Implications of Profiling for Asperger's Assessment

Post Reply New Topic

sorry i do not understand what you mean. can you please be more specific? or are you being facetious again? i cannot tell.

I don't understand your point.

You talk about a problem in the detection of lies and connect it in some way to autism diagnosis. I don't see the connection

my main point was that because of the abstract nature of asperger's it is [difficult] to scientifically test a clinican's ability to profile people with asperger's. thus, it maybe helpful to look at the effectiveness of other forms of profiling. studies have shown that with the use of CBCA lie detection increases to 68%. whilst this is well and good 32% or almost one in three will result in error. knowing this, what case is there to suggests that clinicans profiling people with aspergers are significantly more effective at reducing error?

Melanieeee

Your point as I understand it is that there is a very wide margin of error in many psychological tests, just as in lie-detector tests used by the police. The police wouldn't get a conviction based just on a lie-detector test. (I hope not, at least. Maybe it has happened somewhere.) To solve a crime conclusively a whole range of things can be brought forward (eye-witnesses, murder weapon, forensic tests etc). Is it not the same with psychological testing for AS or musical talent or anything? If someone wants to use just one simple test then there will surely be a really huge chance of getting whatever they are testing for wrong.

Actually, there's a theory that all psychiatric traits are evenly distributed, and any categories we use are artificial. However, even if that's true, the categories are still useful, because if we just designated someone as 'atypical' with no further specifications, we'd know nothing useful about them.

With regards to the specific idea of estimating accuracy of AS diagnosis by accuracy of lie detection, I think that's pointless. Those two are very different phenomena. In lie detection, you're trying to distinguish two groups of NTs who only differ in their current mental state - it's not surprising that it's hard. To distinguish an NT from an autistic, in contrast, you use general personality features that appear in multiple settings and have a pervasive impact on the person.

Since we can't have an 'objective' measure of autism, what researchers usually do is test interrator reliability of diagnosis - in other words, if Person A goes to Psychologist A and gets an autism spectrum diagnosis, then goes to be assessed by Psychologist B (who does not know about the diagnosis) what's the likelihood that Psychologist B will also say the person is autistic? (This is assuming both psychologists are competent to diagnose autism.) This kind of research generally shows that autistic vs not-autistic judgments tend to have high interrator reliability, but classification of different subtypes of autism is extremely unreliable (one person's AS is another person HFA and a third person's PDD NOS - this is why DSM-V has lumped them all together).