Real autism tests

Wednesday 17 September 2003This is close to 20 years old. Be careful.

In a comment on my posting about Newsweek’s autism test a few days ago, Laurent said,

I do not understand why we would categorize people based on test. 50 questions to define a human brain, this seems a little bit simple.

Laurent is right, of course. That quiz was designed for regular people to take as an entertainment. Real diagnosis of autism is a much more nuanced procedure, but perhaps not as much more as you would think.

For example, here’s an experimental PDD screening questionnaire. It contains 44 questions, answered on a five-point scale. The existing standard diagnosis protocol is hard to find online, but here’s an informal summary.

The key phrase in the diagnosis is “6 of 12 symptoms”. This means two kids can both be diagnosed with “autism”, and have no symptoms in common. This is ridiculous.

Imagine we didn’t know how legs and feet really worked inside. Any time someone couldn’t walk, we’d say they had “Can’t Walk Disease”, but to make it sound more official, we’d call it “anambulism” (from the Latin root “ambul”, to walk). If you twisted your ankle, you’d be diagnosed with anambulism. Torn achilles tendon? Anambulism. Pulled hamstring? Anambulism. Broken leg? Severe anambulism. Plantar warts? Mild anambulism.

Of course, the diagnosis of “anambulism” is ridiculous, but only because legs and feet are relatively straightforward devices, and we can see what is really preventing them from working properly. Because we understand the mechanisms, we can provide appropriate and effective treatments for each malady. We can deal with the different underlying problems differently.

But the mind is not straightforward at all. We have no idea how it works, and we have no idea what’s preventing it from working when it doesn’t. When a child behaves in certain bizarre ways, we say he has “In a World by Himself Disease”, but to make it sound more official, we call it “autism” (from the Greek root “aut” for self). The diagnosis is about observable behavior. Many treatments are about modifying observable behavior, or are scattershot techniques that sometimes work, but we don’t know why.

There are other diagnoses besides “autism” for similar afflictions, like Asperger’s and PDD (pervasive developmental disorder), but these are mostly just measures of degree, not of kind. They label different points on the same spectrum, and choosing among them seems to have as much to do with the practitioner’s preference as with real science.

I hope in 50 years we’ll look back on this thinking and chuckle at how we could have been so simplistic. We’ll have split “autism” into at least four or five separate diseases, each thought of as different. Maybe we’ll understand causes. Maybe we’ll have effective treatments. Maybe “autism” will go the way of “anambulism”.


I fully agree with your post. In 50 years from now it will be funny to browse all the older scientific papers. Even though the brain is challenging and fascinating subject, I have the impression that the hypotheses based on those beautiful colored maps of the brain are close to the tales of the ancestral tribes which explained the surrounding world.
Nothing wrong with that, this is the way science work but we have to take this into account when we want to apply the 'scientific' conclusions to the human society.
Hopefully we will laugh at this in 50 years time. Lewis Terman's IQ tests (or more accuratly derived versions of it) are still very pervasives in education and workplaces today. William Shockley (the guy who invented the transistor) even pushed it as far as proposing (in the 60's) that the govt. offer money to people with low IQ for having them sterilized...

Read "How would you move mount Fuji" by William Poundstone for some interesting background on this.
i'm autistic, and the test was very accurate.
i believe it'll only get more accurate, and yes autism will become an outdated "term", but the state or condition, it won't change; the name will.
Autism is not just one disorder but a spectrum so a test for autism is useless in and of itself as it has a diverse grouping of disorders which are affected by certain differences in the brain. so two people with Autism spectrum disorders could share no quantifiable symptoms other than a learning difficulty or social and behavioral problems

Add a comment:

Ignore this:
Leave this empty:
Name is required. Either email or web are required. Email won't be displayed and I won't spam you. Your web site won't be indexed by search engines.
Don't put anything here:
Leave this empty:
Comment text is Markdown.