Invalid measures invalidate us: ciscentrism and ableism in the trans autism literature

The author of this post is a transgender person conducting autism research at a major Midwestern university. Here they reflect on ways cisgender bias may impact neuroscience findings and theories and how transgender and autistic voices and insights could help alleviate these problems.


Two relatively recent* publications (see, here and here) address rates of autism among transgender people, finding that autistic people are over-represented in transgender samples relative to cisgender samples. Both of these studies are informed by the “extreme male brain” theory of autism, which posits that personality traits/cognitive styles are reliably sexually dimorphic, and that autism is associated with extremely “male typical” traits. The extreme male brain theory relies on the assumption that personality traits are gendered AND consistently associated with the brain, and that increased prenatal androgen exposure is a likely cause of these brain differences. There has been a great deal of excellent scholarship (see, for example, here, here, and here) in feminist science and technology studies that critiques and questions these assumptions that I will not rehash.

These trans/autism studies have similar experimental designs: researchers collected data from a sample of transgender individuals receiving care at a gender clinic. These participants completed an assessment form called the Autism Spectrum Quotient (AQ). This form is designed to assess for traits associated with autism, and is divided into social, attention switching, attention to detail, communication, and imagination subscales. Both of these studies used the same dataset of cisgender people as their comparison sample. This cisgender dataset is previously published and includes AQ norms for a large sample of (presumed) cisgender people.

There may be more trans autistic people than would be expected from the prevalence of each of those identities in the broader population. I want to be clear that this is not a “problem” for which we need to determine the “cause.” I do, however, want to problematize the way that assessment tools, designed and normed for cis populations, can lead to invalid claims about transgender people. Importantly, many autistic people, trans and cis, have already critiqued the ways in which representations of autistic people in the research literature and elsewhere do not consider autistic perspectives (see, for example, here and here). Others have emphasized the way that autistic ways of communicating are pathologized in a literature dominated by neurotypical perspectives (see here for example). Measuring social skills by assessing comfort and enjoyment with interacting with neurotypical people misses the point. My critiques of ciscentrism in this literature are greatly indebted to the work of cis and trans autistic activists, writers, and scholars.

Many of the individuals in the current study have reported that they did not fit in with others; indeed, both MtF and FtM cohorts showed more dysfunctional scores in the social skills subscale…supporting a reported sense of impairment.” (Pasterski et al., 2014, p 391).

I am not socially impaired when I have difficulty fitting in cisgender culture or with cisgender people. The AQ has captured trans people’s experiences with marginalization and transphobia. Many of us prefer routines and predictability, one of the traits measured by this scale. Routines become important for many of us as strategies to avoid transphobic violence: this is the restroom I can use safely at school, if I take the 5:30 bus home from work I am less likely to be harassed, I wear my hair this way every day so I am less likely to be misgendered, etc. We “prefer to do things the same way over and over again” because it keeps us safe.

Likewise, questions on the AQ about enjoying childhood play remind us of the way our imaginary play was policed and gendered. Many of us did not enjoy playing imaginary games with our childhood peers, because there was no room for us to imagine our trans selves in a story, or because our favorite toys were taken from us. The AQ also assesses attention to detail with items such as: “I often notice small sounds when others do not” and “I tend to notice details that others do not”. Attention to detail also keeps us safe. Particularly given the high rates of PTSD in trans populations, high could be due to sensory hyper arousal, which can also be present in some autistic people, but is a general construct not necessarily related to autism per se. Items like “I find social situations easy”, “I find it hard to make new friends”, “I enjoy meeting new people”, etc. are all attributable to the difficulty we can experience navigating a cis-dominated world. “Social chitchat” is not enjoyable for me because it so often devolves into invasive personal questions about my transition status or my relationship with my parents.

Personally, I have a complex relationship with “thinking of myself as a good diplomat” because, as the only transgender PhD at my institution, colleagues constantly demand that I represent trans people. On days when I gently correct a colleague for casually insisting that “pronouns aren’t important,” I think of myself as an excellent diplomat. By the third time I’ve been asked to give an uncompensated Trans 101 in a month, not so much. Likewise, I am certain that many of my colleagues and friends are tired of hearing me talk about the poor scientific quality of the transgender medical and biology literatures. “People often tell us that we go on and on about the same thing” because we are compelled to speak ourselves into being in a culture that would prefer we not exist.

I would venture that at least twenty of the fifty questions on the AQ are not valid for transgender people. Because of ciscentric bias, these researchers forgot the most famous maxim in science: “correlation is not causation.” The authors attribute differences they observed in transgender people to be causal rather than correlational; they did not consider the (obvious to any trans person) idea that being transgender mediates social experiences. Attribution of elevated scores on the AQ to an “extreme male brain” among trans people makes several logical leaps.

These leaps aren’t “caught” by cisgender researchers because of their unexamined ciscentrism, although Pasterski and colleagues do acknowledge that the extreme male brain theory doesn’t fit their findings in trans women. Regardless, inclusion of transgender autistic people in the research process (from hypothesis generation to data interpretation) would improve the scientific quality of this work and increase its relevance to trans and autistic people. Chillingly, Jones and colleagues end their paper with the following recommendation: “Clinically, even if only for a minority of individuals considering sex reassignment surgery (sic), the formulation of undiagnosed autism might be a helpful alternative to explore” (p 305).

*It’s 2017 and trans research in psychology and neuroscience still regularly uses the Blanchard typology. 

Facebooktwitterby feather

Leave a Reply

Your email address will not be published. Required fields are marked *