Social Stories in ABA Therapy: How They Are Really Used

Social stories get handed out constantly. They come home in backpacks, they turn up in treatment plans, and there are thousands of free ones online. What is much less clear to most parents is what they are supposed to do, and what they are not.

Short version: a social story is a few sentences describing a situation your child does not understand yet, read before that situation happens. Inside an ABA program, it is preparation, not a teaching procedure. That distinction does most of the work in this article. The research is also thinner than most articles on this topic let on, which I will get to.

The honest summary, up front:

"Social stories" is a trademarked method with 10 specific criteria, and most of what gets called a social story does not meet them.
A well-written one is mostly description, not instruction, and a large share of it should describe things your child already does well.
Inside an ABA program, a story is an antecedent strategy, meaning it sets a child up before a situation rather than teaching a skill on its own.
The evidence is genuinely mixed, including one large randomized trial that found real but narrow benefits.
The most common mistake is using a story where a teaching procedure was actually needed.

A behavior analyst uses a social story as preparation. It is a short, personalized piece of writing that describes an upcoming or confusing situation in plain language, read with a child before the situation happens, so they arrive already knowing what to expect. In an ABA program, it is almost never used alone. It gets paired with direct teaching, prompting, and reinforcement, and the team collects data to see whether it is actually helping.

First, a terminology problem worth knowing about:

This trips up almost everyone, including plenty of professionals.

Social Stories: A trademarked method developed by Carol Gray, who began building the guidelines in 1991. A piece of writing only qualifies if it meets 10 defining criteria covering how it is researched, worded, formatted, and delivered. The criteria have been revised repeatedly over three decades.
Social Narratives: A broader, generic category, and it is the term you will find in the research. The AFIRM team at the University of North Carolina reports that, based on the 2020 systematic review by the National Clearinghouse on Autism Evidence and Practice, social narratives meet evidence-based practice criteria on the strength of 20 single case design studies and one group design study, with demonstrated effects from preschool through high school.

Why this matters to you: the laminated page that came home from school, the free PDF off Pinterest, the story your RBT typed up on a Tuesday. Most are social narratives. Very few are social stories in Gray's sense. That is not automatically a problem, but it does mean research on the trademarked method does not automatically transfer to whatever is in your child's backpack.

What a Good One Actually Looks Like

Gray's criteria are detailed, but a few principles do most of the work, and they are the opposite of what most people's instincts produce.

It is mostly description, not instruction.

Gray's guidance is to use roughly two to five descriptive, perspective, or affirmative sentences for every one directive sentence. Most homemade stories invert this and become a list of rules, which is exactly what a social story is not supposed to be.

Gray has said that for something to count as a social story, half of the stories written on a person's behalf should describe things that person currently does well. That principle came partly from listening to autistic adults who grew up with these stories and found that a lot of what they were given had been correction dressed up as narrative.

It answers the six questions.

Where, when, who, what, how, and why. If a child still would not know what is going to happen after reading it, it needs more description.

Here are the sentence types you will see in a well-constructed story. Take a short one about a fire drill:

On Thursday, my class has a fire drill. A fire drill is practice. It helps everyone learn what to do if there is ever a real fire. A loud bell will ring. Many children do not like the sound. That is okay. I can put my hands over my ears. My teacher will walk with us out to the field. I am good at following my teacher.

Sentence typeWhat it doesExample from the story aboveDescriptiveStates facts about the situation, without opinion"A fire drill is practice for what to do if there is a fire."PerspectiveDescribes what other people think, feel, or know"Many children do not like the loud bell."DirectiveSuggests a response the child can try. Used sparingly"I can put my hands over my ears."AffirmativeReinforces a shared value or reassures"That is okay."Affirmation of competenceNames something the child already does well"I am good at following my teacher."

Nine sentences. Six of them just describe what will happen, one tells the child what he can do, one reassures, and one names an existing strength. It would take about fifteen minutes to write.

Where a Story Sits Inside an ABA Program

This is the part that gets skipped, and it is the part that determines whether the thing works.

In behavior analytic terms, a social story is an antecedent strategy. It happens before the situation, and its job is to change how the child arrives at it. It is not a teaching procedure. It does not, by itself, build a skill through practice and reinforcement the way discrete trial teaching or behavioral skills training does.

So in a well-run program, it shows up as one component of something larger. The BCBA identifies the situation, writes the story with input from the family about what the child actually finds hard, and then pairs it with the parts that do the teaching: rehearsal, prompting, reinforcement for the target response, and a plan for the real event. Data gets taken on the target behavior, not on whether the child sat through the reading. If you want to see how antecedent supports fit into the overall shape of a session, our approach page walks through it.

Situations where our clinicians reach for one most often: a first dentist or haircut appointment, a fire drill, a substitute teacher, a new sibling, a family flight, or a change to a routine the child relies on. What these share is that the difficulty is largely about not knowing what is coming.

What the Research Actually Says

Here is where I want to be careful, because most articles on this topic overstate the case.

The Association for Science in Autism Treatment, which reviews autism treatments on the evidence rather than the marketing, describes the evidence for social stories as mixed. Numerous studies have found them effective, and the National Standards Project classified story-based intervention as an established procedure for increasing social skills and decreasing problem behavior. But other reviews have concluded the supporting evidence is low or questionable, and there is a specific methodological problem: many studies used a story alongside another technique such as prompting, which makes it impossible to say whether the story, the prompting, or the combination produced the change.

The largest test to date was ASSSIST-2, a pragmatic cluster randomized trial across 87 UK primary schools involving 249 autistic children, published in 2025. Six months in, the primary outcome measure showed a small reduction that was not statistically significant. But children in the story group met their individual socio-emotional goals more often than children receiving usual care, and that difference was significant. No differences appeared in anxiety, depression, general health, or parental stress. The authors' conclusion is worth reading slowly: social stories are a low-cost, low-burden intervention, with benefits on individual goals but no clinically evident impact on broader social responsiveness.

That is more useful than it first appears. It suggests these work best when narrow and specific, which is how a good behavior analyst would use one anyway. ASAT's recommendation is that professionals present social stories as having limited scientific support, avoid using them in isolation, and consider procedures with stronger evidence alongside them. They also note something I think about often: a story may partly function to cue the adult about what the expected behavior is.

What We See in Our Own Sessions

Picture a girl we will call Noa, seven, in Bergen County, and the dentist.

Two previous appointments had ended before anyone got near her teeth. Her mother had described the visits to her beforehand each time, in the car, in about four sentences.

Her BCBA wrote a story of about ten sentences. It named the specific practice, described the chair going back and the light and the sound of the suction, said that some children do not like the sound, and that is okay, gave her one thing she could do, which was raising her hand to ask for a break, and noted that Noa is good at telling people what she needs. They read it once a day for two weeks. Then they practiced the hand raise at home with reinforcement, went to the office once just to sit in the waiting room, and only then booked the actual appointment.

The cleaning happened. It was not smooth, and it took two attempts at the chair, but it happened.

I use this example because of what it shows about attribution. Nobody on that team believes the story did it. The story, the rehearsal, the reinforcement, and the desensitization visit all happened together, which is precisely the confound researchers keep pointing at. What the story contributed was that Noa walked in knowing the sequence and knowing she had a way to pause it. That is a real contribution, and a smaller one than "social stories work for dentist visits."

Noa is a composite rather than one specific child, but the pattern is one we see repeatedly.

How to Tell a Good Story From a Bad One

If something comes home from school or arrives from a therapist, here is what I would check.

Count the directive sentences. If more than a quarter of the story is telling your child what to do, it has become a rule list.
Look for the word "don't." Well-written stories describe what will happen and what a child can do. Heavy negation is a sign it was written to correct rather than to inform.
Check whether your child appears in it. A generic story about "children at the dentist" is a worksheet. A story that names the actual office, the actual chair, and something your child is genuinely good at is a support.
Ask what it is paired with. If the answer is nothing, ask what the teaching plan is. A story on its own is preparation without practice.
Watch your child read it. If it makes them anxious rather than settled, that is information, and the story needs rewriting or shelving.

Writing One at Home

You can absolutely do this. Keep it short, and write it in the first person from your child's point of view. Describe the situation concretely, including the sensory parts that are actually the problem. Include one or two things your child can do, no more. Include at least one sentence naming something they already handle well, because that is both the humane choice and part of the method. Read it well before the event, not in the car on the way.

Then show it to your child's BCBA. A five-minute read can catch the two most common problems, which are too many directives and a target that needed a teaching procedure rather than a story. Parent training is built into the programs we run for exactly this kind of thing.

Putting One to Work for Your Family

Social stories are a modest, low-cost, low-risk tool that can genuinely take the edge off a situation your child does not understand yet. They are also not a teaching procedure; they are frequently written in ways that undercut their purpose, and the evidence supports narrow, specific use rather than broad claims. Understanding that distinction is most of what separates a story that helps from a laminated page nobody reads twice.

Centerbrite provides quality ABA therapy services to families in New Jersey, across Bergen, Essex, and Morris counties. Our BCBAs write supports like these into a wider plan, pair them with the teaching that actually builds the skill, and take data so we can tell you honestly whether something is working rather than assuming it is. We will also tell you when a story is not the right tool, which happens more often than you might expect. If you have a situation coming up that your child is dreading, or you would like a second opinion on the supports your child is currently getting, contact us today.

Frequently Asked Questions

Both, in a sense. They originated outside behavior analysis, with Carol Gray, but behavior analysts use them regularly as an antecedent support inside a broader program, and social narratives appear on the NCAEP list of evidence-based practices for autism. Our FAQ covers more on which techniques a program typically draws on.

2. Can I just download one instead of writing my own?

You can, and it is a reasonable starting point, but the personalization is most of the value. A downloaded story about "going to the doctor" does not know your child's clinic, your child's specific worry, or what your child is already good at. Adapt rather than adopt.

3. How long should one be?

Short. Ten sentences is plenty for a young child. The instinct to be thorough usually produces something too long to hold attention and too directive to be reassuring.

4. What if the story does not seem to help?

That is common, and it is not a failure. It usually means the situation needed teaching rather than information. Bring it to your BCBA and ask what procedure would target the skill directly, since a story is preparation rather than practice.

Sources:

https://asatonline.org/for-parents/learn-more-about-specific-treatments/social-stories/
https://afirm.fpg.unc.edu/resource/social-narratives-brief-packet/
https://carolgraysocialstories.com/
https://pubmed.ncbi.nlm.nih.gov/39688069/

Social Stories in ABA Therapy: How They Are Really Used