The Measure of…

…Autogynephilia.

Over the years, certain aspects of autogynephilia and its expression have caused both confusion and consternation in the transgender world.  Some of the quirks found in the data, such as the now well documented (slightly) lower level of reported erotic cross-dressing in “bi-sexual” transsexuals, suggested that perhaps some bisexuals weren’t autogynephilic, etc.  But a new paper, though VERY tentative, give some credence to the idea that there are different subsets of autogynephilia, different expressions of it, that correspond with sexual behavior, specifically, that in addition to the previously well documented four types of autogynephilic expression, a fifth expression exists, that may explain these quirks.

Hsu, et al. have created a new instrument for experimental purposes and did a validity and factorial analysis of a population of known autogynephiles against a control group of men.  The known AGP men were from an internet AGP erotic sharing group, as the authors explained,

“Participants were 149 adult men (M age=34.40 years, SD=11.20) recruited from Internet forums dedicated to sharing and discussing erotic fiction and media depicting autogynephilic fantasies, including cross-dressing, transforming into a woman, and body swapping with a woman. Most of the participants identified as heterosexual (80.54%) although a substantial minority identified as bisexual (14.77 %). Four other men identified as homosexual and one as asexual; the remaining two men selected‘‘Other’’but did not specify their sexual identity. Because participants were recruited from Internet forums catering to men with autogynephilia, all participants were included in the analyses regardless of their sexual identity.  In addition to the 149 participants considered to have autogynephilia, 112 adult heterosexual men (M age=32.63 years, SD=10.88) who reported having never cross-dressed were recruited as a control group from Amazon Mechanical Turk, a website used by people who want to earn small sums of money quickly by taking online surveys.”

However, this study was not directly meant to learn about how AGP and non-AGP men differed, as the authors explained,

“Specifically, our study did not address the issue of whether autogynephilia represents a dimensional or taxonic difference from typical male sexuality. In order to explore that issue, it would be necessary to obtain a representative (and presumably large, given the likely rarity of autogynephilia) sample from the general population (Beauchaine, 2007). Rather, we explored differences among autogynephilic men, assuming that such differences are dimensional, and we focused on describing their dimensional structure. Thus, the primary empirical question that we addressed was not ‘‘How do autogynephilic men differ from other men?’’ but ‘‘How do autogynephilic men differ among each other?’’

The study was most especially NOT about the differences between AGP transsexuals and MTF transkids.  In fact, as I will explain below, this new instrument has several items that make it invalid to such a task.  But in studying autogynephilia in AGP males in general, it is a very good start.

The most exciting thing that can happen when conducing science is to hear this phrase, “That’s odd…”  It means that something unexpected has been found in the data, something new.  In this case, we may have found something, two somethings actually.

It has always been mooted about that ‘transvestic’ autogynephilia is the most commonly found form of autogynephilia, along with ‘anatomic’, ‘behavioral’, and ‘physiological’.  Sometimes, a fifth type is discussed, ‘interpersonal’.  The usual explanation for bisexual or “pseudo-androphilia” sexual behavior is that ‘behavioral’ autogynphilic ideation includes acts of having sex with a man as a woman.  But this paper supports the notion that the fifth type, ‘interpersonal’, is responsible.  The central concept of ‘interpersonal’ autogynephilia is that of narcissistic desire to be admired by other people, as an attractive, sexually desirable woman, emphasis on being a woman.

It has also been suggested that gender dysphoria, the desire to “change sex” is most motivated by ‘anatomic’ autogynephilia and that most Cross-Dressers, who do not have such strong gender dysphoria experience less ‘anatomic’ and more ‘transvestic’ autogynephilia.  But in this study, the reverse seems to hold!

Those interested in the details should read the paper carefully, but to summarize, the authors explored four factor models and five factor models to explain the variance of the data.  They also explored the idea that a single factor underlay the whole.  The best model seemed to be the five factor model with a single factor, “autogynephilia” underlying the whole.  Thus, a generic ‘autogynephilia’ does exist, but also that there are variations on a theme, with (at least) five types identified.  The newly identified type is indeed interpersonal’ and is very strongly correlated with both identifying and being behaviorally bisexual.  The ‘behavioral’ type did NOT correlate with bisexuality / androphilia.

However, a note of caution needs to be introduced here, as two of the four items that had high loading for the ‘interpersonal’ factor specifically relate to dating and having sex with men… thus is essentially measuring the same construct.  It might be interesting in the future to redact these two items to see if the high correlation remained.  I’d even like to add another item to the ‘interpersonal’ set (see below).  Yet another caution needs to be observed… and that is finding EXACTLY who (or rather, what) these men were finding attractive and having sex with.  The “men” these individuals may be having sex with are very likely to be other Cross-Dressers!  IF so, this puts a rather different interpretation on their putative ‘androphilia’.

In looking at the correlations, both zero order and partial regression coefficients, ‘anatomic’ autogynephilia did NOT correlate with gender dysphoria, counter to previous report (Blanchard).  Instead, the highest correlation with gender dysphoria was ‘interpersonal’ !!!  While it may have made some theoretical sense to believe that ‘anatomic’ autogynephilia would be a powerful motivator for “changing sex”, given that both involve anatomic features.  This data suggesting that ‘interpersonal’ autogynphilia would be even more motivating makes sense when one considers the required “real life test” before one may obtain SRS, that social transition and the ‘real life test’ are all about interpersonal aspects of life.  But, again, we need to introduce a note of caution here.  The sample used in this study were NOT sufficiently gender dysphoric as to actually proceed to transition and SRS… only that they might have wished that they could have been born as girls, etc.  This may be more a measure, in this population, of such weaker desires.  This study needs to be repeated with a sample of AGP transwomen to confirm or disconfirm this unexpected result.

The fact that ‘interpersonal’ correlates so highly with both an interest in sex with men (even with the caution of the items I mentioned above) and gender dysphoria might explain why so MANY post-transition APG transwomen experience a “change in sexual orientation” from exclusively gynephilic to being bisexual / (pseudo)androphilic.  Lawrence showed that perhaps 38% experience such a change.  The two may go together, transition and “orientation change”.

Over all, this study suggests ‘anatomic’ autogynephilia may be just as common, if not more common than ‘transvestic’.  Most importantly of all, as the authors explain,

“The finding that a general factor of autogynephilia underlies the five types among the sample of autogynephilic men was not predestined to be true. For example, autogynephilic men may engage in or be invested in behaviors or fantasies of one type of autogynephilia at the expense of those of other types. In contrast, the general factor accounted for a much greater amount of the total variance of the 22 items than did the group factors, suggesting that there is an overall tendency for some men to be more autogynephilic than others. Indeed, scores on the GAS, a measure we constructed by adding all 22 items,were normally distributed. From these results, it appears that the types of autogynephilia that a man has are less important than the degree to which he has autogynephilia.”

To explore their new instrument’s validity as a measure of autogynephilia, they compared the scores of their putatively known autogynephilic men with heterosexual control men:
Scale/subscale                                             AGP                         Controls         Cohen’s d

General Autogynephilia Scale (SD)      3.32  (0.89)        1.16 (0.38)       3.33

The absolute range for the General Autogynephilia Scale was 1–5.   My guess is that most people would find the fact that the scale is from one to five confusing, so if we made the scale from zero to four, the numbers would be 2.32 for AGPs vs. 0.16 for the controls… being more intuitively obvious that the scale works to differentiate AGP from non-AGP males.  And, for comparison purposes, making the scale from zero to eight, like Blanchard’s Core Autogynephilia scale, would be very useful.

As to my earlier comment as to why I believe that this new instrument is not valid for exploring the differences between putatively autogynephilic transwomen and putatively “homosexual” transwomen is the construction of three of the items in the instrument:

How sexually arousing would you find each of the following activities?

9. Having a stranger mistake me for a woman.
10. Picturing myself as a woman having sex with a man.
11. Having a man take me out for a romantic evening.

Item number nine, as constructed, is very likely to be interpreted in a rather different manner than that intended by the authors.  The term “mistake”, in someone who is extremely gender dysphoric and presently identifying as a woman, would have a very negative emotional valence.  Even if that individual might find having strangers who accept / perceive her as a woman as sexual arousing, she is unlikely to endorse this item as she will not experience that act of “passing” as a “mistake”.  This item would need to be modified to have a more acceptable valence.

Items 10 and 11 are problematic in that a truly androphilic and extremely gender dysphoric transwoman would also endorse these items, even in the total absence of autogynephilia, as would any heterosexual natal female for that matter.  The context of the question is not self-evident.  In fact, the construction of 11 doesn’t even specify “as a woman” or “dressed as a woman”… and even if it did… they still would not be interpreted as describing an autogynephilic motivation.  Context is everything.

In the same vein, but on the flip side, I would, for the purposes of strengthening the measure of ‘interpersonal’ autogynephilia, suggest a companion question of,

23. Picturing myself as a woman having sex with another woman.

This, I believe, is a VERY common ‘interpersonal’ autogynephilic fantasy.

And speaking of common fantasies.  The authors also explored correlations between their factors and paraphilic sexual interests.  Not unexpected, there was a slightly increased interest in sexual masochism in the AGP sample.  The correlation was highest with ‘interpersonal’ and ‘transvestic’ autogynephilia.  To explore this better, I would add another item to the instrument:

24. Being forced to wear women’s clothing by another person.

This might answer a question regarding “forced feminization” fantasies, that of which of the two competing hypotheses is correct.  Is “forced feminization” a convolution of ‘transvestic’ autogynephilia and simple sexual masochism?  Or… Is “forced feminization” a means for reducing the guilt and shame of ‘transvestic’ autogynephilia, without experiencing masochism.

Hypothesis were meant to be tested.

 

Reference:

Kevin J. Hsu, A. M. Rosenthal, J. Michael Bailey, “The Psychometric Structure of Items Assessing Autogynephilia”
Archives of Sexual Behavior, DOI 10.1007/s10508-014-0397-9