Science, Your Bias Is Showing: Autism and Gender

Science is objective, right? That’s what we love about it. Facts are facts, whether you agree with them or not, and science is really a great way of finding out facts.  The best part of science is when we can find out facts that help people, like modern medicine. SCIENCE! YAY! IT’S NEVER BIASED!

This is what some skeptics sound like to me.

Now there are some elements of this that are true and that are premises that are fundamental to skepticism. Yes, facts are not influenced by our feelings about them, and yes, science is hugely helpful and beneficial in many ways. But science is not without bias. Beyond the ways in which women are excluded from doing science, or have their accomplishments and contributions dismissed, the people doing science have biases and that affects the outcomes that they see, the experiments they choose to run, and the questions they want to ask.

And as one might expect, the place that this hurts people most is in the modern medicine that pro science folks hold up as so amazing. When bias comes through in medicine, people die. Physical medicine is getting more aware of the fact that some diseases or problems present differently in men and women, but psychiatry and psychology? These fields are in their infancy and are rife with bias.

There are far too many examples to talk about all of them, but let’s look at a few representative examples. Autism is much in the news of late, and there are a few interesting facts about the demographics of people diagnosed with autism. Autism is about 5 times more common in boys than in girls (although in recent years the rates of autism have been in flux, so these are hard to pin down). Some minority populations have far higher diagnoses of autism (for example Somali children), while black children are underrepresented in autism diagnoses.

A similar example is that borderline personality disorder is far more commonly diagnosed in white women than any other demographic. Depression tends to be underdiagnosed in men, who get diagnoses related to anger disorders when presenting with the same symptoms as women.

Ok, you might say, but how do we know that these are biases in the science and not just examples of the science bearing out bias? First, most definitions of mental illnesses are basically made up to put a name to a common constellation of symptoms. That means they often circulate on a stereotype or the assumptions of the doctors.

The best example I have of this is anorexia. Until the MOST recent version of the DSM, which was only released a couple years ago, amenhorrea was one of the diagnostic criteria. For those who don’t know, amenhorrea is the cessation of one’s period. That means that anyone but cis women of child bearing ages literally could not get an anorexia diagnosis because the physical symptoms were created based on young, female patients. Now if someone has all the symptoms of anorexia except for amenhorrea because they are, in fact, a man, it seems highly unlikely that they actually have another disorder. There are many similar examples that aren’t quite as extreme. Boys with autism are more likely to have very visual, physical behaviors like hand flapping or spinning, while girls are more likely to have high anxiety, but the diagnostic criteria focus more on the physical behaviors. Which symptoms are given high importance or included at all in a diagnosis is political, and can vary over time.

Since autism is a Big Deal right now, let’s talk about how the diagnosis has changed over time, why, and how politics played a role.

Kanner, often credited with discovering autism, was very invested in his condition being a rare and specific one. He famously bragged about turning away 9 out of 10 children from his clinic without a diagnosis. Since that time, we’ve come to understand the concept of the spectrum, and realized that it doesn’t help anyone to very narrowly define the spectrum rather than looking at the wide variety of symptoms that tend to cluster together and calling that autism. We’re also finding that many individuals who present at the higher end of the spectrum struggle with similar issues as those who are lower functioning, but have developed better coping skills. The more children we look at, the more we see that the traits aren’t what we thought they were, and the original conception of autism was created by someone who was sexist in his conception of autism (considering it a male disorder only) and in his conception of what caused it (refrigerator mothers). That conception of autism reigned for decades. If human beings are the ones who get to define what a disorder is, then of course the science around that disorder will be biased.

New research about women and autism is coming out constantly, showing that there are more women and girls who are not getting services that they need. There are studies that find women who have similar underlying difficulties to men with autism present in a different way. When psychiatrists take the time to interview them they find that these girls likely do have autism, but they’re not presenting with the same rigidity, repetitive behaviors, or even sometimes aggressiveness that boys do. There is more anxiety and more shyness. With in depth studies, it’s becoming clear that despite looking different, the underlying issues are similar, something we would miss if we accepted the science as is, that autism looks one way.

We even have to be wary about allowing these flawed scientific principles to influence our conceptions of the mind and gender. Autism is often called “the extreme male brain” because it is so common in boys, and because when it is poorly understood it is believed to be an extreme rational brain with a deficit of emotion and empathy, clearly boy things. We then assume that boys are more rational and girls are more emotional, because “the science” shows that a deficit of empathy is something that happens in boys.

These are the kinds of difficulties that come when a field is in its infancy and cannot look at the underlying causes of things as much as they can look at symptoms. We group symptoms together based on what we can see as human beings. That means that the science is influenced by what we perceive, what we’re willing to group together as “the same,” which makes it subject to all the pitfalls of human perception. That includes bias.

Psychology isn’t the only field that includes bias not just in its hiring, academic, and HR practices. But it’s one of the most obvious. So let’s use this autism and gender problem as a case study to remind ourselves that science is not perfect. I would rather live in a world where I can criticize science and have it grow than one in which the idea of flawed science is anathema.



Olivia is a giant pile of nerd who tends to freak out about linguistic prescriptivism, gender roles, and discrimination against the mentally ill. By day she writes things for the Autism Society of Minnesota, and by night she writes things everywhere else. Check out her ongoing screeds against jerkbrains at www.taikonenfea.wordpress.com

Related Articles


  1. An example I always used was bipolar vis-a-vis schizophrenia dx. It’s been shown that white, Asian, or American Indian pseudopatients are more likely to get be diagnosed bipolar than black or Hispanic pseudopatients presenting the same symptoms, who will instead be diagnosed schizophrenia.

    That shows a cognitive bias on the doctors’ parts. It’s something they’ll have to recognize and figure a way to correct for.

    The same case here with women with autism or men with BPD or…

      1. The first thing you learn is, everyone has a bias.

        It’s especially dangerous in mental health. You’re not actually measuring levels of serotonin or whatever because there’s no safe way to do that. Instead the diagnostic criteria are subjective.

        1. There’s now also studies showing a strong environmental link to ADHD and autism resulting from exposure by pregnant mothers to a variety of chemicals and a bill in Congress currently to more strongly control exposure to these chemicals. Rather than just going by corporate data- which is usually slanted in their favor.

  2. I think you are misunderstanding the older DSM criteria for anorexia nervosa. One of the DSM IV criteria was amenorrhea, specifically _in postmenarchal females_. That does not mean that only postmenarchal females can ever have anorexia, it means that in that subgroup of patients there was an _additional_ criterion of amenorrhea. This would not exclude cis men, therefore, because they are not postmenarchal females. Nor would it exclude postmenopausal cis women, because they are amenorrheic. Nor, for that matter, would it exclude trans women (whether you think of them as_all_ amenorrheic, or don’t consider them postmenarchal). It also specified amenorrhea in the absence of hormone administration, so having periods while on OCPs would not exclude the diagnosis. The only group of patients that this criterion incorrectly excluded were anorexic postmenarchal cis women without amenorrhea while not on hormones.

    The amenorrhea criterion was not present prior to the DSM IV, and was removed in the DSM V because it was recognized that not all postmenarchal females who had the behaviors of anorexia nervosa had amenorrhea; you could be an anorexic cis woman of childbearing age who still had periods.

    Whether clinicians might have discounted the possibility of a diagnosis of anorexia nervosa for a cis male patient because of a tendency to think of anorexia as a ‘female’ disorder is a fair question. However, the DSM IV criteria did not require you to be a woman capable of menstruating in order to be diagnosed with anorexia.

  3. And the Merck researchers who doctored Vioxx studies and how slow the FDA was to react? Yes, capitalism corrupting scientists.

Leave a Reply to Jon BrewerCancel reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Back to top button