Accuracy of online Aspergers' diagnosis tests

Post Reply New Topic

I know many people who think they might have Aspergers take online tests to see if they fit the criteria. The question is, how are accurate are these tests? Do you believe they correctly distinguish between Aspies and NTs?
For example, the AQ test seems to be quite popular but I read in an article somewhere that it misidentifies many NTs as Aspies.
Anyways, I'm interested to hear your thoughts on the Aspie Test, the BAP test, RAADS-R, and any others related to Aspergers and autism.

I think they can help point you in the right direction, but they are no replacement for a professional who specializes in Autism Spectrum. I took a couple of them myself before I went to be evaluated and diagnosed a couple of weeks ago. Both of the ones I took -- I don't remember which ones -- had me as an Aspie, but that doesn't mean that they would work for everyone. Some people don't answer the questions 100% honestly, and some can't because they don't recognize certain actions or traits within themselves.

The RAADS-R is actually the official version clinicians use to diagnose in my country. That one is legitimate - if you can get yourself a copy from someone who has an actual hard copy at home.

There is no such thing as an online Asperger's diagnosis test. There is also no such thing as an Asperger's diagnosis test. What you refer to are screening tools that are used in conjunction with a qualified clinician's thorough interview and background information gathering.

Therefore no-one should rely on the results of just the test itself. There are other conditions which can mimic the traits and symptoms of AS which the test does not pick up.

The results of the tests can be relied upon if all the other information your clinician gathers also shows AS.

I don't trust those quizzes.
I don't trust internet tests for mental disorder in general.
Once I took a test for borderline PD, and the result was that I very likely have it.
The same thing happened with a test for multiple personality disorder, one for schizophrenia, one for OCD and another one for narcissistic PD.
They're silly and unaccurate, and I'll never take one of them seriously.

I find it's useful but as others have said, not enough to distinguish whether one has AS or not for sure. The best way to find out is to see a professional but even then it's not 100% foolproof, so you can see online tests have even more limitations.

If you take the same test more than once you may find your results change. As an example I did the AQ 4 times with results of 32, 24, 22 and 30 (most recent). While it could indicate you've changed over time, there's also a good chance it's influenced by your state of mind at the time of taking the test. I find that sometimes when doing such tests there are times where I'm stuck between two choices (eg being given the choices 0, 1, 2 and wanting to input something like 0.5 or 1.5) and then having to 'lean' towards one way (which often causes me to lean in the same way on other borderline questions), even if I'm trying to be honest.

There are real clinicians tools such as the AQ, EQ, SQ, FQ etc, from the Autism Research Centre so they are not all "silly". The point is that they cannot be trusted alone, they are to be used by professionals trained in using them.

l really hope people don't use these to self diagnosis. l'm sure l've taken all of them.

lf people don't like being them described as "silly", maybe consider that they are useful indicator that you have autistic traits but in no way a qualified diagnostic tool by themselves. Clinicians don't really on these tests alone, if they do they are sloppy.

The AQ and the others will pick up a lot of traits also related to OCD and sensory issues IME, as well as general introversion. l scored as an aspie from HELL on the RAADS-R and got a 33 AQ.

l'd say many other NT introverts with a few quirks and especially ADHD could have similar scores. Even underlying disorders like stealth, undiagnosed dyslexia(which in reality l probably do have and was mislabled with ADHD in the first place) cause traits and sensitivities very similar to AS in some cases.

l DO think these tests easily sort an averag NT from a group of related disorders like ADHD,OCD, etc very well but that many disorders will score as AS on these tests.

If you mean me - erm, that's exactly what I said, before you, if you read my posts.

ok. l honestly don't read all posts, l can of scan and see words pop out. l did read your post afterward. Sorry...l wasn't associating it with your username or anything.

edit:you're basically saying what l was anyway. l mean...sorry lol. Didn't think it would be an issue.

It depends on the test.

Many of these tests are the exact same ones that are used in the research literature. If you plug the test name into Google Scholar, you'll find out how accurate they are.



Actually, the opposite is more likely. See this link. The AQ has a high validity, but false negatives occur for a subset of autistics.



The Aspie Test is not a test from the research literature, it's a test designed by an Aspie out of his own interest. However, from what I've seen, the results from it often agree with the research literature tests, so it's probably pretty accurate.

The BAP test is accurate, but the cutoffs are designed to detect BAP, not autism, so scoring above cutoff does not actually mean you're autistic. (Though an autistic person would score above cutoff, so would many people who don't meet criteria for autism.)

RAADS-R is scientifically validated as a screening measure for autism spectrum, and according to research, it has good sensitivity and specificity (meaning it detects who does and doesn't have autism well).

It's important to keep in mind, though, that no screening scale is a replacement for a full assessment. The main reason is that screening scales mostly distinguish AS from NT, but not AS from other clinical conditions. For example, a person with schizophrenia will often score in the AS range on screening scales, but would not be considered AS unless the AS clearly predated the schizophrenia and had an early childhood onset. In addition, some people misunderstand the scales or lack good self-awareness, and a psychologist can separate those things out through observing behavior, checking understanding and talking to informants (eg parents).

The RAADS-R is probably the best self-report test for HFASD in adulthood. The AQ for BAP. But these tests are best used in conjunction with behavioral observation by a psychologist to diagnose ASD.

As others have mentioned, both the AQ and RAADS-R are scientifically validated screening tests. As screening tests the scoring is biased to give more false positives than false negatives. If you score above the thresholds in either of these tests then you display enough characteristics to warrant further investigation. That is all they are designed to do. They are not designed to be true diagnostic tools.

Having said that, if you suspect you are on the autistic spectrum and you score reasonbly high on the AQ, RAADS and Aspie test it is pretty likely you are on the spectrum. If you score over 120 on the RAADS I personally think it is very likely.

As a side note, one of the authors of the RAADS-R is now analyzing over 1600 test results from the online version of the RAADS-R on the Aspie Tests website. I am hoping he will get back to me with the results at some point.

This. I've only done one online test, the fairly popular "Aspie Quiz". I took it a total of 5 different times and then averaged the scores. If I were to take another test (probably the AQ), I would do the same.

Also, while Aspie Quiz is not a professional screening tool, it is at least as accurate as the AQ and has been validated through multiple iterations and widespread usage.

Many online tests are not very accurate and are mostly hobby creations, but the professional screening tools are valid, although they are not used to diagnose.

I actually have a fairly narrow range of AQ scores.

A few years ago I took it and had no idea what several of the questions were supposed to mean along with some Dunning-Kruger effects on my perception of my social skills. I scored 24. Then I took the Aspie Quiz, which asked similar questions more clearly and scored something in the 150s for Aspie, and went back to the AQ and scored in the high 30s. I think lowest to highest for me has been 37-42.

Agreed on having trouble with the needing to pick 0 or 1 when I want 0.5. I hate that.