Social narratives

• Approach family: Social narratives
• Primary purpose: Provide accurate, accessible and individualised information about a specific social situation, event, perspective or response option
• Primary discipline: Education and psychology, often with speech and language therapy contribution
• Contributing disciplines: Occupational therapy, health, social care, family support and whole-setting practice
• Typical settings: Education, home, health, care, transport, community and work-transition settings
• Typical users: Young people who want or need clearer, more durable information about a particular social or practical situation
• Typical tier: Universal information design, targeted narrative support or individualised narrative intervention

Encyclopaedia topic connections

These working tags describe the approach’s focus. Conditional connections depend on the particular barrier, goal and application. Professional delivery roles remain recorded separately.

Direct focus

• 03 Communication & connection — The narrative makes a particular social situation, cues, perspectives or response options more understandable.

Conditional connections

• 01 Emotional life & regulation — Relevant when the chosen narrative addresses uncertainty or fear around a specific event.
• 02 Thinking, attention & action — Relevant when a narrative explicitly prepares for a transition or changing sequence.
• 06 Agency, safety & participation — Relevant when content addresses access, safe options, saying no or joining a chosen activity.

Related Encyclopaedia reading

These entries offer related reading for the situations described. A link expresses the specific connection below; it does not establish individual suitability or an effect of the approach.

• E043 Difficulty managing uncertainty — Read alongside this approach when a truthful account of an uncertain situation may make information easier to use. Preserve uncertainty where it exists; a narrative must not promise that events or other people’s intentions are certain.
• E098 Difficulty joining shared activities or groups — Read alongside this approach when a young person wants information about joining a shared activity or leaving it. A narrative can describe available options and adult responsibilities without making participation compulsory.
• E005 Difficulty with transitions — Read alongside this approach when a transition becomes difficult because the sequence, destination or available options are unclear. Check the environment and plan as well as the wording of the narrative.

What belongs to this approach

A social narrative is a deliberately authored, usually short and reviewable account that helps a young person understand a particular situation. It may describe what happens, why it happens, relevant cues, what different people may know or feel, what the young person can do, what others will do, and how uncertainty or change will be handled. It may use text, pictures, symbols, audio, video or a combination.

“Social Stories™” is a named subset associated with defined authorship criteria. The broader label social narratives includes other individualised narrative formats, but those should not automatically inherit evidence or fidelity (delivery of the approach as intended) claims from the trademarked approach. The implemented record must identify the format, authorship process, content rules, delivery and outcome.

A021 is not:

• a generic picture schedule or A006 Visual and anticipatory support, although a narrative may use visual information;
• A020 Explicit instruction and modelling, although a narrative may sit within direct teaching or rehearsal;
• ordinary fiction, story time, a comic strip, a behaviour contract or a list of rules;
• a disguised warning, threat or demand for compliance;
• a replacement for environmental adaptation, communication access, safeguarding, health investigation or adult responsibility;
• evidence that the young person lacks empathy or must learn to imitate neurotypical social behaviour.

The intervention boundary is the specific authored narrative plus its collaborative use. Simply giving a child a commercially available story, reading a script repeatedly, or adding clip art to adult instructions does not establish that a social-narrative intervention was delivered.

Purpose and evidence boundary

Social situations can be difficult to interpret because information is implicit, fast, inconsistent or known only to other people. A durable narrative can make relevant information available before, during or after a situation. Its purpose is to reduce the need to work everything out in the moment and make questions, options and repair easier to discuss.

The narrative must be truthful and proportionate. It should explain the environment as well as the young person’s options. If the fire alarm may be loud, the account should include what staff will do and how the person can access protection or leave—not merely say that they should remain calm. If a peer interaction involves uncertainty, the story should avoid declaring what another person definitely thinks.

The strongest large UK trial did not find a clear effect on its broad primary social-responsiveness outcome at six months, but did find improvement in individual goal-based outcomes. That distinction fits the likely use of the approach: a social narrative may help with a well-defined situation without changing a global trait or every social context.

A021 therefore asks: What precise uncertainty or goal is being addressed, what information does the young person say would help, is the narrative accurate and accessible, how will it be used without coercion, and does it improve the intended outcome or experience?

Core components

Define the specific situation and purpose

Identify one meaningful situation, question or transition. Describe what currently happens, what remains unclear, whose goal it is and why a narrative might help. “Improve social skills” is too broad. A usable goal might concern joining a chosen lunchtime activity, knowing what will happen at a blood test, recognising when a lesson plan changes, or understanding how to ask for space during group work.

Gather perspectives and facts

Observe the setting and ask the young person, family and relevant people what information is missing. Separate observable facts from interpretations. Identify sensory, communication, emotional, cultural and practical conditions. Where another person’s perspective is included, use qualified language rather than mind-reading.

Co-author where possible

Invite the young person to shape the question, wording, images, voice, format, privacy and use. They may correct adult assumptions, identify irrelevant details or reject the intervention. Co-authorship does not require the person to write every sentence; it requires meaningful influence and an accessible route to consent, dissent and revision.

Write accurately and respectfully

Use concrete, age-respectful language and enough context to make the situation intelligible. Describe more than direct. Include variation and uncertainty where real. Avoid moral labels, shame, exaggerated consequences and promises adults cannot guarantee. State what adults and environments will do, not only what the young person should do.

Offer realistic options

Describe more than one acceptable response when possible, including asking for help, taking time, using augmentative and alternative communication (AAC), leaving safely, declining, trying later or repairing a misunderstanding. The narrative should expand predictability and agency, not narrow behaviour to a single performance.

Make the format accessible

Match language, length, layout, symbol or picture type, sensory presentation, literacy and communication mode to the person. Use photographs only with permission and attention to privacy. An adult may read or model the story, but comprehension should not depend on inaccessible speech.

Introduce it at a workable time

Preview the narrative when the person has sufficient regulation and time, commonly before the relevant event rather than at peak distress. Explain its purpose and offer a choice about reading or revisiting it. Repetition should serve recall and confidence, not become a demand or punishment.

Practise, use and update carefully

If the person wants, pair the narrative with questions, rehearsal, role-play, a visit, video, visual schedule or environmental change. Use it in the real context and notice whether the information matches events. Correct inaccuracies promptly; revise as circumstances change; retire a story that is no longer helpful.

Monitor the intended outcome and experience

Measure the specific goal, use and acceptability. Include the young person’s understanding, control, distress or effort; adult fidelity; transfer to the real situation; and any unintended pressure or stigma. Do not equate quiet compliance with understanding or wellbeing.

What this may look like

An individualised social narrative may:

• explain the sequence, people, sensations and choices involved in an unfamiliar appointment;
• describe what usually happens when a lesson is taught by a substitute and what staff will do if the plan changes;
• clarify how to enter, pause or leave a chosen group activity;
• explain that different people may interpret the same message differently;
• describe how a friend might be unavailable without claiming to know the reason;
• show a range of ways to say “no”, ask for clarification or request more time;
• prepare for a fire drill while also specifying hearing protection, location and adult support;
• explain a new school rule, including its purpose, exceptions and route for questions;
• support repair after a recurring misunderstanding without deciding blame in advance;
• use first person, second person or neutral wording according to the person’s preference;
• combine concise text with photographs, symbols, drawings, audio or video;
• include blank or editable sections for details that vary;
• state “sometimes”, “may” and “usually” where outcomes are uncertain;
• identify what adults, peers and services will do to make the situation accessible;
• be read independently, shared with a trusted partner or used as a reference in the moment;
• be revised or withdrawn when the person says it is wrong, embarrassing or unhelpful.

A narrative is not made therapeutic by its tone. “When I am good, I keep my hands still and look at my teacher” remains a compliance demand even if it is illustrated. The content and power relationship matter.

Tiered use

Tier 1 — Universal accessible explanation

Make routines, changes, expectations and available support clear to everyone through accessible information. A general narrative about a new building or shared event may sit here if it describes rather than disciplines. Universal materials should not disclose an individual’s needs or replace direct communication.

Tier 2 — Targeted narrative support

Develop a narrative for a recurring situation shared by a small group, or adapt a base account for individuals. Check that the purpose, language and options fit each user. A premade template can support production but should not substitute for fact-finding and individualisation.

Tier 3 — Individualised social narrative

Co-author a personalised narrative for a specific, significant or high-demand situation; coordinate it with communication, environmental and professional plans; and monitor a defined outcome. Where trauma, severe anxiety, safeguarding, healthcare or restrictive practice is involved, a narrative is only one possible support within a wider plan.

Practice example

This is a constructed illustration of the approach, not a reported case or an observed outcome.

Imagine a pupil who wants to try a science club but is unsure how partners are chosen. The pupil and teacher prepare a short, truthful account of the sequence, the options for working with someone or independently, and how the pupil can ask for clarification or a pause. The narrative says what is known and what may change; it does not claim to know another pupil’s intentions. The teacher also makes the actual pairing process clearer.

Review would check whether the account stayed accurate, whether the pupil wanted to use it and whether the available options were usable. The pupil’s view, participation and distress would be recorded separately. Because the pairing process also changed, any improvement could not be attributed to the narrative alone.

1. Why the evidence boundary matters

The literature uses “Social Stories”, “social narratives”, “social scripts” and related terms inconsistently. Studies vary in sentence rules, point of view, illustrations, rehearsal, comprehension checks, reinforcement and additional supports. Some target communication or participation; others target reductions in behaviours selected by adults.

A broad positive conclusion can therefore disguise incompatible interventions and goals. Phoenix treats the authored narrative, fidelity criteria, delivery context and target outcome as essential evidence fields. Brand-specific findings remain brand-specific unless the active components and implementation are genuinely shared.

A defensible narrative answers a real information need. It describes the setting, important cues, possible perspectives, likely variation, available responses and the support other people will provide. It makes unspoken expectations clearer without assuming that only the young person must change.

Description should outweigh direction. Instructions may be included, especially for safety or a practical sequence, but they should be accurate, justified and paired with accessible choices and adult responsibilities. A story that tells a person to tolerate preventable distress is not made acceptable by friendly wording.

2. What systematic reviews show

Reviews have often found positive results across selected behaviours, but much of the evidence comes from small single-case studies. A 2018 review applying quality standards found only seven of 22 studies supplied strong or moderate causal evidence, and different synthesis methods produced different conclusions about whether the practice met an evidence-based threshold.

A 2024 scoping review identified 56 articles and found that most, but not all, reported positive outcomes. It also highlighted limitations in designs and generalisation. A review of reviews found continuing variability and uncertainty, reinforcing the need to distinguish precise outcomes and implementation rather than treating the approach as settled across all uses.

3. What the school trial adds

The UK Autism Spectrum Social Stories in Schools Trial 2 (ASSSIST2) was a pragmatic cluster randomised trial. It included 249 primary-aged autistic children across 87 schools. Trained school staff developed and used Social Stories to established criteria, with a minimum intended dose across four weeks. At six months, the difference in the teacher-rated broad social-responsiveness primary outcome was small and not statistically significant.

Individual goal-based outcomes did favour the Social Stories group. There were no clear broad gains across several secondary measures, and delivery occurred across the disruption of the COVID-19 period. The most responsible interpretation is neither “the stories worked” nor “the stories failed”: a structured narrative may support an individualised goal while not changing global social responsiveness.

4. Individualisation is an active component

The intervention begins before writing. Observing the context and hearing the young person can reveal that the adult’s assumed target is wrong. A narrative about turn-taking will not solve uncertainty about room noise, partner assignment or ambiguous instructions. The fact-finding process may lead to an environmental change instead of, or alongside, a narrative.

Templates can reduce workload, but generic language can also omit the decisive detail or impose normative assumptions. Individualisation means more than changing a name and photograph. It changes the problem definition, information, options, format and criteria for success.

5. Plausible mechanisms and their limits

A narrative may make transient spoken information durable, reduce uncertainty, preview sensory and social demands, support memory, provide language for repair and allow rehearsal away from real-time pressure. These mechanisms are plausible and useful for case formulation.

They do not prove that the narrative caused change. Adults may also alter their behaviour while writing it; the environment may become more predictable; the person may receive extra attention; or repeated exposure may account for some outcome. Combined intervention is often appropriate, but the record should not award the narrative sole credit.

6. Implementation determines what was tested

Record who gathered information, who authored and approved the narrative, which criteria were used, its length and mode, when and how often it was offered, whether use was voluntary, what additional supports occurred and how inaccuracies were corrected. A nominal story that was not read, could not be understood or described a different reality was not a fair test.

Fidelity must not become rigidity. If the young person reports that wording is wrong or repetition is aversive, revision is better fidelity to purpose than enforcing a schedule. Record adaptations so later reviewers know what actually happened.

7. Minimum monitoring

Monitoring should include:

• implementation: availability, actual use, timing, comprehension and fidelity to the intended format;
• specific outcome: the observable or reportable goal named before the intervention;
• experience: usefulness, accuracy, autonomy, distress, shame, effort and willingness to continue;
• generalisation: whether learning or access appears in the real situation, with relevant people and over time;
• maintenance: what remains after the story is faded or no longer needed;
• environmental response: whether adults delivered the supports promised in the narrative;
• unintended effects: literal misinterpretation, rigidity, dependence, exposure of private information or pressure to mask.

Where the outcome is adult-rated, add the young person’s report or an accessible proxy for experience wherever possible. A reduction in visible behaviour can coexist with increased internal distress.

8. Cautions, safeguarding and equity

Do not use a narrative to secure obedience, prevent disclosure or make a young person tolerate bullying, pain, sensory harm, unwanted touch or unsafe practice. Eye contact, stillness, smiling and suppression of harmless self-regulation should not be framed as moral obligations. A narrative must never replace action on abuse, discrimination, unmet communication need or an inaccessible environment.

Protect privacy when photographs, diagnoses, bodily information, family events or peer details are included. Avoid cultural stereotypes and one “correct” interpretation of social behaviour. Provide home-language and AAC access. Check that images and language are age-respectful. Narratives generated by artificial intelligence (AI) require human fact-checking, safeguarding review and meaningful young-person input; fluent text is not evidence of accuracy or consent.

High-stakes situations require real safeguards. A story about getting lost is not a safety plan. A story about a medical procedure does not replace consent, pain relief or reasonable adjustment. A story about bullying does not shift responsibility from adults to the person experiencing it.

9. What we can say

• Carefully specified social narratives have improved some focused, individual outcomes.
• The largest pragmatic school trial found a goal-based benefit without a clear broad social-responsiveness effect.
• Individualisation, accurate description, accessible format and implementation review are central to a fair trial.
• A narrative makes information available beyond a fast spoken explanation; whether it supports preparation or repair needs review in the particular situation.
• Involving the young person provides a way to check accuracy and acceptability; recipient voice remains under-measured in the research.
• Social narratives may be combined with environmental, communication and teaching supports when the combined plan is recorded honestly.

10. What we cannot say

• We cannot claim that social narratives broadly improve autism, social responsiveness, friendship, behaviour or wellbeing.
• We cannot assume findings for Social Stories™ apply to every illustrated script or generic downloadable resource.
• We cannot infer understanding or reduced distress from quiet behaviour or compliance.
• We cannot guarantee what other people will think, feel or do.
• We cannot attribute change solely to the story when rehearsal, adult attention or environmental adaptation changed at the same time.
• We cannot use a narrative to transfer adult or system responsibility to the young person.
• We cannot assume repeated reading is harmless or that refusal indicates a need for more repetition.

11. Review, exit and escalation

At review, ask whether the narrative answered the person’s actual question, matched what happened, was used by choice, improved the defined outcome and reduced rather than increased uncertainty or pressure. Ask what the young person would change. Review whether adults fulfilled their part and whether a different environmental or communication change is now indicated.

Stop or revise when the narrative is inaccurate, disliked, shaming, overly literal, no longer needed or ineffective after a fair agreed trial. Fade when the information has become familiar, while keeping it available if the person wants a reference.

Escalate to relevant clinical, education or safeguarding support when the situation involves trauma, severe anxiety, pain, loss of skills, bullying, abuse, dangerous behaviour, restrictive practice, significant communication barriers or a high-stakes health or legal decision. A narrative is not a reason to delay assessment or protection.

12. Phoenix judgement

Social narratives may be useful for a defined information need. Their usefulness and burden depend on the individual account and how it is used. Their evidence is strongest when the question, narrative criteria and individual outcome are clear. The literature does not justify broad claims that stories change general social ability, and the large pragmatic trial cautions against such expansion.

A time-limited trial can be considered when the young person wants or accepts clearer information, the story is factual and individualised, adults also change what is within their control, and success is defined around the specific goal and experience. The narrative should illuminate a situation, not rewrite the young person into compliance.

Shared evidence profile fields

For social narratives, identify the authored material, intended information and any named authorship criteria. Evidence for Social Stories™ remains attached to that defined intervention rather than every narrative format.

The following fields make a specific evidence or implementation record comparable with other Atlas profiles. They do not supply missing study data or replace the source-specific judgements in this entry.

• Identity and version: Name the component or programme, its edition or protocol where relevant, and the purpose being evaluated.
• Source role: Separate direct comparative evidence, descriptive implementation findings, guidance, theoretical support and firsthand experience.
• Population and setting: Retain the studied age, communication/access characteristics and setting. State when an intended application lies outside them.
• Intended and actual delivery: Record the planned components, frequency and duration (dose) separately from what was received, by whom, and with what support or interruptions.
• Comparison and other support: Name the actual comparison and concurrent changes. If there is no comparison group or condition (comparator) or it is unreported, say so rather than supply an assumed usual-care condition.
• Outcome and reporter: Record the outcome, time point and reporter. Keep the young person’s account separate from someone else’s judgement on their behalf (proxy report), and access or participation separate from task performance.
• Delivery as intended (fidelity) and adaptation: Record whether the core components were delivered and describe adaptations; distinguish an adapted application from the protocol tested.
• Burden and unwanted effects: Record time, effort, practical access, cost or workload where known, alongside refusal, fatigue, distress, restriction and other unwanted consequences. Unreported burden is not evidence of no burden.

• Wright and colleagues — ASSSIST2 randomised trial and economic evaluation
• Wright and colleagues — Social Stories in mainstream schools
• NIHR — ASSSIST2 clinical effectiveness results
• Como and colleagues — Social narratives for autistic people: scoping review
• Camilleri, Maras and Brosnan — Review of Social Story reviews
• Qi and colleagues — Social Story intervention systematic review
• Karkhaneh and colleagues — Social Stories systematic review
• Carol Gray Social Stories — definition and criteria

The research retains the findings and limits of the September 2026 source review. This is a targeted review, not an exhaustive systematic review, a formal evidence grade or completed specialist consultation. Research findings do not determine which support is right for an individual young person.