Reply to Lawrence and Bailey (2008)

Lawrence and Bailey (2008) were critical of our inclusion of scales that were said to be ‘‘unrelated to autogynephilia’’ in our cluster analysis to classify transsexual participants into two groups. These scales, Attraction to Transgender Fiction and Attraction to Feminine Males, were included because in the taxometric analysis performed by Veale, Lomax, and Clarke (2007) these scales were able to differentiate between latent taxa (two fundamentally distinct categories) greater than the level of d = 1.25 recommended by Ruscio, Haslam, and Ruscio (2006) if these taxa do exist. That is, if there do exist two distinct categories of male-to-female transsexuals in this sample, then the Attraction to Transgender Fiction scale would be able to distinguish between these groups at d = 2.06 and the Attraction to Feminine Males scale at d = 1.90. The corresponding d scores for the Core Autogynephilia and Autogynephilic Interpersonal Fantasy scales were 2.15 and 1.47, respectively, and none of the other sexuality related scales in the Veale, Clarke, and Lomax (2008) study reached the d = 1.25 threshold. Although the sample size was smaller than what is recommended for a taxometric analysis, generally a sample size of 169 should be considered sufficient for calculating Cohen’s d scores for the difference between two groups. Lawrence and Bailey claim that neither the Attraction to Feminine Males nor the Attraction to Transgender Fiction ‘‘scale bears any theoretical or intuitive relationship to the concept of autogynephilia.’’ Admittedly, the Attraction to Feminine Males scale would not have been our initial choice of a scale to differentiate between transsexual subtypes—this is probably a correlate rather than a core component of autogynephilic sexual attraction—nevertheless, our reasons for including this variable were based on methodological reasons outlined above. On the other hand, we believe that the Attraction to Transgender Fiction scale is measuring a core component of autogynephilic sexual attraction—in its manifestation in erotic narratives. As evidence for this was the notable correlation coefficient between Core Autogynephilia and Attraction to Transgender Fiction of q = .52 reported in Veale et al. (2008). In the original study, which used a Likert response scale for the Core Autogynephilia scale, this correlation coefficient was r = .67 (Veale, 2005). Lawrence and Bailey also suggest that ‘‘this might have been the only way they could create two groups out of what is actually one relatively homogenous, autogynephilic group.’’ However, as we will describe below, finding a way to create two groups was not the difficulty—a conclusion of whether splitting the sample into two groups is valid or not was more difficult. Lawrence and Bailey were critical of our labeling the transsexual group that scored lowest on the four variables as ‘‘non-autogynephilic.’’ Indeed, as they point out, this group scored comparably to Blanchard’s (1989) ‘‘nonhomosexual’’ (autogynephilic group) on the Core Autogynephilia scale and, as they don’t point out, this group scored significantly higher then the biological female group on this scale on the ANCOVA in our study, suggesting that they were not actually ‘‘non-autogynephilic’’ at all. We would like to give an explanation of the history of our use of these labels. The origin of this grouping of transsexuals into ‘‘autogynephilic’’ and ‘‘non-autogynephilic’’ came from the master’s thesis this Journal article originated from (Veale, J. F. Veale (&) D. E. Clarke School of Psychology, Massey University, Albany Campus, Private Bag 102-904, North Shore Mail Centre, Auckland, New Zealand e-mail: [email protected]