Yes, and there is direct evidence, but the story has to be shorter and more concrete than the one you would write for a same-age peer. A 2024 multiple-baseline study taught classroom social skills to children with an intellectual disability using a story-based intervention alone. The practical constraint is time: in a 2024 survey of 16 school SLPs, pediatric OTs, and parents, 94% spent 30 or more minutes on one story, and adapting it takes longer.
What does the research actually say?
The most direct study is a 2024 multiple-baseline evaluation by Piper, Reeve, Deshais and colleagues in Education and Training in Autism and Developmental Disabilities. They taught four classroom response classes to children diagnosed with an intellectual disability: saying "please," saying "thank you," appropriate hand raising, and giving assistance. Results indicated the story-based intervention may be effective for increasing classroom social skills, and notably it was evaluated as a standalone intervention rather than inside a prompting and reinforcement package. That matters, because most of the older literature bundles stories with other supports and cannot separate the effects.
Zoom out and the picture is consistent but hedged. The AFIRM social narratives brief packet lists social narratives as an evidence-based practice for learners aged 3 to 22, and the NCAEP evidence brief reaches the same conclusion. The 2026 Frontiers in Psychology meta-analysis reported a moderate overall effect (Tau-U 0.743) that was significant for social, safety, and problem behaviors but not for classroom adaptive skills. Expect the story to help with a discrete social response before it helps with broad independence.
What prerequisite skills should you check first?
Cognitive level is a weaker predictor than these three concrete skills. Check them before you write, not after the story fails.
| Prerequisite | How to check it in 3 minutes | If it is absent |
|---|---|---|
| Sustained shared attention | Read any familiar picture book. Does the student stay for 4 pages? | Cut the story to 3 pages, one sentence each |
| Receptive one-step language | "Show me the door." "Touch the cup." | Move to symbols plus a single word per page |
| Picture-to-referent matching | Show a photo of the classroom sink. Can the student point to the real sink? | Use only photos of the real setting, never illustrations |
A student who fails picture-to-referent matching will not benefit from a cartoon character standing in for them. That is the failure mode school teams misread as "social stories do not work for this kid."
A pediatric OT in the community research put the same point plainly: "Some of our students might not yet be able to relate the abstract characters from the stories to their personal experience yet and may benefit from more directed wording and pictures to help them apply the information." That is a format problem, not a diagnosis problem.
How do you adapt the story when comprehension is the barrier?
Five changes, in the order worth making them:
- Cut length before anything else. Three or four pages, one sentence per page. Roughly half a typical K-5 story.
- Replace every illustration with a real photo of the actual room, adult, and object, shot at the student's eye level.
- Strip abstract emotion vocabulary. "Sad" and "frustrated" are hard. "My body feels tight" and a photo of clenched hands are concrete.
- Use the same sentence frame on every page across all of the student's stories, so the format itself becomes predictable.
- Read it in the actual setting, not in your therapy room. Generalization is the weakest link and reading in place shortens the transfer.
Can you add more directive sentences if the student needs more direction?
No. That is the wrong lever, and it is the most common mistake here. Carol Gray methodology holds the ratio at a minimum of two descriptive or perspective sentences per directive sentence, and the reason survives the adaptation: a story made of instructions stops describing the situation and becomes a compliance script. When a student needs more support, shorten the story, simplify the vocabulary, and pair it with modeling or rehearsal. Do not stack "I will" sentences.
What if the student has no reliable narrative comprehension at all?
Then a social story is the wrong tool right now, and that is a legitimate finding to document. Drop to a picture sequence with one word or symbol per step and call it what it is: a visual support. Build narrative comprehension separately using familiar routine sequences the student already lives (arrival, snack, dismissal). Reintroduce the story format once the student can answer a simple what-happens-next question about a known routine. This is not giving up on the intervention. It is sequencing it correctly.
How long should you run it before deciding it did not work?
Give it at least six exposures across two to four weeks, the dosage used in the ASSSIST-2 cluster randomized trial of 249 autistic children across 87 schools. Track two data streams, not one: the target behavior, and a simple comprehension check ("Where does this happen? What do I do?"). If the behavior has not moved but comprehension is solid, keep going and check whether the story matches what actually happens in the room. If comprehension is failing, stop and fix the format. Running the same story a fourth week without checking comprehension is how a term of service time disappears.
Frequently Asked Questions
Is there research on social stories for students with an intellectual disability?
Yes. A 2024 multiple-baseline study by Piper, Reeve, Deshais and colleagues in Education and Training in Autism and Developmental Disabilities taught four classroom social skills to children with an intellectual disability and reported the story-based intervention increased those skills as a standalone intervention, without a bundled prompting or reinforcement package.
What prerequisite skills does a student need before a social story will help?
Three things: sustained attention to a shared book or screen for the length of the story, receptive understanding of simple one-step statements, and the ability to match a picture to a real object or place. If any one is absent, adapt the format before you conclude social stories do not work for the student.
How short should the story be?
One sentence per page and three to four pages total is a reasonable starting point, which is roughly half a typical K-5 story. Add pages back only after the student attends to the whole story twice without redirection. Length is the first thing to cut and the last thing to add back.
Should you use real photos or illustrations for these students?
Real photos of the actual room, adult, and object, taken from the student's eye level. Abstract characters require the student to map a generic figure onto themselves, which is exactly the transfer step that is hardest here. Photos of the real setting remove that step entirely.
Can you add more directive sentences if the student needs more direction?
Adding directives is the wrong lever. Carol Gray methodology keeps at least two descriptive or perspective sentences per directive because a story that is mostly instructions stops being informative. When a student needs more support, cut total length, simplify vocabulary, and pair the story with modeling or rehearsal rather than stacking commands.
What do you do if the student has no reliable narrative comprehension?
Drop to a picture sequence with one word or symbol per step and treat it as a visual support rather than a social story. Build narrative comprehension separately with familiar routine sequences. Reintroduce the full story format once the student can answer a simple what-happens-next question about a known routine.
How long should you run the story before deciding it is not working?
Give it at least six exposures across two to four weeks before judging, which matches the dosage used in the ASSSIST-2 trial. Track comprehension separately from the target behavior. If comprehension checks fail, the problem is the story format, not the intervention.
A practical stack for adapting stories across a mixed caseload
One approach for school SLPs short on time is to keep a 5-tool stack: a Carol Gray sentence-type checklist you audit every draft against, a short 3-page slide template kept separate from your standard 6-page one, a folder of real classroom photos sorted by room rather than by scenario, an AI text drafter (ChatGPT, Claude, MagicSchool, or Emoquest for one-sentence-in story output), and a delivery format your district already uses. The adaptation above is mostly subtraction, so the fastest path is having a shorter template ready rather than trimming a long story under time pressure.