Module 6 — RBQ-2A — Repetitive Behaviour and Stimming, Explained Kindly

 

What the twenty questions actually ask, why stimming is better understood as regulation than as a symptom, how insistence on sameness tracks uncertainty rather than stubbornness, and what this score honestly cannot tell you.

 

Autistic Self-Discovery · Part Three — Masking and the Shape of Your Mind · 13 min read, about 26 min with the workbook

 

The big idea

 

If you have ever eaten the same lunch every day for a year, taken the longer route because it is the route, or looked down and found your hand doing something with a pen that nobody asked it to do — this module is about the part of autistic life that gets written down as symptoms and is very often doing a job. The questionnaire counts these things. It does not judge them, and neither does this page.

 

Routine is not the illness. Routine is what a nervous system builds when the next hour is unpredictable and the cost of being wrong is high. If you want the routines to loosen, the thing to work on is almost never the routine.

 

The instrument

 

RBQ-2A — Adult Repetitive Behaviours Questionnaire-2

 

Who it is for: Adults who want stimming and routine counted plainly, without either being treated as a symptom

Length: 20 statements, 5–7 minutes

What you get back: One total and the band it falls in

The cut-off line: 36 of 60. No research threshold exists for this questionnaire; the bands here are editorial, and Section D says so.

 

People usually arrive at this one because…

You have eaten the same lunch every day for a year and never thought of it as anything.

You look down and find your hand doing something with a pen that nobody asked it to do.

A plan changing at short notice drops your stomach in a way other people do not seem to feel.

You were told you could not be autistic because you never lined up your toys.

You spent childhood being told to sit still and you want to know what the movement was doing.

 

What it does well. It gives a vocabulary and a rough sense of scale for something many people have never had words for, and it asks only how often, not whether it is odd. It also splits the ground into two halves that behave differently: what the body does, and what the week does.

Where it falls down. It has no norms and no published sensitivity or specificity, so nothing in the literature supports reading any score as pointing at autism. The group difference it rests on came from 29 autistic adults compared with 37 non-autistic adults.

How much weight it carries. Tier 1 of 4 — the best evidence in this course — published, and independently tested more than once.

A low score is not a no. The two halves do not have to rise together, and one number flattens the difference between someone whose hands are never still and someone who cannot eat off the wrong plate. It is also a load reading as much as a trait: the sameness half climbs when the next month gets less predictable, so a quiet year reads lower than a hard one.

 

FAQ

 

Is stimming something I should be trying to stop?

The autistic adults who have been asked directly say no. In interviews and focus groups with 32 of them, stimming was described as a way of regulating intense emotion and thought, and as a group they opposed interventions meant to eliminate it. Holding it in because of how other people react has a cost, and this questionnaire counts the behaviour without asking you to justify it.

Everybody has routines. Does scoring on those mean anything?

Almost everybody scores something, and the questionnaire was never built to say otherwise: in the first sample of ordinary adults the average was 1.51 on the published one-to-three scale. What it looks at is how much of a day these things take up and what it costs when they are interrupted. Routine and a favourite subject are not autistic property.

What happens to my answers?

Your screener answers and your score are kept, so that if you ever want to talk the result through with someone, it is there to look at. Your workbook is different: it saves in this browser only and never reaches us.

 

Step 1 — The lesson

 

A. Two families, twenty questions

 

The RBQ-2A asks about twenty things you might do, and about each one it asks the same dull, useful question: how often. Not whether it is odd. Not whether it bothers anyone. How often.

 

The twenty cover more ground than most people expect. Rocking or pacing. Fiddling with an object — a pen, a ring, the cord of a hoodie. Repeated hand or finger movements. Being drawn to a particular texture, a particular sound, the way light moves on water. And then a second sort of thing entirely: needing the same route, the same seat, the same order of tasks; eating the same meal for months; the drop in your stomach when a plan changes at short notice; one interest that quietly takes up a great deal of your attention.

 

In 2015 Sarah Barrett and colleagues gave the questionnaire to 161 adults and looked at how the answers grouped. They did not group into one thing. They grouped into two, which the authors named Repetitive Motor Behaviours and Insistence on Sameness.¹

 

Diagram — A · Two families, twenty questions. The left column is what the body does. The right column is what the week does. They are counted by the same questionnaire and they are not the same thing — which is why one total number is a poor summary of either.

 

Two things follow, and both matter more than whatever total you were handed. The first is that almost everybody scores something. In that first sample of ordinary adults the average total was 1.51 on a scale running from 1 to 3.¹ Routine, fiddling and a favorite subject are not autistic property. What the questionnaire looks at is how much of a day they take up, and what it costs when they are interrupted.

 

The second is that the two families do not have to rise together. You can be someone whose hands are never still and who is genuinely relaxed about a canceled plan. You can be someone who has never rocked in your life and who cannot eat off the wrong plate. A single total cannot say which of those you are, and the rest of this module keeps pulling the two apart on purpose.

 

B. It is doing something

 

For most of the history of this field, everything in the last section was recorded as a symptom — a behavior to be counted and, in a great deal of early intervention work, reduced. That framing has not survived contact with the people doing the behavior.

 

In 2019 Steven Kapp and colleagues ran interviews and focus groups with 32 autistic adults about stimming. What came back was not an account of a deficit. Participants described stimming as a way of regulating themselves — of soothing or letting out intense emotion and thought — and as a group they opposed interventions designed to eliminate it.² The title of that paper is a quotation from one of them: "People should be allowed to do what they like".

 

A second study came at the same question from the other end. Emma Collis and colleagues interviewed twelve autistic adults about repetitive behavior generally, and four themes came out: self-regulation, positive impacts, negative impacts, and suppression. The clinical split between so-called higher-order and lower-order repetitive behavior did not match how participants described their own lives. For some behaviors in some settings the expression was largely, though not wholly, under voluntary control — and negative reactions from other people led several participants to hold it in, at a psychological cost.³

 

Diagram — B · It is doing something. Five jobs, one purpose. This is the reframe that matters most in the module: the behavior on the left-hand side of your score is usually a regulator, and regulators are not removed without something taking their place.

 

Two more recent surveys of autistic adults — 131 people, then 117 — found something the regulation account on its own misses. Stimming also communicates. About 73 per cent of the first sample agreed that when they see another neurodivergent person stimming they can usually tell what that person is feeling, and connection to the autistic community was the strongest predictor of seeing stimming that way at all.⁴ A stim is not only a valve. It is sometimes a sentence.

 

None of this makes a high motor score good news or bad news. It makes it a description of something you have been doing, with a function, whether or not anyone ever told you that is what it was.

 

C. Sameness is usually an answer

 

The other half of the questionnaire needs a different explanation, because insistence on sameness is rarely about liking things tidy. It is far more often about how expensive it is when the next hour is unpredictable.

 

In a study of 176 autistic adults and 116 non-autistic adults, both intolerance of uncertainty and anxiety were raised in the autistic group, and intolerance of uncertainty sat in the middle of the chain: between sensory sensitivity and anxiety, and again between anxiety and insistence on sameness.⁵ The pattern held across age groups.

 

A larger study of 426 autistic adults took the idea further and found the two halves of the questionnaire run on different engines. For repetitive motor behavior the route ran from sensory difference, through alexithymia, through intolerance of uncertainty, to anxiety. For insistence on sameness, intolerance of uncertainty did the work directly — and anxiety played no part in that second route at all.⁶

 

Diagram — C · Sameness is usually an answer. The score at the top is the reading on a gauge. The boxes underneath are what pushes the needle. Two independent studies put intolerance of uncertainty closest to the needle, which is why arguing with the routine itself so rarely works.

 

A third study, of 187 adults across a clinical autistic group, a community autistic group and a non-autistic comparison group, adds the piece about mood. Autistic community participants scored 2.28 on insistence on sameness against 1.42 for the non-autistic group, and were higher on rumination and on obsessing. The link between insistence on sameness and depression-anxiety ran through that repetitive thinking rather than around it, with a total indirect effect of 0.31 (95% CI 0.22 to 0.41).⁷

 

This is also why the same score can mean two different things a year apart. Move house, change job, lose a person, and the sameness half climbs — not because you became more autistic, but because there is more uncertainty to hold. Read a high score as a load reading before you read it as a trait.

 

D. What the number can and cannot tell you

 

Here is the honest part, and it is longer than you might want it to be.

 

The published instrument does not produce a total between 20 and 60. Barrett and colleagues score the RBQ-2A as a mean across the items, running from 1 to 3.¹ Adding the twenty item scores instead gives a range of 20 to 60, which is what the screener on this site reports back to you. It is the same information in different clothes; it is worth knowing which one the research uses.

 

There is no published cut-off. The 2015 paper proposes none. It shows that a group of autistic adults scored higher than a group of non-autistic adults, and it stops there, because that is what the data support.¹ Any band boundary attached to this questionnaire anywhere online — including the three on our own screener page — was chosen by whoever built that page. Read the bands as rough description. Do not read them as a line you have crossed.

 

The validation samples were small. The two components came out of 161 UK university students with an average age of 21. The comparison that showed the questionnaire could separate the groups used 29 autistic adults and 37 non-autistic adults. The separation was large (r = −.67) but 29 people is 29 people, everyone had an IQ above 80, and in the non-autistic group the internal consistency of the Insistence on Sameness subscale was .55, which is poor.¹

 

The two-factor structure is contested. In 2024 an Australian team tested it properly with confirmatory factor analysis in 368 university students and 283 autistic adults. The original two-factor model fitted badly in both samples (RMSEA .075 and .089). A four-factor model fitted well — repetitive motor behaviors, interest in sensations and objects, insistence on sameness, and restricted interests — with hierarchical omega between .77 and .80.⁸ Against that, a 2018 study of 275 autistic adults did recover two components, Insistence on Sameness and Repetitive Sensory and Motor Behaviours, with alphas of .81 and .70.⁹ Two respectable answers to the same question. The field has not settled it, and a page that told you it had would be lying to you.

 

The sex difference is smaller and narrower than the stereotype. A 2024 meta-analysis pooled 25 studies of autistic children, adolescents and adults. Males scored higher on stereotyped behaviors (SMD 0.21, 95% CI 0.09 to 0.33) and on restricted interests (SMD 0.18, 0.07 to 0.29). On insistence on sameness there was no difference at all (SMD 0.01, p = .68), and none on sensory experiences either.¹⁰ The study of 275 autistic adults found no gender differences on this questionnaire.⁹ If you are a woman who was told you could not be autistic because you never lined up your toys, the sameness half of this questionnaire does not agree with whoever told you.

 

A high score is not specific to autism. A study of 260 children aged four to eight, referred for behavioral, emotional or cognitive difficulties but carrying no diagnosis, found repetitive behavior above community levels, and the severity of anxiety symptoms was significantly associated with both subtypes.¹¹ Repetitive behavior is a good deal of what distress does in a nervous system. It is not an autistic signature.

 

Diagram — D · What the number can and cannot tell you. Three ticks, four blanks. The blanks are the honest part of this instrument: nobody has published a cut-off, nobody has reported how often it is right or wrong, and no item asks what any behavior is for.

 

What survives all of that is still worth having. The RBQ-2A gives you a vocabulary and a rough sense of scale for something a lot of people have never had words for, and it splits it into two halves that behave differently. That is a real service. It is not a verdict, and it was never built to be one.

 

INSIDE THE INSTRUMENT

 

RBQ-2A — The Adult Repetitive Behaviours Questionnaire-2

 

Where it came from

 

Developed by Sarah Barrett, Mirko Uljarevic, Emma Baker, Amanda Richdale, Catherine Jones and Susan Leekam, and published in the Journal of Autism and Developmental Disorders in 2015 as the adult self-report member of the Repetitive Behaviours Questionnaire family.¹ In 2024 the same research group published the RBQ-3, a 20-item measure that consolidates the RBQ-2 and the RBQ-2A into a matched self-report and informant-report pair for use across the lifespan, keeping the item content and making the response scaling consistent.¹² For new work, look at the RBQ-3 first.

 

What it is made of

 

Twenty items answered on a mixture of four-point and three-point scales, with the four-point responses collapsed to three for scoring; the published score is a mean between 1 and 3, and the instrument is used dimensionally.¹ The original principal components analysis produced Repetitive Motor Behaviours (6 items) and Insistence on Sameness (5 items) — that is 11 of the 20 items, so nine items contribute to a total that no published subscale covers.

 

How well it performs

 

In the 161-adult development sample, alpha was .83 for the total, .71 for Repetitive Motor Behaviours and .73 for Insistence on Sameness. In the clinical comparison the autistic group's total alpha was .91 and the non-autistic group's .73, with the non-autistic Insistence on Sameness subscale at .55. The autistic group scored higher with a large effect (Z = −5.43, p < .001, r = −.67), mean totals 1.87 against 1.41.¹ In 275 autistic adults aged 18 to 66, a polychoric principal components analysis again produced two components, Insistence on Sameness (alpha .81) and Repetitive Sensory and Motor Behaviours (alpha .70), both correlating moderately with AQ scores (rs = .25 and .42), with the sensory and motor component falling with age (rs = −.21).⁹ For comparison, the successor RBQ-3 reports total alpha and omega of .90 in 110 adults seeking assessment, self-informant agreement of rs = .71, and convergent correlations of rs = .45 to .54 with the restricted and repetitive behavior domain of a blind clinical interview, with a weaker association to the social communication domain.¹²

 

Where it was validated

 

Structure: 161 UK university students, mean age 21.3. Group difference: 29 autistic adults against 37 non-autistic adults, all with IQ above 80.¹ Since then: 275 autistic adults recruited online, predominantly female, predominantly white and mostly diagnosed in adulthood⁹; 368 Australian university students and 283 self-identified autistic adults online⁸; and adult clinical and community samples in the RBQ-3 work¹². No sample in this literature includes autistic adults with intellectual disability or limited language, and the authors of the RBQ-3 say so explicitly.

 

What it cannot do

 

It has no cut-off, no norms and no published sensitivity or specificity. The development paper proposes no threshold and reports no classification statistics.¹ Nothing in the literature licenses a statement of the form "a score above X indicates autism". Any banding you meet online, including ours, is editorial.

 

Its factor structure does not survive confirmatory testing. The original two-factor model fitted poorly in confirmatory factor analysis in both a university sample (RMSEA .075) and an autistic sample (RMSEA .089); a four-factor solution fitted well (RMSEA .036, CFI .989, TLI .982 in the university sample) with hierarchical omega of .77 to .80 and configural, metric and scalar invariance across groups.⁸ Reporting two subscale scores is defensible practice; treating them as established structure is not.

 

It cannot separate autistic repetitive behavior from repetitive behavior driven by anxiety or by repetitive thinking. Insistence on sameness sits downstream of intolerance of uncertainty in two independent adult samples⁵⁶, its link to depression and anxiety is mediated by rumination and obsessing⁷, and non-diagnosed children referred for difficulties show elevated repetitive behavior tracking anxiety severity.¹¹ A raised score is a reason to ask what is being held, not a diagnostic pointer.

 

It records frequency, not function. No item asks what a behavior does for the person doing it. Qualitative work with autistic adults describes self-regulation, communication, pleasure, and a real cost to suppression — none of which the score can see.²³⁴ Two people with identical totals can be in entirely different clinical situations.

 

It has been superseded by its own authors. The RBQ-3 replaces both the RBQ-2 and the RBQ-2A, adds an informant version, and reports stronger psychometrics on larger and more clinically relevant samples.¹² The RBQ-2A remains usable and interpretable; it is no longer the current instrument in this family.

 

What a clinician does with it

 

Use it as a structured prompt rather than a score. Read the two halves separately, then ask the question the instrument does not: what is each behavior for, when did it get heavier, and what happens when it is prevented. Where insistence on sameness is high, formulate around intolerance of uncertainty and repetitive negative thinking rather than around the routines themselves; where repetitive motor behavior is high, ask about sensory load and about suppression in public settings. Do not report a band, and do not put the total in a diagnostic formulation.

 

Validity tier: 1 — validated and published, and re-examined by labs with no stake in it. It was developed by external researchers, tested against clinically confirmed diagnosis, published in a peer-reviewed journal and independently re-analyzed in Australia and the UK — but it is a dimensional measure with no cut-off, no norms and a disputed factor structure, so Tier 1 here means the instrument is real, not that the number is calibrated.

 

Diagram — E · Where your score lands. These are the bands the screener will show you, word for word. They are a reasonable editorial description of a dimensional score and they are not a research threshold, because no research threshold for this questionnaire exists.

 

STRENGTHS LENS

 

You have been regulating yourself for years without a manual and without permission.

 

Look again at what the first column of that questionnaire is actually describing. A body that reaches for movement when the input gets too loud. A hand that finds something to do while the rest of you thinks. A route you do not have to spend attention on, so the attention goes somewhere better. Nobody taught you any of it. You worked it out, mostly alone, mostly while being told to sit still, and it has been keeping you functional ever since.

 

The second column has a similar story. Among autistic adults, about two thirds report having special interests, most of them several rather than one, and the motivation behind engaging with an interest was associated with higher subjective wellbeing and with satisfaction in areas including social contact and leisure; only very intense engagement went the other way.¹³ The depth is not the problem. The problem has usually been a world that treats depth as a phase and stillness as good manners.

 

F. What helps

 

Nothing here is about doing less of what you do. It is about knowing what each thing is for, so you can protect the ones that are working and go after the load underneath the ones that are not.

 

1. Work out what each one is doing before you change anything.

 

Take three behaviors the questionnaire counted and ask, for each: what happens if I do not do it for an hour. If the answer is nothing much, it is a preference. If the answer is that everything gets louder, it is a regulator, and regulators do not get removed, they get replaced or made socially cheaper.

 

2. Read a rigid week as a reading on the uncertainty in it.

 

When routines tighten, look at what got less predictable rather than at the routines. In adult samples, intolerance of uncertainty sits between sensory load and anxiety, and between anxiety and insistence on sameness. Removing one unknown — a confirmed time, a written agenda, a decided answer — usually does more than any amount of trying to be flexible.

 

3. Stop paying the suppression tax where you can.

 

Autistic adults describe holding stims in because of how other people react, and describe a cost for doing it. Find the settings where you can let your hands do what they do, and stock a pocket with the discreet versions for the ones where you cannot yet. Suppression is a fee, not a skill.

 

4. Protect the interest rather than rationing it.

 

The interest is one of the few things reliably associated with wellbeing in autistic adults. Put it in the week on purpose, before it becomes the thing you do instead of sleeping. Rationing it as a reward for getting through the rest tends to produce a week made entirely of the rest.

 

5. Ask for predictability in specifics, not for patience.

 

"I do not cope well with change" earns sympathy and changes nothing. "Send me the agenda the day before and tell me if the time moves" changes the week. Your Insistence on Sameness answers are the raw material for two or three sentences exactly like that, which is what the workbook below is for.

 

Step 2 — Take the screener

 

Twenty questions, about five to seven minutes, free and confidential. You will get a total between 20 and 60 and a band to go with it. One honest steer before you start: this questionnaire has no published cut-off score, so treat the band as a description of where you sit on a dimension, not as a line you have or have not crossed.

 

Take the RBQ-2A

 

Before you start

 

The RBQ-2A is a published, peer-reviewed self-report measure of restricted and repetitive behaviors developed for adults, and it is used dimensionally in research rather than as a test with a pass mark. It has no published cut-off, no population norms and no reported sensitivity or specificity, so the score bands you see here are our own editorial description and not a research threshold. Repetitive behavior is also common outside autism, particularly where anxiety is high, so a raised score does not indicate that you are autistic and a low score does not indicate that you are not. This is a screen, not a diagnosis. Only a qualified clinician can diagnose.

 

Step 3 — Your workbook

 

Your workbook saves to this device only — we cannot see a word of what you write. This module helps you split your score into its two halves, work out what each behavior is actually doing for you, and turn the sameness half into two or three specific things you can ask other people for.

 

Your RBQ-2A results

 

Took the screener? Put your total in below. Worth knowing before you do: this questionnaire has no published cut-off, so the bands are description rather than a line. Entirely optional — skip it if you would rather just read.

 

Score bands: 20–25 = Low Prevalence of Repetitive or Restrictive Behaviors; 26–35 = Mild Prevalence; 36–60 = High Prevalence

 

Fields: RBQ-2A · The Adult Repetitive Behaviours Questionnaire-2, 20 items; Total score (20–60); My total (enter 20–60)

 

Which half is louder

 

Section A. The total flattens the one thing worth knowing — whether your score comes from the body column, the sameness column, or both.

 

Fields: Where most of my score comes from: The motor half — hands, movement, textures, sounds / The sameness half — routines, order, one deep interest / Both, about evenly / I genuinely cannot tell; Three things from the questionnaire I actually do, in my own words

 

What each one is for

 

Section F, item 1. Take three behaviors and ask what happens if you do not do them for an hour. Preference, or regulator?

 

Fields: Behavior one, and what it does for me; Behavior two, and what it does for me; Behavior three, and what it does for me; Tick: At least one of these is a regulator I would not want to lose

 

The uncertainty underneath

 

Section F, item 2. When the routines tighten, something usually got less predictable. This is where to look.

 

Fields: What is unpredictable in my life right now; One unknown I could actually close this month; When did the routines get heavier: They have always been like this / After a specific change or loss / Gradually, over the last year or two / They are lighter now than they were

 

The suppression tax

 

Section F, item 3. Autistic adults describe holding stims in because of how people react, and describe a cost for it. This is a private page.

 

Fields: What I hold in, and where; What it costs me afterwards; One place I could stop holding it in

 

The interest, protected

 

Section F, item 4. Not rationed as a reward. Put in the week on purpose.

 

Fields: The interest; When it goes in the week; What it gives me that nothing else does

 

The ask, in specifics

 

Section F, item 5. One or two sentences somebody could act on tomorrow. Not "I do not cope well with change."

 

Fields: Who I would say it to; The exact sentence

 

Appendix — Research companion

 

Peer-reviewed research

 

1. Barrett SL, Uljarevic M, Baker EK, Richdale AL, Jones CRG, Leekam SR (2015). The Adult Repetitive Behaviours Questionnaire-2 (RBQ-2A): a self-report measure of restricted and repetitive behaviours. Journal of Autism and Developmental Disorders, 45(11), 3680-3692. DOI 10.1007/s10803-015-2514-6. View the paper The development paper. Study 1 gave the questionnaire to 163 UK university students (final N = 161, mean age 21.32); principal components analysis produced two components, Repetitive Motor Behaviours (6 items) and Insistence on Sameness (5 items), with total alpha .83. Study 2 compared 29 autistic adults with 37 non-autistic adults, all IQ above 80, finding higher autistic scores on total and both subscales (Z = -5.43, p < .001, r = -.67; mean totals 1.87 vs 1.41) and total alphas of .91 and .73. The paper scores the measure as a mean from 1 to 3 and proposes no cut-off. Limitation: the structure rests on a young student sample and the clinical comparison on 29 autistic adults, with no norms, no cut-off and no sensitivity or specificity reported anywhere in the paper.

 

9. Barrett SL, Uljarevic M, Jones CRG, Leekam SR (2018). Assessing subtypes of restricted and repetitive behaviour using the Adult Repetitive Behaviour Questionnaire-2 in autistic adults. Molecular Autism, 9, 58. DOI 10.1186/s13229-018-0242-4. View the paper Online study of 275 autistic adults aged 18 to 66 (100 men, 171 women, 5 non-binary), the largest adult RBQ-2A sample at the time. Polychoric principal components analysis produced two components, Insistence on Sameness (alpha .81) and Repetitive Sensory and Motor Behaviours (alpha .70), both correlating moderately with Autism-Spectrum Quotient scores (rs = .25 and .42); the sensory and motor component declined with age (rs = -.21), most clearly after 50, and no significant gender differences were found. Limitation: an online convenience sample, predominantly female, predominantly white, mostly diagnosed in adulthood and highly educated, with no independent confirmation of diagnosis and self-report measures only.

 

8. Brett JD, Peden B, Preece DA, Whitehouse A, Becerra R, Maybery MT (2024). Assessing restricted and repetitive behaviours in online-sampled autistic and non-autistic individuals: factor structure of the Repetitive Behaviours Questionnaire for Adults (RBQ-2A). Journal of Autism and Developmental Disorders, 54(6), 2138-2147. DOI 10.1007/s10803-023-05977-w. View the paper Two studies, 368 Australian university students (mean age 19.44) and 283 self-identified autistic adults recruited online (mean age 32.28), comparing the original two-factor model against a four-factor model and a hierarchical variant using confirmatory factor analysis. The Barrett two-factor model fitted poorly in both samples (RMSEA .075 and .089), while the four-factor model - repetitive motor behaviours, interest in sensations and objects, insistence on sameness, restricted interests - fitted well (RMSEA .036, CFI .989, TLI .982 in study 1) with hierarchical omega of .77 to .80 and configural, metric and scalar invariance across groups. Limitation: the autistic sample was self-identified and recruited online, the groups were unmatched on demographics, and neither sample includes autistic adults with intellectual or language disability.

 

7. Cooper K, Russell A (2025). Insistence on sameness, repetitive negative thinking and mental health in autistic and non-autistic adults. Autism, 29(2), 424-434. DOI 10.1177/13623613241275468. View the paper Comparison of 67 autistic adults with moderate to severe depression recruited clinically, 54 autistic community participants and 66 non-autistic community participants, using the RBQ-2A insistence on sameness items with measures of rumination, obsessing and depression-anxiety. Autistic community participants scored higher on insistence on sameness (M = 2.28 vs 1.42), rumination and obsessing, and mediation across all groups showed the association between insistence on sameness and depression-anxiety ran through repetitive negative thinking, total indirect effect 0.31 (95% CI 0.22 to 0.41). Limitation: convenience sampling, a predominantly white and highly educated sample, and a cross-sectional design that cannot establish the direction of the mediation.

 

10. Edwards H, Wright S, Sargeant C, Cortese S, Wood-Downie H (2024). Research Review: a systematic review and meta-analysis of sex differences in narrow constructs of restricted and repetitive behaviours and interests in autistic children, adolescents, and adults. Journal of Child Psychology and Psychiatry, 65(1), 4-17. DOI 10.1111/jcpp.13855. View the paper Systematic review of 46 studies with 25 entering meta-analysis, examining sex differences separately for each narrow restricted and repetitive construct rather than for a single composite. Autistic males scored modestly higher than autistic females on stereotyped behaviours (SMD 0.21, 95% CI 0.09 to 0.33) and restricted interests (SMD 0.18, 95% CI 0.07 to 0.29), while there was no difference on insistence on sameness (SMD 0.01, 95% CI -0.03 to 0.05, p = .68) or on sensory experiences (SMD -0.09, 95% CI -0.27 to 0.09). Limitation: English-language studies only, 36% of meta-analysed studies rated unsatisfactory for unrepresentative samples or weak diagnostic confirmation, substantial heterogeneity, and planned moderator analyses for age and IQ could not be run.

 

13. Grove R, Hoekstra RA, Wierda M, Begeer S (2018). Special interests and subjective wellbeing in autistic adults. Autism Research, 11(5), 766-775. DOI 10.1002/aur.1931. View the paper Study of autistic adults drawn from the Netherlands Autism Register examining special interests and subjective wellbeing. About two thirds of participants reported having special interests, most holding several rather than one, with computers, autism, music, nature and gardening among the commonest topics; motivation for engaging with an interest was associated with higher subjective wellbeing and with satisfaction in domains including social contact and leisure, while very high engagement intensity was associated with lower wellbeing. Limitation: a cross-sectional register-based self-report sample of adults able to complete online questionnaires, so it cannot show that interests cause wellbeing or that the pattern holds for autistic adults with higher support needs.

 

5. Hwang YI, Arnold S, Srasuebkul P, Trollor J (2020). Understanding anxiety in adults on the autism spectrum: an investigation of its relationship with intolerance of uncertainty, sensory sensitivities and repetitive behaviours. Autism, 24(2), 411-422. DOI 10.1177/1362361319868907. View the paper Self-report survey of 176 autistic adults (mean age 42) and 116 non-autistic adults, using the RBQ-2A alongside measures of intolerance of uncertainty, sensory sensitivity and anxiety. Intolerance of uncertainty and anxiety were both elevated in the autistic group, all four constructs were positively correlated, and intolerance of uncertainty mediated the relationship between sensory sensitivity and anxiety and between anxiety and insistence on sameness, with results stable across age groups. Limitation: cross-sectional self-report in a volunteer sample, so the mediation models describe association rather than sequence in time.

 

12. Jones CRG, Livingston LA, Fretwell C, Uljarevic M, Carrington SJ, Shah P, Leekam SR (2024). Measuring self and informant perspectives of restricted and repetitive behaviours (RRBs): psychometric evaluation of the Repetitive Behaviours Questionnaire-3 (RBQ-3) in adult clinical practice and research settings. Molecular Autism, 15, 24. DOI 10.1186/s13229-024-00603-7. View the paper Psychometric evaluation of the RBQ-3, a 20-item measure that consolidates the RBQ-2 and RBQ-2A into matched self-report and informant-report versions for use across the lifespan, in 110 adults seeking autism assessment and an online sample of 151 autistic and 151 non-autistic adults. Two factors emerged consistently, Repetitive Sensorimotor Behaviours and Insistence on Sameness; total alpha and omega were .90 in the clinical study, self-informant agreement was rs = .71 with informants reporting lower scores than self-raters, and convergent correlations with the restricted and repetitive domain of a blind clinical interview were rs = .45 to .54 against a weaker association with the social communication domain. Limitation: samples were autistic adults without intellectual disability, ethnically homogeneous, with no non-autistic diagnostic outcomes and no test of the lifespan claim in children.

 

11. Keating J, Van Goozen S, Uljarevic M, Hay D, Leekam SR (2023). Restricted and repetitive behaviors and their developmental and demographic correlates in 4-8-year-old children: a transdiagnostic approach. Frontiers in Behavioral Neuroscience, 17, 1085404. DOI 10.3389/fnbeh.2023.1085404. View the paper Study of 260 children aged 4 to 8 (66.9% male) referred by mainstream schools for behavioural, emotional or cognitive difficulties but carrying no clinical diagnosis at the time of testing, using a parent-report repetitive behaviour measure. Repetitive behaviour was higher than in community samples though lower than in autistic samples, insistence on sameness scores exceeded repetitive sensory-motor scores, and anxiety severity and male sex were significantly associated with both subtypes, with lower socioeconomic status and younger age linked to higher insistence on sameness. Limitation: a single parent-report measure with a single informant in a sample referred for a particular purpose, with no assessment of autistic traits for comparison, so it shows that repetitive behaviour is not diagnosis-specific rather than quantifying overlap.

 

6. Moore HL, Brice S, Powell L, Ingham B, Freeston M, Parr JR, Rodgers J (2022). The mediating effects of alexithymia, intolerance of uncertainty, and anxiety on the relationship between sensory processing differences and restricted and repetitive behaviours in autistic adults. Journal of Autism and Developmental Disorders, 52(10), 4384-4396. DOI 10.1007/s10803-021-05312-1. View the paper Survey of 426 autistic adults in the UK, including people formally diagnosed and people awaiting assessment, using the RBQ-2A with the Toronto Alexithymia Scale, the Intolerance of Uncertainty Scale-12 and the Hospital Anxiety and Depression Scale. For repetitive motor behaviours a serial pathway ran from sensory processing through alexithymia and intolerance of uncertainty to anxiety; for insistence on sameness, intolerance of uncertainty acted directly and through alexithymia, with anxiety playing no role. Direct effects from sensory processing to both subscales were also significant. Limitation: cross-sectional self-report with no non-autistic comparison group, so the two pathways are correlational and cannot establish that different mechanisms cause the two behaviour types.

 

Lived experience

 

3. Collis E, Gavin J, Russell A, Brosnan M (2022). Autistic adults' experience of restricted repetitive behaviours. Research in Autism Spectrum Disorders, 90, 101895. DOI 10.1016/j.rasd.2021.101895. View the source Interview study with 12 autistic adults without intellectual disability, analysed thematically, testing three assumptions built into the research literature. Four themes emerged - self-regulation, positive impacts, negative impacts and suppression - and participants' accounts did not map onto the higher-order versus lower-order distinction used in clinical models. Expression of some behaviours in some contexts was largely, though not wholly, under voluntary control, and negative social evaluation led some participants to suppress visible repetitive behaviour at a psychological cost. Limitation: twelve participants, all without intellectual disability, recruited in the UK, so the themes are indicative rather than representative.

 

2. Kapp SK, Steward R, Crane L, Elliott D, Elphick C, Pellicano E, Russell G (2019). 'People should be allowed to do what they like': autistic adults' views and experiences of stimming. Autism, 23(7), 1782-1792. DOI 10.1177/1362361319829628. View the source Qualitative study using interviews and focus groups with 32 autistic adults, analysed thematically, and conducted with autistic involvement in the research team. Participants described stimming as a self-regulatory mechanism used to soothe or communicate intense emotions and thoughts, valued it as a coping strategy, and opposed interventions aimed at eliminating it while accepting the need to manage settings where it is unsafe or unwelcome. This is the key source for reading repetitive motor behaviour as function rather than symptom. Limitation: a small purposive UK sample of adults able to take part in interviews and focus groups, so it does not describe the experience of autistic people with intellectual disability or limited speech.

 

4. Morris IF, Sykes JR, Paulus ER, Dameh A, Razzaque A, Vander Esch L, Gruenig J, Zelazo PD (2025). Beyond self-regulation: autistic experiences and perceptions of stimming. Neurodiversity, 3. DOI 10.1177/27546330241311096. View the source Two participatory online survey studies of autistic adults, 131 participants (mean age 24) and 117 participants (mean age 31), using mixed methods. Around 73% of the first sample agreed that they can usually tell what another autistic or neurodivergent person is feeling when that person stims, and roughly 60 to 63% endorsed stimming as important in friendship and in connection with other autistic people; connection to the autistic community was the strongest predictor of viewing stimming as socially communicative. Limitation: online recruitment over-represented adults with a positive autistic identity and strong community connection, under-represented people with higher support needs, and did not verify diagnoses.

 

Peer-reviewed = checked by independent experts before publication. Clinical model = an established professional framework, not a single study.

 

Up next

 

Module 7 - The Monotropism Questionnaire — Autistic Attention and Deep Focus

 

All modules in Autistic Self-Discovery

 

Repetitive does not mean broken. This course was built by clinicians who are part of the New Path Family. If a lifetime of routines and busy hands has only ever been described to you as a problem, it is worth talking through with someone who reads them as regulation first. Therapy for clients in California and coaching worldwide, all by telehealth, are offered by our sister company New Path Family of Therapy Centers, Inc. A conversation costs nothing and there is no pressure. Saving this for later counts too. Talk with the New Path team