Show the full module text
Module 1 — Am I Allowed to Call Myself This?
Where self-identification came from, what the research actually found about the people who do it, the strongest case against it, and how to hold a description nobody handed you.
Self-Identification · Part One — Am I allowed?
The big idea
Short on capacity today? The big idea: self-identification and an in-depth assessment are two different things a person can do, not two rungs of a ladder. The research finds that adults who self-identify meet the same trait thresholds as diagnosed adults, report worse wellbeing, and are mostly stuck in a queue or priced out. A trait threshold is still not a diagnosis. Hold the description firmly enough to act on and loosely enough to revise.
Step 1 — The lesson
A. The question underneath the question
Almost nobody asks am I allowed to call myself this about a broken wrist.
The question only arises when a description does two jobs at once. Autism and ADHD are, at the same time, a clinical category with criteria and a gatekeeper, and a way of accounting for your own life to yourself. The first has a queue in front of it. The second does not, and never has.
When somebody says you cannot call yourself autistic without a diagnosis, they are usually talking about the first job. When somebody says I have known since I was thirty-four, they are usually talking about the second. Both statements can be true in the same room. Most of the heat in this argument is generated by not noticing that.
Diagram — A · One word, two doors. The description does two jobs at once, and the figure puts both doors in the same wall under the same sign. The clinical category has criteria and a gatekeeper, has a queue in front of it, and is what people are defending when they push back. Your own account of your life has no gate and never has had one, is available today and costs nothing, and is what you are actually describing. Both statements can be true in the same room. The question is not whether you are allowed; it is what the description is for, in your hands.
There is a real question in here, and this module takes it seriously rather than reassuring you past it. But it is not the question am I allowed. It is: what is this description for, in my hands, and what would I do differently if I turned out to be wrong?
B. Who actually self-identifies
This has been studied, which surprises people on both sides of the argument.
In 2026, researchers compared 147 U.S. adults who identified as autistic without a diagnosis against 115 who had one, on traits, wellbeing, support needs and beliefs about autism.
The headline finding is the one worth sitting with. More than 93 per cent of both groups scored at or above 14 on the RAADS — the same threshold, on the same instrument, that Module 2 of the autism course is built around. On the traits themselves, the two groups were not meaningfully different populations.
Three other findings from the same study matter here.
Self-identifying adults were more likely to be women — 53.5 per cent cisgender female, against 27.0 per cent of the diagnosed group.
They reported worse mental wellbeing, and were far more likely to have support needs going unmet: 71.4 per cent, against 47.0 per cent.
And 68.7 per cent of them wanted a diagnosis and could not get one.
Diagram — B · The same line, a different load. Ahuvia and colleagues, 2026, compared 147 U.S. adults who identify as autistic without a diagnosis against 115 who have one. More than 93 per cent of both groups scored at or above 14 on the RAADS, the same threshold on the same instrument, which is why both figures in the drawing stand at the same line. Support needs went unmet for 71.4 per cent of the self-identifying group against 47.0 per cent of the diagnosed group. The self-identifying group was 53.5 per cent cisgender female against 27.0 per cent of the diagnosed group. And 68.7 per cent of the self-identifying group wanted a diagnosis and could not get one. Same traits, worse wellbeing, less support.
That last number is the one that dissolves the caricature. The picture people argue about — a person casually adopting a label because it is interesting — is not what the data describes. It describes people who meet the trait threshold, are doing worse than the diagnosed group, are disproportionately women, and are mostly stuck in a queue or priced out.
C. The case against, taken seriously
If this module only gave you the evidence above, it would be doing the thing these courses were built to avoid.
So: there is real evidence that a high score on a self-report trait measure, in an online sample, does not identify the same population as a clinical assessment does.
A 2025 study compared 56 adults with clinically confirmed autism, 56 online participants with high self-reported autistic traits, and 56 with low traits. The high-trait online group reported substantially elevated social anxiety and avoidant personality symptoms relative to the clinically assessed group. On behavioural tasks, the two groups came apart: only the clinically assessed group showed the specific differences the tasks were designed to detect.
And one finding from inside that study should give everyone pause. Within the clinically assessed sample, self-reported autistic traits had no significant relationship with the traits a clinician rated using a structured observation. How autistic you feel and how autistic you look to an assessor are not the same measurement — in either direction.
Diagram — C · A balance with a broken beam. Banker and colleagues, 2025, compared 56 adults with clinically confirmed autism, 56 online participants with high self-reported autistic traits and 56 with low traits. A high score on a self-report trait measure, in an online sample, does not identify the same population that a clinical assessment does. The high-trait online group reported substantially elevated social anxiety and avoidant personality symptoms relative to the clinically assessed group, and on the behavioural tasks only the clinically assessed group showed the specific differences the tasks were designed to detect. Inside the clinically assessed sample, self-reported autistic traits had no significant relationship with the traits a clinician rated using structured observation — which is why the beam between the two pans is drawn snapped. A trait threshold is not a diagnosis, and the strength of your own recognition is not evidence of how an assessment would go.
Read together, the two studies say something more useful than either alone.
People who self-identify are not making it up — they meet the same trait thresholds. But a trait threshold is not a diagnosis, social anxiety produces a very similar self-report pattern, and the strength of your recognition is not evidence of how an assessment would go.
Which is, almost word for word, what the screener courses say about every questionnaire in them.
D. Why the queue exists
The reason self-identification became a norm in this community is not philosophical. It is administrative.
In England, in June 2026, 294,792 people had an open referral for suspected autism. Of those, 256,017 — 86.8 per cent — had been waiting at least thirteen weeks. Of the people waiting longer than thirteen weeks, 3.8 per cent had been seen within the thirteen weeks the guidance recommends.
Diagram — D · The corridor, June 2026. NHS England, Autism Statistics, June 2026 position. In England, 294,792 people had an open referral for suspected autism. Of those, 256,017 — 86.8 per cent — had been waiting at least thirteen weeks, which is why the thirteen-week line is painted near the front of the corridor and almost the whole queue is behind it. Of the people waiting longer than thirteen weeks, 3.8 per cent had had a first appointment within the thirteen weeks the guidance recommends. The metric counts first appointments rather than completed assessments. These are people who did everything they were told to do: they asked, they were referred, they are in the system, and the system is not moving.
Those are people who did everything they were told to do. They asked. They were referred. They are in the system, and the system is not moving.
Where there is no queue there is a price instead. A private adult assessment is typically a four-figure commitment, and the range is wide enough that the same person can be quoted twice the figure a hundred miles away.
Then there is who gets missed in the first place. In childhood, girls meet criteria at roughly half the rate of boys, and the ratio moves close to equal in adulthood — a shift that points at recognition rather than biology. Clinical referral ratios have run anywhere from 3:1 to 16:1 in favour of boys against a community ratio of around 3:1. A description built from the children who were referred was never going to fit the ones who were not.
None of that makes self-identification correct. It makes it predictable — and it explains why 68.7 per cent of the self-identifying group in the 2026 study wanted the assessment they could not have.
E. So what is the word doing?
Here is the practical position this course takes, and everything after this module rests on it.
Self-identification and an in-depth assessment are two different things a person can do. They are not two rungs of a ladder with the assessment at the top. They answer different questions, they cost different amounts, and they buy different things.
An assessment buys you a clinician’s judgement, a document, and — in most places — access to accommodations and legal protection that a self-description does not unlock. That is not a small list, and Module 11 is about deciding whether you want it.
Self-identification buys you an explanation you can use on Monday. It is available today, it costs nothing, and it is the thing that actually changes how you run a week.
What self-identification does not buy is certainty. Nor does an assessment, entirely — but it buys more of it. Holding a description honestly means holding it the way you would hold any other good explanation: firmly enough to act on, loosely enough to revise.
Module 2 takes the word apart properly — because “label” is doing four separate jobs, and once you can see which one is being argued about, most of the argument stops.
Diagram — E · A fork, not a ladder. Self-identification and an in-depth assessment are two different things a person can do, not two rungs of a ladder with the assessment at the top. They answer different questions, they cost different amounts, and they buy different things. An assessment buys a clinician’s judgement, a document, and in most places access to accommodations and legal protection that a self-description does not unlock. Self-identification buys an explanation you can use on Monday: it is available today, it costs nothing, and it is the thing that changes how you run a week. What it does not buy is certainty, and holding a description honestly means holding it firmly enough to act on and loosely enough to revise.
F. What helps
Five things to do with the question, rather than five answers to it.
1. Say it in the sentence that is actually true for you.
“I am autistic” and “I think I am autistic and I am working it out” are both honest, and one of them is easier to say out loud in a year when you are surer. There is no rule that you must pick the strongest available claim on day one.
2. Write down what you would do differently if you were wrong.
This is the single most useful exercise in the module. If the answer is nothing much — I would still need quiet after work, I would still need instructions in writing — then the accommodation you are making is not resting on the label, and the stakes of the question are lower than they feel. If the answer is I would stop looking for another explanation, then keep looking.
3. Notice which job you are being challenged on.
When somebody pushes back, they are almost always defending the clinical category. You are almost always describing your life. Naming that out loud — I am not claiming a diagnosis, I am telling you what I have worked out — ends more of these conversations than arguing does.
4. Take the anxiety possibility seriously rather than defensively.
The 2025 study found substantially elevated social anxiety and avoidant symptoms in the high-trait online group. That is not an accusation, and it is not either/or — plenty of autistic people are also anxious. But if you have never had anyone look at that question properly, it is worth looking at, and it is a reason to keep going rather than to stop.
5. Do not carry it alone for a year.
The self-identifying group in the 2026 study reported worse wellbeing and more unmet support needs than the diagnosed group. Whatever you end up calling this, the isolation is doing damage on its own account. Modules 9 and 10 are about that.
Up next
Module 2 — What a Label Is, and What It Is Not