Q&A #1: How to Diagnose Autism
Q: What do you think ARE the real symptoms of autism/differences from NTs? What do you think would be a good way to know if you do/don't have it?
--Anonymous
A: You guys aren't starting me off easy, are you?
Suffice it to say, this is actually a hugely controversial question in the field of autism research. Defining autism properly is a pretty difficult proposition: Autism varies widely from person to person; the people with the most obvious symptoms also have the most difficulty explaining what they are experiencing; and the origin and progression of autism, though known to be genetic, is still not fully defined.
There are two basic approaches to defining autism: Internal experiences and external, observable symptoms.
Today's psychologists are focusing increasingly on observable symptoms. Why they do it is obvious: They want to make sure psychology is a true science, dependent on things that can be seen and quantified. They would much rather depend on an objective set of diagnostic criteria than their own gut feelings, because as scientists, most psychology researchers know very well that appearances can deceive and "common sense" isn't always right. It was "common sense" that led to the Refrigerator Mother theory, and we all know how many people that hurt. (If you don't: The refrigerator-mother idea is that autism comes from having been emotionally neglected by one's mother. Obviously, this made it much worse for mother and child alike--a mother who needed help raising an autistic child got blamed for the autism, and the child's treatment was designed pretty horribly.)
Objectively observing someone with autism, talking to them and talking to the people they know and live with, is the easiest way for a psychologist to make a diagnosis. The criteria they use nowadays are either the DSM-IV (a list of definitions of psychological and mental disorders) or the ICD-10 (a list that does the same with disease, injury, etc. in general).
If you want to read those, here are some links:
ICD-10: Childhood Autism
ICD-10: Asperger's Syndrome
DSM-IV: Autistic Disorder
DSM-IV: Asperger's Disorder
DSM-IV: PDD-NOS
There are multiple autism diagnoses (I've left out Rett's and CDD, both of which are quite rare) because of how we figured out what autism was. First we discovered classic autism (also called Kanner's syndrome); very shortly thereafter, Asperger's syndrome. There wasn't originally much difference between the two groups--Kanner's patients were more likely to be speech-delayed, though most of Asperger's patients would today be diagnosed with classic autism. The difference was that we lost track of Asperger's research, and only classic autism was widely known of for quite a while, and thought to be quite rare because we had quite narrow diagnostic criteria. After a while, "autism" came to mean severe speech delay and obvious disconnect from the outside world; and to make the group even smaller, if you had a developmental delay, you were usually simply diagnosed with cognitive impairment and left off the autism spectrum entirely.
The rediscovery of Asperger's syndrome changed all that. Gradually we came to realize that autism was not a small group of unusual cases, but a broad, extremely diverse spectrum. Speech could range from fluent and pedantic to entirely absent; socialization could be anywhere from extroverted to oblivious to others' existence. One autistic might be fascinated with a piece of string; another, with oceanography. Not only did we begin to realize that autism was diverse--we also understood that Asperger's and classic autism were not really well-defined subgroups after all. PDD-NOS (atypical autism; this is the group they use when you are autistic but don't fit into a specific category) is now the most common autism spectrum diagnosis.
So psychologists, with their desire to have a specific set of autism symptoms to check off for their patients, are a bit stuck. On the one hand, they'd love to have proper lists of criteria; but autism is such a slippery diagnosis to pin down that it's pretty much a matter of looking at the diagnosis and saying, "Well, your child has many of these traits; yes, he's got some that aren't on the list; and yes, he hasn't got some of the ones that are on it; but there's no better diagnosis to get him the therapy he needs, so we'll say autism and leave it at that."
A lot of Internet sites have autism screening tests, and I'll post some links to those. But take the links with a grain of salt! Remember that screening tests are not diagnostic. Say that there's a screening test which catches 99% of cases of autism, and only misdiagnoses 1% of neurotypicals as autistics. That's a very good screening test. But let a thousand people take the test. Of those, ten will be autistic, and all of them will probably score autistic. Of the 900 neurotypicals, nine will score autistic. That means that if you score as autistic, the odds are a little less than half that you are actually neurotypical.
Another thing to remember about screening tests is that they don't usually take into account a very important diagnostic criterion: The question of whether these traits cause significant impairment. A diagnosis is something you use to describe a person's traits that are causing problems for them: If there are no significant problems, then there is no reason for a diagnosis. You can score positive on a screening test, have autistic traits, and still not be diagnosable because they don't cause disability. There's no reason to diagnose that group--though of course they fit in quite well with autistics and autistic culture because of how similar their thinking styles are.
Screening tests:
The AQ Test
PDD Assessment Scale
The AQ test works best for adults; the PDD assessment scale works better for kids. They are not diagnostic, but a positive result on a screening test, combined with problems in everyday life, are good reason to make an appointment with a professional.
The second way to tell the difference between autistic and neurotypical is the internal experience of being one or the other. Psychologists generally don't like this because it depends on our telling them what it's like to be autistic, and explaining things they can't directly observe. However, since behavior comes from internal states, the way autistic people think, experience the world, learn, and make decisions is actually much more fundamental to being autistic than the behavior you can observe in someone who thinks in an autistic way.
The differences that I've observed between autistics and neurotypicals come in three broad categories: Learning and cognition; sensory processing; and language/communication.
Let's look at thinking and learning first. Autistic people are very diverse in the way we learn; so the only thing I can say categorically about autism is that autistic people are almost certain to learn differently than NTs. What those differences are is anybody's guess and really depends on the person.
Common learning differences between autistics and NTs include:
In general, a diagnosis of autism is made when someone determines that you have a lot of autistic traits, and that these traits mean you need help with something, and that a diagnosis would help. But defining autism--the essential cognitive style--is a project I think we'll take decades to finish.
Disclaimer: I am not a doctor and cannot provide medical advice. If you have autistic traits that are creating difficulty for you, an autism specialist is your best bet. Q&A updates Monday and Friday (if I have questions to answer). Enter a question by replying to the "Q&A Submissions" post, which can be found under the Tags heading.
--Anonymous
A: You guys aren't starting me off easy, are you?
Suffice it to say, this is actually a hugely controversial question in the field of autism research. Defining autism properly is a pretty difficult proposition: Autism varies widely from person to person; the people with the most obvious symptoms also have the most difficulty explaining what they are experiencing; and the origin and progression of autism, though known to be genetic, is still not fully defined.
There are two basic approaches to defining autism: Internal experiences and external, observable symptoms.
Today's psychologists are focusing increasingly on observable symptoms. Why they do it is obvious: They want to make sure psychology is a true science, dependent on things that can be seen and quantified. They would much rather depend on an objective set of diagnostic criteria than their own gut feelings, because as scientists, most psychology researchers know very well that appearances can deceive and "common sense" isn't always right. It was "common sense" that led to the Refrigerator Mother theory, and we all know how many people that hurt. (If you don't: The refrigerator-mother idea is that autism comes from having been emotionally neglected by one's mother. Obviously, this made it much worse for mother and child alike--a mother who needed help raising an autistic child got blamed for the autism, and the child's treatment was designed pretty horribly.)
Objectively observing someone with autism, talking to them and talking to the people they know and live with, is the easiest way for a psychologist to make a diagnosis. The criteria they use nowadays are either the DSM-IV (a list of definitions of psychological and mental disorders) or the ICD-10 (a list that does the same with disease, injury, etc. in general).
If you want to read those, here are some links:
ICD-10: Childhood Autism
ICD-10: Asperger's Syndrome
DSM-IV: Autistic Disorder
DSM-IV: Asperger's Disorder
DSM-IV: PDD-NOS
There are multiple autism diagnoses (I've left out Rett's and CDD, both of which are quite rare) because of how we figured out what autism was. First we discovered classic autism (also called Kanner's syndrome); very shortly thereafter, Asperger's syndrome. There wasn't originally much difference between the two groups--Kanner's patients were more likely to be speech-delayed, though most of Asperger's patients would today be diagnosed with classic autism. The difference was that we lost track of Asperger's research, and only classic autism was widely known of for quite a while, and thought to be quite rare because we had quite narrow diagnostic criteria. After a while, "autism" came to mean severe speech delay and obvious disconnect from the outside world; and to make the group even smaller, if you had a developmental delay, you were usually simply diagnosed with cognitive impairment and left off the autism spectrum entirely.
The rediscovery of Asperger's syndrome changed all that. Gradually we came to realize that autism was not a small group of unusual cases, but a broad, extremely diverse spectrum. Speech could range from fluent and pedantic to entirely absent; socialization could be anywhere from extroverted to oblivious to others' existence. One autistic might be fascinated with a piece of string; another, with oceanography. Not only did we begin to realize that autism was diverse--we also understood that Asperger's and classic autism were not really well-defined subgroups after all. PDD-NOS (atypical autism; this is the group they use when you are autistic but don't fit into a specific category) is now the most common autism spectrum diagnosis.
So psychologists, with their desire to have a specific set of autism symptoms to check off for their patients, are a bit stuck. On the one hand, they'd love to have proper lists of criteria; but autism is such a slippery diagnosis to pin down that it's pretty much a matter of looking at the diagnosis and saying, "Well, your child has many of these traits; yes, he's got some that aren't on the list; and yes, he hasn't got some of the ones that are on it; but there's no better diagnosis to get him the therapy he needs, so we'll say autism and leave it at that."
A lot of Internet sites have autism screening tests, and I'll post some links to those. But take the links with a grain of salt! Remember that screening tests are not diagnostic. Say that there's a screening test which catches 99% of cases of autism, and only misdiagnoses 1% of neurotypicals as autistics. That's a very good screening test. But let a thousand people take the test. Of those, ten will be autistic, and all of them will probably score autistic. Of the 900 neurotypicals, nine will score autistic. That means that if you score as autistic, the odds are a little less than half that you are actually neurotypical.
Another thing to remember about screening tests is that they don't usually take into account a very important diagnostic criterion: The question of whether these traits cause significant impairment. A diagnosis is something you use to describe a person's traits that are causing problems for them: If there are no significant problems, then there is no reason for a diagnosis. You can score positive on a screening test, have autistic traits, and still not be diagnosable because they don't cause disability. There's no reason to diagnose that group--though of course they fit in quite well with autistics and autistic culture because of how similar their thinking styles are.
Screening tests:
The AQ Test
PDD Assessment Scale
The AQ test works best for adults; the PDD assessment scale works better for kids. They are not diagnostic, but a positive result on a screening test, combined with problems in everyday life, are good reason to make an appointment with a professional.
The second way to tell the difference between autistic and neurotypical is the internal experience of being one or the other. Psychologists generally don't like this because it depends on our telling them what it's like to be autistic, and explaining things they can't directly observe. However, since behavior comes from internal states, the way autistic people think, experience the world, learn, and make decisions is actually much more fundamental to being autistic than the behavior you can observe in someone who thinks in an autistic way.
The differences that I've observed between autistics and neurotypicals come in three broad categories: Learning and cognition; sensory processing; and language/communication.
Let's look at thinking and learning first. Autistic people are very diverse in the way we learn; so the only thing I can say categorically about autism is that autistic people are almost certain to learn differently than NTs. What those differences are is anybody's guess and really depends on the person.
Common learning differences between autistics and NTs include:
- Learning disabilities; being very, very bad at learning a particular subject, like reading or math.
- Extreme talent or even savant skills--being "a natural" for some skill or another, often quite a narrow skill, as though your brain were built for learning that skill (which may be exactly the case). Hyperlexia is common--a combination of extreme skill for reading, combined with a lot of difficulty with comprehension.
- Difficulty with learning from other people; being an autodidact; not fitting into classroom style instruction very well.
- Having a very strong learning style--very strong visual learner; very strong verbal learner; very strong auditory; etc. Personally, I learn best from reading--several times as efficiently than from any other method.
- Having a strong memory and tending to use brute-force memorization to accomplish complex tasks.
- Having problems learning symbolic, non-concrete things. Can range from being unable to recognize a photo as a representation of an object to much slighter problems, like having a tendency to take things literally if one doesn't stop to think first.
- Executive functioning differences. Multitasking, attention control, planning and organization all tend to be problematic for autistics. This can express itself multiple ways, though--anything from an ADHD-like disorganization (or even ADHD proper, a very common comorbidity) to a need for order and routine if you want to get anything at all done, to the point that a simple change can send you into a tailspin. Autistics are often very good at organization and pattern-recognition, maybe thanks to this need for order.
- Difficulty figuring out which stimuli are significant. This can lead to filtering out too much information (leading to a look of being "in one's own world") or not enough (leading to sensory overload).
- Processing sensory stimuli as not strong enough--for example, someone who is undersensitive to cold may be comfortable in a T-shirt despite freezing weather.
- Processing sensory stimuli too strongly--for example, touch sensitivity can lead to a need to wear only very soft clothing with no seams.
- Being unable to locate sensory stimuli properly--for example, having an injured ankle, knowing you feel bad, but being unable to figure out that it's your ankle that hurts.
- Being unable to multitask different sensory input streams. This may be why we don't make a lot of eye contact--we're not good at multitasking visual and auditory, so we stop looking so that we can listen better.
- Not having a good sense of where we are and how we are moving (proprioception and balance). Some kinds of stimming might have to do with getting better input on where your body is located. This can lead to clumsiness.
- Problems recognizing objects, especially faces. Prosopagnosia is common; less commonly, people might have to think consciously before being able to identify an object.
- Problems with learning language. This is quite possibly related to sensory issues (processing auditory data), executive function (having to think of what to say on the spot), and symbolic reasoning (understanding that words are symbols for ideas and objects).
- Problems with connotation. "Reading between the lines" can be quite difficult for someone who thinks literally. This could be because they're having issues with multitasking (understanding the words at the same time as thinking about the meanings) or not having learned some of the finer points of symbols (assuming one symbol-> one meaning; not understanding that combinations of symbols can create meanings that none of the symbols can create alone).
- Understanding body language and facial expression. Once again, this could be a symbol issue, or a multitasking issue, or an executive function issue. It could just be because the autistic brain isn't built to be an expert at understanding these things.
- Understanding social norms. These are harder to learn because unlike neurotypicals, autistics don't easily pick up the between-the-lines communication. Social norms are communicated to children when they're young; if the person can't receive that communication, they won't learn the social norms.
- Having problems imitating someone. Probably another multitasking issue--having to watch them at the same time as you do the thing yourself. On the other hand, some autistics are great at imitating, even using imitated speech to communicate. These folks have probably used their good memories to their advantage.
- Generating original speech might be more difficult. Many people solve this by using echolalia or scripting, or thinking of what they want to say ahead of time.
In general, a diagnosis of autism is made when someone determines that you have a lot of autistic traits, and that these traits mean you need help with something, and that a diagnosis would help. But defining autism--the essential cognitive style--is a project I think we'll take decades to finish.
Disclaimer: I am not a doctor and cannot provide medical advice. If you have autistic traits that are creating difficulty for you, an autism specialist is your best bet. Q&A updates Monday and Friday (if I have questions to answer). Enter a question by replying to the "Q&A Submissions" post, which can be found under the Tags heading.