Neurodivergent Reflection Profiler

Observed: in ADHD, performance depends unusually heavily on how interesting, urgent or novel a task is, rather than on how important it is. Time estimation and forward planning are affected, and difficulty is with doing what is known rather than with knowing it.

Models: Barkley (1997) put behavioral inhibition at the center, with working memory, self-directed speech, emotional self-regulation and reconstitution downstream of it — this is where the now-common phrase “time blindness” comes from. Volkow and colleagues (2009) found differences in dopamine markers in the reward pathway in adults with ADHD, which is consistent with motivation being the bottleneck.

Not established: that any individual's difficulty can be attributed to a specific dopamine level, or that a scan or blood test could show it. There is no biomarker for ADHD, and group-average differences do not transfer to individuals.

Observed: sensory differences are common enough in autism that they are part of the DSM-5-TR criteria, and they occur in people with no diagnosis at all. Both over- and under-responsiveness turn up, often in the same person across different channels.

Models: several predictive-processing accounts converge on the same idea from different angles — that sensory input is weighted more heavily relative to prior expectation. Pellicano and Burr (2012) framed it as weaker priors; Van de Cruys and colleagues (2014) as inflexibly high precision on prediction error; Lawson, Rees and Friston (2014) as aberrant precision. They make overlapping but not identical predictions, and which one fits best is unresolved.

Not established: a specific thalamic “gating failure” as the cause. That is a plausible implementation, not a finding, and it is often stated on pages like this one as though it were settled.

Practical upshot that does not depend on the model: an environment that is genuinely harder to be in is harder to be in. Changing the environment is the intervention with the best return.

The model: Murray, Lesser and Lawson (2005) proposed that autistic attention tends to be allocated to few interests at a time, very deeply, rather than spread across many. A single account then covers a lot of ground: intense focus, the cost of interruption, difficulty with transitions, and the flat exhaustion of environments that demand constant switching.

Why it is worth knowing: it reframes the same behavior without pathologizing it. Trouble stopping is the same mechanism as being very good at the thing you cannot stop doing.

Status: influential, developed substantially by autistic researchers, and increasingly studied — but a theoretical framework rather than an established mechanism.

Double empathy: Milton (2012) argued that the breakdown between autistic and non-autistic people is a two-way mismatch of expectation, not a one-sided deficit in social ability. Crompton and colleagues (2020) gave it empirical support: information passed along a chain as effectively among autistic people as among non-autistic people, and less well in mixed chains.

Masking: Hull and colleagues (2019) developed and validated the Camouflaging Autistic Traits Questionnaire, which is the measure in this area — a real instrument, which the items on this page are not. Camouflaging is consistently associated with exhaustion and poorer mental health in that literature. Most of it is cross-sectional, so “associated with” is doing real work in that sentence.

Burnout: Raymaker and colleagues (2020) produced the first formal definition of autistic burnout from autistic people's own accounts — chronic exhaustion, loss of skills that were previously reliable, and reduced tolerance to stimulus. That is a qualitative definition, which is the appropriate first step and not yet a measure.

Well supported: emotional dysregulation is common in ADHD, frequently the most impairing part for adults, and under-represented in diagnostic criteria written around childhood behavior.

Has a research base: rejection sensitivity as a personality construct, with a validated questionnaire, from Downey and Feldman (1996).

Does not: “rejection sensitive dysphoria” as a distinct entity. It is a clinical-community term popularised in ADHD writing, absent from the DSM-5-TR, with no validated measure and no real peer-reviewed evidence base. It is a useful word for an experience people recognize instantly. That is different from being a thing that has been shown to exist as described.

Not established: the amygdala-and-prefrontal-cortex account of it that circulates widely. Nobody has imaged this. It is left out here for that reason, not because the experience is doubted.

Kapp and colleagues (2019) asked autistic adults directly. Repetitive movement was described as a useful self-regulatory tool, and the main harm reported came from being stopped from doing it — the suppression, not the movement.

This is the item on this page most likely to be scored as a problem and most likely not to be one. Unless a behavior causes injury or genuinely prevents something the person wants, the default is to leave it alone.

The DSM-5-TR lists nine inattentive and nine hyperactive-impulsive symptoms. A diagnosis needs six or more in a domain for children up to 16, or five or more from age 17 — plus several symptoms present before age 12, impairment in two or more settings, clear interference with functioning, and the judgement that nothing else explains it better.

The items here are paraphrases, deliberately. The DSM-5-TR criteria are copyrighted text belonging to the American Psychiatric Association, and reproducing them verbatim in a web page is not something this site will do. Paraphrase also matters clinically: the official wording is written for a clinician who is trained to interpret it, and a layperson reading it cold reads it differently.

What this page counts: how many symptoms you rated “often” or “very often”, in each domain, against the threshold for the age band. That is the symptom count and nothing more.

What it cannot do — all of it required for an actual diagnosis:

• Criterion E: establish that nothing else explains the picture better. This is most of the work, and it needs a clinician.
• A developmental history, and where possible a school report or a parent's account from childhood.
• Direct observation, and collateral information from someone other than the person answering.
• Any judgement about whether reported difficulty amounts to clinically significant impairment.

Meeting the count here means your answers have the shape the criteria describe. That is a good reason to book an appointment. It is not a result.

Winnie Dunn's model (1997) crosses two axes: neurological threshold (how much input it takes to register) against behavioral response (whether the person acts to change their input or lets it happen). That produces four quadrants:

• Low registration — high threshold, passive. Misses things: names not heard, bumps not noticed.
• Sensation seeking — high threshold, active. Goes and gets input: movement, noise, texture, strong flavours.
• Sensory sensitivity — low threshold, passive. Notices everything and is worn down by it, without necessarily escaping it.
• Sensation avoiding — low threshold, active. Actively limits input: leaves the room, cuts the tags out, dims the lights.

Most people sit somewhere on all four. A high score in two apparently opposite quadrants is normal, not a contradiction — thresholds differ by channel, so seeking movement while avoiding noise is an ordinary combination.

This is not the Sensory Profile 2. That is Dunn's own published, normed, proprietary instrument, distributed by Pearson, with age-banded norms and a scoring manual. The items here were written for this page against the published model, carry no norms, and produce no comparable score. Where the quadrants matter — an occupational therapy referral, a school accommodation plan — the real instrument is what gets used, administered by someone qualified to interpret it.

The original version of this page offered an “ADOS-2 Observational Coding Adaptation.” That framing has been removed, and this is the most substantive change in this rewrite.

The ADOS-2 (Lord et al., 2012) is a semi-structured assessment in which a trained clinician runs standardised activities designed to create social opportunities, and codes what the person does in response. The codes depend on the press: what is being measured is behavior under specific, controlled conditions, scored by someone trained to research reliability on that instrument. It is also proprietary, and restricted to qualified users for good reasons.

A questionnaire cannot reproduce any of that. A self-report cannot observe the person answering it, and an observer report is a memory rather than an observation. Calling such a thing an ADOS-2 adaptation implies a correspondence that does not exist, and someone could reasonably take a result from it as meaning something about an autism assessment. It does not.

What is here instead: fourteen plain-language probes about social communication, reciprocity and repetitive behavior — the behavior domains the diagnostic criteria describe, which are public. They produce a description you can take to an appointment. They produce no score, no threshold and no classification, because none of those would mean anything.

If an autism assessment is what you are after, validated screening questionnaires exist and are free or cheap: the AQ and AQ-10 (Baron-Cohen et al., 2001 and after), the RAADS-R for adults, the SCQ for children. Those have published psychometrics. Ask a clinician which is appropriate; each is a screener too, not a diagnosis.

• No validation. No reliability study, no factor analysis, no comparison against any established measure. The five dimensions are a sensible organizing scheme, not empirically derived factors.
• No norms. “62% of maximum” is not high or low relative to anyone, and no amount of care changes that — there is no comparison group and this page will never have one. What is done about it: the report leads with a within-person ranking (your five areas against each other), which is interpretable without any reference sample, and the percentage sits below it with an explicit denominator.
• Arbitrary thresholds. Every band label on this page is a judgement call made while writing it, and is labeled as such where it appears.
• Acquiescence bias. Nearly every item is worded so that agreeing indicates more difficulty, so someone inclined to agree scores higher throughout. What is done about it: five Tier 2 items are worded the opposite way round, one per dimension, and the report gives an agreement count. They are deliberately kept out of the dimension scores — folding them in would change what every other number on the page means. Five items is a weak check rather than a validity scale, and it is reported as a plain count instead of being dressed up as one.
• No validity scales. Nothing detects inconsistent, careless or deliberately skewed responding.
• Frequency is not impairment. The items ask how often; the DSM asks how often and how much harm. A single impact item per tier is a gesture at this, not a solution.
• Self and observer reports diverge, and this tool cannot tell you which is closer to the truth. Often neither is.
• Overlap. Trauma, anxiety, depression, sleep disorders and several medical conditions produce these same answers, and this tool cannot separate them — that is a clinician's job and it stays one. What is done about it: a twelve-item rule-out screener asks about sleep and apnea, hearing, vision, iron, thyroid, mood, trauma, medication and plain overload, then turns whatever it flags into a list naming the actual test to ask for — ferritin and a full blood count, a thyroid panel, audiometry, a full eye exam, a sleep study. It diagnoses nothing. It exists to stop a fixable cause going unexamined because attention was the thing everybody looked at first.
• Cultural and linguistic narrowness. Written in English, in one cultural frame. What reads as poor eye contact or excessive directness varies enormously by culture.
• Children's items are read by adults. The child self-report wording assumes a reading level not every ten-year-old has.

None of that makes the exercise pointless. Writing down what happens, how often, and in which areas is genuinely useful preparation. It is only worthless if presented as measurement, which is why it is not.