Could I be neurodivergent?

I agree that screening tests have to strike a balance, but that doesn’t make the RAADS-R more definitive.

In principle, anyone could design a test that has 100% sensitivity: it would just flag everyone as potentially autistic. That would catch every autistic person, but only because it caught everyone else, too.

Sensitivity (how good a test is at correctly flagging those who do get a diagnosis) has to be considered alongside specificity (how good it is at correctly flagging those who don’t). 

In the study, specificity was only 3.03% - ie almost everyone who was not diagnosed as autistic also scored above the test’s threshold.

Per what I said in response to your earlier reply, a lower-than-needed threshold wasn’t the issue here, because the non-diagnosed group had a higher median score.

Hence the researchers' conclusion that, when used as a self-report tool, it had “no clinical value”.

Martin said:

That is because the RAADS-R cutoff  (>65) has been set, admittedly by the developers, at a ridiculously low level. I have an ASD diagnosis and scored 167. Had the testers set the cut off at a more reasonable level, say 100, the predictability of the test would, obviously, improve. Those scoring 65 and over have a 34.7% chance of receiving a clinical diagnosis. Increase the cut-off value and the test becomes automatically more predictive with no changes in the wording of the test.

That doesn’t follow from the findings. Raising the cut-off would only make the test more predictive if higher scores were meaningfully associated with getting an ASD diagnosis. But the study found the opposite:

• Median score of those who were not diagnosed as autistic: 154
• Median score of those who were diagnosed as autistic: 138

So the finding wasn’t that the cut-off was too low. It was that the RAADS-R didn't meaningfully separate the ASD and non-ASD groups. So raising the cut-off wouldn't have obviously improved its predictive value.

Martin said:

"Negative predictive value (NPV) determined that 100% of those who did not score above the threshold did not receive a clinical diagnosis", which is in itself useful. 

The test is not at fault, it is the over-optimistic threshold level that most definitely is.

The NPV was based on one person - the only one who scored below 65 - so I’d be cautious about placing too much weight on that.

That also sits alongside their finding that there was only a 3.03% specificity in detecting the absence of ASD in those who didn't receive a clinical diagnosis.

All of this explains why their conclusion was: “When used as a self-report tool, the RAADS-R had no clinical value.”

All such tests have to strike a happy medium between two non-complementary goals, including all autistic people and excluding all non-autistic people. No test can do both perfectly. The RAADS-R test cut off has gone for including all autistic people at the expense of not excluding all allistic people. In the paper the test identified 100% of the people who were subsequently diagnosed. I very much doubt that the AQ50, with fewer questions is in any way superior in this regard.

Yes, just to be clear, I do have a formal diagnosis but this was just for myself rather than for the purposes of proving it to anyone else. I'm not sure why I needed it either but I haven't interrogated that too much. It felt important to me so I did it.

Thank you, this is great to know that they needn’t be involved 

Thank you, this is helpful. I did find many questions ambiguous and almost wanted to ask the person for more clarity and possibly more options for potential answers to click…..maybe that’s all the answer I need 

I am diagnosed with ASD and score between 37 and 40 on the AQ50, depending on how I treat ambiguous and situational questions. I score 167 on the RAADS-R. I was told that I would be getting an ASD diagnosis about 20 minutes into the assessment, so was fairly obviously autistic just from my questionnaire answers.

I had my wife support my assessment. Anyone who knows you fairly well can do this. Evidence for the existence of early childhood traits can come from yourself, it is not a court of law requiring witness statements. Or at least it should not be, though some assessors can harbour very strange notions.

That is because the RAADS-R cutoff  (>65) has been set, admittedly by the developers, at a ridiculously low level. I have an ASD diagnosis and scored 167. Had the testers set the cut off at a more reasonable level, say 100, the predictability of the test would, obviously, improve. Those scoring 65 and over have a 34.7% chance of receiving a clinical diagnosis. Increase the cut-off value and the test becomes automatically more predictive with no changes in the wording of the test. "Negative predictive value (NPV) determined that 100% of those who did not score above the threshold did not receive a clinical diagnosis", which is in itself useful. 

The test is not at fault, it is the over-optimistic threshold level that most definitely is.

Your AQ50 score is high enough to be referred by a GP, if you wanted to try that route.

Wanting external confirmation is fine. Doubting yourself and having imposter feelings is normal, even after formal diagnosis.

Thank you so much, appreciate your reply. 

I’ve done the RAAD-R test today, result 174 and the AQ50 scored 42, both of which are seemingly pretty high. I’m unsure why I feel the need for formal diagnosis but I guess I doubt myself.

Hi LynnG, a lot of those certainly sound like autistic traits and some sound like ADHD. I am AuDHD myself and can relate to almost everything you said. Martin has recommended the tests at Embrace Autism and these are really good, plus there's loads of other useful stuff on there. I would recommend you do a few of the tests and see what they come up with.

I can sympathise with family not taking neurodivergence seriously. I haven't bothered mentioning it to any of my family apart from my wife, as my parents are, I feel, very unlikely to take it seriously. There is a chance I could educate them and get them to understand it more but, to be quite frank, I can't be bothered! I am autistic and don't have any interest in proving it to anyone.

Wholeheartedly agree about it not being a fashion but this is what I’m up against,’especially when a formal diagnosis may require the same family members involvement. Thank you for your kind reply 

It depends on which neurodivergency you mean as it's an umbrella term for quite a lot of conditions.

If you are talking about autism, I'd start by looking at the diagnostic criteria and see if you fit and as others have said, the tests available, rather than a long list of your own characteristics that might, or might not, fit into one or another condition.

https://www.autism.org.uk/advice-and-guidance/diagnosis/assessment-and-diagnosis/criteria-and-tools-used-in-an-autism-assessment

All the best.

Thank you for your kind reply. 

Hello.

This would seem consistent with having ASD, but only a clinician can diagnose officially through a formal assessment.

If you wanted to get more of a feel, you may like to do some of the questionnaires. The AQ50 is the main autism screening test (50 questions). Personally, I think struggling with the questionaires and overthinking them is enough on its own, but that's an aside.

A good single place to look for them as they are all collected together is the embrace-autism.com website. They also give info on the scores and an indication of how reliable they are.

Note, the questionaires don't prove anything, but they can have high correlation.

Martin said:

but the RAADS-R seems to be more definitive

Recent research - involving NHS autism assessments - has thrown doubt on its validity:

"the RAADS-R was not able to differentiate between service users who receive an ASD diagnosis after full assessment and those who do not.

When used as a self-report tool, the RAADS-R had no clinical value."

The Effectiveness of RAADS-R as a Screening Tool for Adult ASD Populations

There is nothing 'fashionable' about neurodivergence. Mere fashion does not make people wade through the obstacle course and ridiculous waiting times that is the NHS route to diagnosis, nor pay the very considerable sums of money needed for a private assessment. A real need to find answers to lifelong issues is what drives people to seek a diagnosis.

My advice would be to take some online autism tests. AQ10 and AQ50 (sometimes called just 'AQ') are the most commonly used by clinicians in this country, but the RAADS-R seems to be more definitive. They can be found here:  https://embrace-autism.com/autism-tests/
If you score in the autism range it will give you more certainty about following up on a diagnosis.

Hi LynnG and thanks for posting. Welcome to our community. Each neurodivergent person is unique but much of that long list resonates with me (I am autistic and ADHD) and your research on the topic is perhaps another marker in itself. My recommendation is not to be put off by family members but continue researching and doing any of the established tests you can find. I found Embrace Autism to be a good source for resources and tests. Of course the National Autistic Society is excellent too, so have a good look around their site and the threads on here. Pursuing diagnosis via your GP is a personal choice, and some people prefer to self-identify. Good luck and do keep posting and asking questions.

HI Lynn, and welcome to the online community!  I am sure you will find a lot of friendly and helpful people here.

While waiting for others to share their thoughts, you may like to have a look at our About Autism page which gives a lot of information on many core characteristics of autism.  More information can also be found on that here.

I hope you find these links helpful!