared their rates of autogynephilic fantasies, he naturally used the 212 participants who were transsexual. But when he defined the construct of autogynephilia (using a factor analysis), he used all 302 participants who had ever felt like woman (such as only when cross-dressed). Just because the "core autogynephilia scale" is a single unified construct among this larger sample does not mean it is also a single unified construct among transsexuals. One consequence of Blanchard's model is that cross-dressers and "non-homosexual" transsexuals are different by degree instead of kind. Yet by assuming the same constructs would apply to both transsexuals and cross-dressers before the factor analysis, he begs the question about this issue.

This is page 6 of 13.          [1]   [2]   [3]   [4]   [5]   [6]   [7]   [8]   [9]   [10]   [11]   [12]   [13]