Social Skills and Relationships: Fostering Meaningful Connections for Autistic Children

A friend of mine in Edinburgh — a woman with two children, the older of whom is autistic — once said something I have been thinking about ever since. He is not failing at being a child, she said, he is succeeding at being a particular kind of child, and the bit that is failing is the room around him.

I want to begin with that observation because almost all of the parenting writing I have read about social stories for autism — and about autism social skills more broadly — assumes the work to be done is in the autistic child. It is not, on the consensus of the most recent research, where most of the work to be done actually is. The double empathy problem — proposed by the autistic researcher Damian Milton in 2012 and now supported by ten years of subsequent evidence — finds that the social-communication breakdowns we have for decades treated as autistic deficits are bidirectional. Both groups, autistic and non-autistic, struggle to read each other; autistic people communicate effectively with other autistic people; the breakdown sits in the middle of mixed-neurotype interactions, and the responsibility to bridge it belongs to both sides.

That reframe matters because it changes the entire shape of this article. The old framing produced social skills training, in which the autistic child is the project. The newer framing produces social translation and connection support, in which the work is collaborative and the autistic child is, importantly, allowed to remain the kind of child they are. The strategies below — social stories, friendship through special interests, self-advocacy phrases, bullying response templates, sibling check-ins — all sit inside that newer frame.

Let me say what I mean by social stories and where they actually fit, because the term is doing a great deal of work in this domain and a good deal of that work is sloppy.

Social Stories — What They Are and What They Don't Do

A Social Story™ — the trademark belongs to Carol Gray, who formalised the form in the early 1990s — is a short, personalised narrative written in the child's voice that describes a specific situation: who will be there, what is likely to happen, what kinds of responses tend to work, and what the child can do if it gets hard. They are usually five to ten sentences long. They are read aloud immediately before the situation. They are revised every few months as the situation changes.

The honest version of the evidence: social stories are useful, narrowly. The 2025 ASSSIST-2 pragmatic randomised controlled trial, the largest UK trial of Social Stories™ in mainstream primary schools, found that children receiving the manualised intervention met individual situational goals (a new bus route, a fire drill, a hairdresser appointment) significantly more often than children receiving usual care. It did not find clinically meaningful changes in general social responsiveness, anxiety, depression, parent stress, or general health.

The strongest version of the case for social stories is therefore the narrow one. They are a useful instrument for specific situational goals — this dental visit, this fire drill, this substitute teacher next Thursday. They are not a general social-skills cure, and articles that claim them as one are overselling. The strongest version of the case against them is that they have, historically, been deployed to enforce neurotypical norms ("look at the speaker", "say hello on entry"); used that way, they slide into the masking territory I will address later in this article. Used for narrow, child-relevant, child-acceptable goals, they are one of the cheaper and better-evidenced tools in the autism social-support kit.

A workable structure for a parent writing one at home, drawn from Carol Gray's framework:

• Descriptive sentences — what is going to happen, factually. On Thursday I will go to the dentist. The dentist's name is Sarah. There will be a chair that goes up and down.
• Perspective sentences — what other people will be doing or feeling. Sarah will smile. She will ask me to open my mouth.
• Directive sentences (used sparingly) — what the child can do. I can wear my headphones if it gets loud. I can raise my hand if I need a break.
• Affirmative sentences — calm reassurance. It is okay to be a little nervous. It will not last very long.

A good social story is short, written in the child's voice, and rereads naturally aloud the morning of the event. The mistake parents most often make, in my experience, is making them too long.

Ten Social Stories You Can Adapt

The single most useful thing I think this article can offer is a starter set of social stories you can adapt to your own child's specific situation. Each one below is a working draft — change the names, simplify the language for younger children, add child-specific detail. The point is the structure, not the exact words.

1. Going to the dentist. On Thursday I will go to see Sarah, the dentist. The dentist's office is bright and has a few different sounds. Sarah will smile and ask me to open my mouth so she can count my teeth. There will be a chair that goes up and down. If it gets loud, I can wear my headphones. If I need to stop, I can raise my hand. It will not take very long, and afterwards we will go to the park.

2. The fire drill at school. Sometimes my school practises what to do if there is a fire. There is no real fire. The alarm will be very loud. The teacher will tell us to walk in a line outside. I can put on my headphones if my bag is open. We will walk to the playground. We will wait quietly. After a few minutes the bell will stop and we will walk back in. The drill is a kind of practice. It does not mean anything bad is happening.

3. A substitute teacher. Sometimes my regular teacher is not at school for the day. When that happens, a different teacher comes in for the day. They are called a substitute. They will know our names. They might do things a little differently. The schedule will be the same. If something feels wrong, I can use my "I need a quiet break" card.

4. The school bus. On weekday mornings I take the school bus. I get on at the same place each day. The driver's name is Miguel. The bus is sometimes loud because other children are talking. I can sit near the front, where it is a little quieter. I can listen to my music. I do not have to talk to other children if I do not want to. When the bus stops at school, I get off.

5. Going to the hairdresser. On Saturday I am going to get my hair cut. The person cutting my hair is called Anna. She is very good and she works gently. The room will smell a little different and there will be background music. The clippers might tickle. If it gets too much, I can ask Anna to stop and take a break. The haircut takes about ten minutes. Afterwards my head will feel different for a while, and that is normal.

6. A sleepover at a friend's house. On Friday night I am sleeping at Lila's house. Her house is different to mine. The bathroom is in a different place. Her parents have different rules. I will bring my own pillow and my own pyjamas. I can text mum or dad if I want to come home. If something feels too much, that is okay — I do not have to stay all night.

7. The supermarket on a busy day. Saturday at the supermarket can be loud and bright. There will be many people moving in different directions. The lights are very bright. We will get the things on our list — six items today. I can hold the trolley or hold mum's hand. If it gets too loud, I can squeeze the small ball in my pocket. We will not stop in the middle of the aisle. When we have the things on our list, we will go.

8. A friend conflict. Sometimes I and a friend have a disagreement. This does not mean we are not friends anymore. It means we are two people who do not agree right now. I can tell my friend, calmly, what bothered me. I can listen to what they say. We do not have to fix it immediately. We can take time and come back to it later. Most disagreements get fixed.

9. The lunchroom. At lunchtime I go to the lunchroom. It is loud because everyone is talking and eating. I can sit at the same table I usually sit at. I do not have to talk to anyone if I do not want to. I can eat my lunch with my headphones on. If it gets too loud, I can go to the quiet room. I have twenty-five minutes for lunch.

10. Transitions between classrooms. In school I move between classrooms during the day. The bell will ring when it is time to move. I will pack my things into my bag. I will walk in the hallway with the other children. The hallway is sometimes crowded. If I need a moment, I can stand against the wall and let people walk past. I will get to the next classroom on time.

Each of these can be cut to half its length for younger children, expanded for teenagers, or supplemented with photographs of the actual venue when possible. The version your child accepts is the right version. The first draft is rarely it.

Loading image...

Friendship: Through Interest Tunnels, Not Around Them

Most parenting writing about autism friendships frames the work as helping the autistic child make friends the way other children make friends. The recent neuro-affirmative literature is unusually direct that this is the wrong frame.

The 2025 Frontiers paper by Brosnan and Camilleri draws on monotropism theory — the proposal that autistic attention is deeper but narrower — and recommends what it calls joining the autistic individual in their attention tunnel rather than trying to get them out. Translated into friendship strategy, that means building connection through the child's existing intense interests, not in spite of them. A working set of practical moves:

• Parallel play around a shared interest. Two children, one shared activity (Lego, drawing, building a jigsaw, watching the same documentary), little expectation of constant verbal exchange. The interaction is in the shared focus, not in the small talk.
• Interest-swap days. A friend brings the autistic child into a topic they care about deeply for thirty minutes; the autistic child reciprocates the next session. Reciprocity made structural.
• Special-interest meetups. Lego clubs, train-spotting groups, fan communities, animal-welfare junior volunteer programmes, music-listening groups. Friendship anchored to a topic the child already loves is dramatically easier to sustain.
• One stable 1:1 friendship rather than a wide circle. The Mazurek line of research — replicated several times since 2014 — finds that quality of friendship predicts reduced loneliness, depression, and anxiety in autistic adults far more than quantity does.
• Structured activities over open-ended play. Board games, cooperative video games, baking, building. The structure of the activity does the work that small talk does for neurotypical friendships.

Find Autistic Peer Connection

A specific subset of the friendship strategy worth its own paragraph. The Crompton and colleagues research series finds that autistic-to-autistic information transfer is statistically as effective as neurotypical-to-neurotypical; the breakdown in mixed-neurotype communication is what drives the apparent "social difficulty" we have historically located in the autistic child.

The practical implication, which most parenting writing skips: time spent with other autistic children is not less social time. It is, on the data, higher-bandwidth communication. Concrete pathways worth pursuing:

• Local neurodivergent peer groups (often run through autism advocacy organisations or specialist schools).
• Autism community events and family camps in your region.
• Parent-network playdates with families whose autistic children share an interest with yours.
• Special-interest meetups that often quietly skew neurodivergent (board game cafés, certain museum programmes, model-building clubs).
• For older children and teens: online communities anchored to specific interests, with appropriate parental scaffolding around online safety. For autistic teens, a well-moderated online friendship is, frequently, the more stable and reciprocal form.

The standard parental instinct is to push the autistic child toward neurotypical peers in the hope they will "learn from" them. The newer evidence suggests that parallel investment in autistic peer connection produces the more sustainable mental-health outcome.

A Self-Advocacy Phrasebook for Kids

The phrase self-advocacy gets used to describe a great many things, and most of them are too abstract for an eight-year-old. What an autistic child actually benefits from is a small, ready-to-deploy set of phrases they can practise in low-stakes moments and reach for in difficult ones.

A working age-banded phrasebook, drawn from neurodiversity-affirming practice and informally pressure-tested with families I have spoken to:

Ages 5–7.

• "I need quiet."
• "I need a break."
• "Too loud."
• "I'm done." (signalling end of social tolerance, not rudeness)
• "Help, please."

Ages 8–12.

• "I'm overstimulated, I need a break."
• "Please rephrase that — I didn't follow."
• "I'd rather write it than say it."
• "Can we do this another time?"
• "I'm not being rude, I'm focused on something."

Teen.

• "Please email me instead of calling — I process written better."
• "I need a few seconds to think before I answer."
• "I'm not flirting — I'm sharing my interest with you."
• "I can do this for thirty minutes; I'd like to leave after that."
• "If I stop responding, I'm not ignoring you, I'm regulating."

The phrasebook is at its most useful when the child has been allowed to practise the phrases at home, repeatedly, in low-stakes moments, until they sit close to the surface. Asking a child to invent self-advocacy language in the moment of distress is asking the impossible. Drawing from a known set is asking the manageable. The whole point of the Autistic Self Advocacy Network's long-running phrase Nothing About Us, Without Us is that these are the child's own words, used on their behalf, by them. Self-advocacy is not what we do for an autistic child. It is what we equip the child to do for themselves.

Bullying: A Response Toolkit

Autistic children and teenagers face bullying at roughly two to three times the rate of their non-autistic peers, per the Maïano et al. meta-analysis cited across the autism research literature. That number alone justifies treating bullying response not as an emotional support task but as a paperwork-and-process task. The work in the moment is documentation, written reporting, and IEP-amendment language. The emotional work follows; without the procedural work, the emotional work has nowhere to land.

A working three-part toolkit:

Part one — your initial reporting email (send to teacher AND principal; verbal complaints lose the paper trail):

Dear Teacher Name and Principal Name,

I am writing to report a bullying incident affecting my child, Child Name, on date and time, at location. The incident involved brief factual description: who was involved, what was said or done, who witnessed it. My child reported the incident to me at time, in what setting.

I am requesting (1) an investigation into the incident under the school's anti-bullying policy, (2) a written summary of the school's response by date, typically within 5 school days, and (3) a meeting to discuss adjustments to my child's IEP/504 plan to address ongoing safety.

Please reply by email to confirm receipt and timeline for next steps.

Thank you for your prompt attention.

Your Name

Part two — the teacher follow-up script (for the meeting):

Three things I want to understand:

1. What is the school's process from here, and what timeline?
2. What changes will be made to my child's day to ensure their safety in the meantime?
3. How will the IEP be amended to reflect ongoing anti-bullying supports?

One thing I am asking for: a designated safe adult my child can go to at any moment of the school day without prior negotiation, and a written check-in protocol so we both know if and when something else has happened.

Part three — IEP/504 language to add:

Anti-bullying support: Student will have a designated safe adult (named in this plan) who is available to the student at any moment of the school day for de-escalation and reporting. Any subsequent bullying incident will be documented in writing within one school day and shared with the parent. Student will be permitted, without consequence, to leave any social setting in which they feel unsafe. The IEP team will reconvene within 30 days of any reported incident to assess the adequacy of supports.

Bullying is not a moment to improvise. It is a moment for prepared language. The school is required to take a written report seriously in a way it is not required to take a verbal one.

Siblings: Three Check-In Scripts

A short note on siblings, because they are the most often-overlooked person in autism family writing. The non-autistic sibling of an autistic child is doing real, often unrecognised, daily work — accommodating, explaining to friends, sometimes acting as informal interpreter, sometimes feeling guilty about wanting attention themselves. They benefit, in my experience and in the family-research literature, from one specific thing: a regular, brief, dedicated check-in that is theirs alone.

Three age-banded scripts you can adapt as a weekly fifteen-minute conversation, just the two of you, no agenda:

• Ages 5–8 — What was the best part of your week? What was the hardest part? Is there anything you wish more people understood about our family?
• Ages 9–12 — What's been hardest about having a sibling who is autistic this week? Is there one thing you wish more people understood? When do you need a break that's just yours?
• Ages 13+ — How are you, honestly? What have you been carrying that I haven't asked about? What kind of support would feel useful from me right now — and what would feel like too much?

Acknowledge complicated feelings without correcting them. Love and frustration coexist; both are honest. Connect older siblings with peer support — Sibshops, online sibling communities specifically for siblings of autistic children — so they meet other people who understand the specific shape of the day. The single most stabilising thing I have heard non-autistic siblings name, in clinical and research settings, is being asked, regularly, by someone who actually wants to know.

Romantic Relationships, for Autistic Teens

Most parents I have spoken to who are raising autistic teens describe the romantic-relationships territory as the place where they feel least equipped. The conventional dating-and-relationship education often assumes a level of fluency in coded social cues that autistic teens, on the whole, do not share — and the well-meaning advice ("just be yourself") is unhelpful when the being yourself part is the part that other people are misreading.

A few principles drawn from the small but growing literature on autistic teen relationship education and from the AANE (Association for Autism and Neurodiversity) approach:

• Explicit-language preference. Many autistic teens prefer direct over coded social cues. Teach phrases like I need to know what you mean exactly and I am sharing this with you because I trust you, not as a flirt as legitimate, not as social failures.
• Consent in plain terms. Consent is best taught with concrete language — yes / not yet / no; if I am not sure, I check; if they are not sure, I stop — rather than with the vague atmospheric framing common in mainstream sex education.
• Sensory considerations are real. Dating involves a great deal of physical proximity, eye contact, perfume, ambient noise. An autistic teen's preferences here are not quirks to be educated out of; they are real information about what makes a relationship sustainable.
• The double empathy point applies here too. A successful autistic relationship — with another autistic person or with a non-autistic person — is, importantly, a mutual translation project, not a one-sided fitting-in.

What We Don't Recommend

I want to spend a paragraph being explicit about the strategies the current research does not support, because they appear in many older articles on this topic and produce measurable harm. The honest version of this advice belongs in writing, not implied.

The 2023-2025 literature on masking (or camouflaging — the conscious or unconscious suppression of autistic traits to fit in) is unusually direct. Masking is associated with elevated anxiety, depression, autistic burnout, and suicidality, particularly in autistic girls, late-diagnosed adults, and teens. The implication is uncomfortable for some traditional social-skills approaches.

Specifically, the strategies we should treat as not the goal:

• Forced eye contact. Looking at or near the speaker's face is fine; eye contact specifically is, for many autistic people, sensorially overwhelming and cognitively expensive. Permit alternatives.
• Stim suppression. Stimming serves regulatory and (per 2025 SAGE research) social-communication functions. Suppress only what is actively self-injurious; for everything else, allow.
• Scripted small-talk drilling. Producing mock conversations to drill turn-taking and small-talk patterns may produce surface-level fluency at the cost of real exhaustion. The double empathy literature suggests this is the wrong direction — both sides need the work, not only the autistic child.
• "Indistinguishability from peers" as a goal metric. This was the defining metric of older intensive intervention programmes. The current evidence treats it as masking-by-another-name, and the long-term outcomes are not good.

These are not radical claims. They are, by 2026, the consensus of the field. Articles that still implicitly chase indistinguishability are working from a 2010-era playbook.

PEERS and Other Manualised Programmes

A short note for parents of teens specifically. PEERS® (the UCLA Program for the Education and Enrichment of Relational Skills) is the most evidence-based manualised social-skills programme for autistic adolescents and young adults — a fourteen-session, ninety-minutes-per-week curriculum, parent-included, delivered in twelve or more languages and a hundred and fifty plus countries, with durable improvement at three- and six-month follow-up.

A 2024 set of PEERS adaptations explicitly addresses earlier criticisms by reframing the programme as deconstructing social interactions rather than teaching the autistic teen to perform neurotypical social skills, and by minimising programme language that could encourage masking. This makes the current version of PEERS a defensible referral for motivated teens who have asked for friendship support. It is not the right fit for a teen who is being pushed into it to act less autistic — that is the masking-harm pattern current research is moving away from.

If you are interested, the PEERS Clinic at UCLA maintains a locator for licensed providers internationally.

A Closing Note

The friend in Edinburgh I mentioned at the start of this article said one other thing, in the same conversation, that I want to end on. I have stopped trying to make him a child the world is ready for, she said. I am trying to make him a child who knows the world is not always going to be ready, and who has the language and the people he needs anyway.

That, in a sentence, is what most of this article is trying to do. The social stories are vocabulary. The friendship strategies are people. The self-advocacy phrasebook is language. The bullying toolkit is the procedural infrastructure that protects the child while the rest of the work is being done. The double empathy framework, the monotropism reframe, and the masking critique are the moral architecture inside which all of these specific moves make sense.

The sharper version of the question I started with is, I think, this: what kind of social world is one in which an autistic child does not have to spend the day translating in order to belong? The answer to that question is the long, slow, patient, ordinary work of friendships and outings and accommodations and brave, repeated, awkward conversations — the work the rest of this article tries to make practical. I do not think it is a finished question. I do not think it is supposed to be.