Thread by @AutSciPerson

I love how "measuring autism" using questionnaires from decades ago created by cis white neurotypical researchers, & created with cis white hetero boys who don't mask in mind..

is seen as objective. 😂

Do they know how bad the AQ is?
"I usually notice car number plates"

1/9

https://twitter.com/DrMBotha/status/1422208106948448262

Here are a few other gems:
"I am fascinated by dates." (??)
"When I'm reading a story, I can easily imagine what the characters might look like." (HOW is this related?)
"I don't particularly enjoy reading fiction."
"I am not very good at remembering people's date of birth."

2/9

I bet you people have not updated these questions in their research with this questionnaire. I bet you people still use the old AQ.

And oh, look! Turns out they're not actually helpful!

Link here - embrace-autism.com/autism-spectru…

3/9

The fact that autistic people were not involved in making these diagnostic tests in the first place means they're -really not objective- or helpful in diagnosing autistic people!

It is mind-boggling how backwards autism research is. Do you know how many AQ's I've filled out?
4/9

Do you know how many researchers tell me "Well we had to have something there to quantify autistic traits so reviewers were happy" (even with an "official" diagnosis as part of the criteria).

Do you know how many brain scan papers correlate to bullshit surveys like this?

5/9

They'll conclude "autistic people who have worse theory of mind have less activation here.." and use the EQ as the correlate - The EQ is a survey that asks "What would a neurotypical do?" and the right answer is always the neurotypical one.

The AQ is even more nonsensical.

6/9

Other research correlates ERPs or EEGs or fMRI activation based on the short version of the AQ, or worse, surveys that they give to parents - and have the parents fill it out about their autistic kid! Asking about how anxious they are, or sensory sensitivities, etc.

7/9

There's research that has shown that parent reports don't always correlate with autistic children's/adolescent's self-report.

And sometimes, in "objective" autism research, they don't even ask the autistic child whether the intervention they performed actually helped them.

8/9

There are so many reasons why what neurotypical people consider "objective" autism research isn't objective at all.

9/9

Also - the children's version of the AQ shows a noticeable gender differences in assumed-to-be cis NT children.

I actually can't believe they got away with publishing this.

10/

"This may be because in the clinical groups, many more boys were in the study than girls. This is a common issue since the high male to female ratio of ASC limits the number of females available to participate in research studies."
docs.autismresearchcentre.com/papers/2008_Au…

11/

It's surprisingly difficult to even find the number of autistic girls they had in the study. Even using the search bar is difficult. The only reference I see in it is Table 3:

They only had 36 assumed-to-be-cis autistic girls, and 156 and/or 312 assumed-to-be autistic boys
12/

And when they graph the results, they put the autistic boys and autistic girls together in one group, while the neurotypical boys and girls are separated into their own on the graph.

13/

You'd think if they noticed a difference in the neurotypical population by assumed gender, they would consider that autistic kids might show the same pattern, & with a larger sample size, would maybe see that the autistic girls had a lower AQ score just due to the questions.
14/

But instead of doing any of that, they just say it's hard to find autistic girls in the first place and that they aren't as common.

And this is the "standard test" in research for measuring autistic traits in autistic people.

You really can't make this up.

15/

Also, there is absolutely 0 information about race or ethnicity. I can't find anything about race whatsoever.

Which honestly, that tracks in "objective" autism research.. Sigh.

I'm just going to assume they were probably all white.

16/

One more thing -

That paper I linked is the Child version of the Autism Quotient -

So it's parent-report. I wonder if that explains the gender difference in the NT population.

Now let's look at the original AQ questionnaire created for adults.

Here's the original paper from 2001 -

They only used "high-functioning" autistic people (i.e. speaking with average IQ),

and had 45 autistic "males" & 13 autistic "females"

docs.autismresearchcentre.com/papers/2001_BC…
18/

13 AFAB people is totally enough to tell if a questionnaire like this is valid in that population of people, right?

...Right?

They had a way larger sample of controls with a more even split between AFAB and AMAB people.
19/

They also similarly found a difference in score between AMAB people and AFAB people in the control group (people assumed to be cis men scored higher on the AQ),
and saw no difference in the autistic group, though the sample size was much, much smaller.
20/

I also skipped over a large portion of the methods, wherein the researchers say that the autistic people in their sample are really "high functioning" and can read and therefore won't ever interpret the statements on the AQ "incorrectly."
21/