Autism-pedia / Therapies and treatments / Developmental and naturalistic approaches (NDBIs)brianspringer.com
Therapies and treatments

Developmental and naturalistic approaches (NDBIs)

The early childhood programs grouped under this name, including the Early Start Denver Model, JASPER and PACT, combine behavioral teaching with developmental science. What the trials show, whether the programs differ from ABA, and what autistic adults say about their goals.

These are programs that help young autistic children by teaching through play. Three well-known ones are the Early Start Denver Model, JASPER and PACT. This page shows what the studies found, and how these programs are like, and unlike, ABA.

Last reviewed 12 October 2026 · one person's reading of the research · 32 research and 6 community sources

What the approaches are Established

Naturalistic developmental behavioral interventions (NDBIs) are early childhood programs that combine behavioral teaching techniques with developmental science. The term was codified in 2015, when Laura Schreibman and colleagues grouped a set of programs that had grown up under different names, to give families, providers and agencies a way to compare them.1

The programs are delivered in a child's ordinary settings, in play and daily routines rather than at a table in a clinic. The therapist or parent follows the child's interests, shares control of the session with the child, and uses behavioral techniques such as prompting and reinforcement to teach developmentally appropriate skills. The skills targeted are usually social communication, joint attention, language and play.1

Three programs are named most often. The Early Start Denver Model (ESDM) is a comprehensive program for young children, starting around a child's first birthday, that targets communication, cognition, motor skills and adaptive behavior. JASPER (Joint Attention, Symbolic Play, Engagement and Regulation) is a focused program that targets joint attention and play. PACT (the Preschool Autism Communication Trial) is parent-mediated: a therapist coaches parents to build back-and-forth communication with their child.1,2,3

Each program under this label has its own manual, its own training and its own developer, so a service that offers NDBI may be running any of them.1,4

Naturalistic developmental behavioral interventions (NDBIs) are programs for young autistic children. They mix two ideas. One is behavioral teaching. The other is what we know about how children grow and learn. The name was settled in 2015. Laura Schreibman and other researchers put several programs into one group. The programs had grown up under different names. The group name helped families and services compare them.1

These programs happen in a child's normal places, during play and daily routines. They are not done at a table in a clinic. The adult follows what the child likes. The child and the adult share control. The adult uses behavioral methods, such as prompts and rewards, to teach skills a child is ready for. The skills are usually talking and communicating, sharing attention with others, language and play.1

Three programs are named most often. The Early Start Denver Model (ESDM) is a full program for young children. It starts around a child's first birthday. It works on communication, thinking, movement and everyday skills. JASPER works on sharing attention and play. PACT is for parents. A therapist coaches parents to build back-and-forth talk with their child.1,2,3

Each program has its own guide, its own training and its own maker. A service that says it offers NDBI may be using any of them.1,4

Suggest a correction
Established

The description comes from the publication that named the family and from the papers that define each program.

Word
Naturalistic

Delivered in a child's ordinary settings and routines, following the child's lead, rather than at a table in a clinic.

Established

This comes from the paper that named these programs and from the papers that describe each one.

Word
Naturalistic

Done in a child's normal places and routines. The adult follows the child's lead. It is not done at a table in a clinic.

The three named programs Emerging

ESDM. The first randomized trial, of 48 children aged 18 to 30 months, compared ESDM with a referral to the providers available in the community. Over two years the ESDM group gained 17.6 points on a standard score of intellectual ability against 7.0 in the comparison group, and kept pace with non-autistic children on adaptive behavior while the comparison group fell further behind. The trial was run by the program's developers.5 A three-site trial by the same group, of 118 children aged 14 to 24 months, found a language advantage for ESDM at two of the three sites but none at the third; its planned analysis pooling all three sites found a significant language advantage, and it found no difference in developmental quotient, autism severity or adaptive behavior. Its authors call it a partial replication.6

The largest independent test reached a different result. In a 2025 trial in France and Belgium, 180 children aged 19 to 36 months were assigned to ESDM added to treatment as usual or to treatment as usual alone, with the main outcome assessed blind to group. The ESDM group did not improve significantly more on a measure of overall development, with a mean difference of 3.82 points and a confidence interval that included zero. The dose was lower than in the earlier trials, about 12 hours a week, and the two groups were unequal. The authors conclude that ESDM added to treatment as usual "cannot be universally recommended".7 Pooling 12 studies of 640 children, an ESDM meta-analysis found an overall effect of 0.36, mostly from gains in cognition and language, and no significant effect on autism symptoms, adaptive behavior, social communication or repetitive behavior. Its senior author is an ESDM developer, so it is not an independent synthesis.8

Smaller independent studies have reported gains. A community preschool study in Israel, of 51 children, found greater progress in thinking, language and everyday communication than a comparison program, and a 2026 non-randomized trial in Japan, of 47 toddlers, found a small group difference in overall development from a low-intensity program, with no difference in autism traits. Both were non-randomized, and one author of the Israeli study works on these programs.9,10

JASPER. The earliest randomized study in this line of research tested joint attention and symbolic play interventions in separate arms.11 A systematic review of 19 studies found that most reported greater improvement than comparison on at least one of joint attention, joint engagement, play and language, that no predictors or mediators were consistent across studies, that none met all of a standard set of quality indicators, and that most of the measures were taken close to the therapy rather than in a child's daily life.2 In one developer-run study, joint engagement mediated the effect on later joint attention initiations, and post-intervention initiations predicted later language gains.12 The one long follow-up in this line found that, five years on, 80% of the children used spoken language, that early play skills predicted later language, and that being assigned to the joint attention or the play group predicted later language among the children who were speaking at age eight. It followed 40 of the 58 children, and still stops in childhood.13

PACT. The first PACT trial, of 152 children aged 2 to 4 years 11 months in the United Kingdom, did not find a reduction in autism symptom severity on its main outcome, though parent-child communication improved, including parents' responses to their child and the child's own initiations. The authors concluded that PACT could not be recommended for reducing autism symptoms on the strength of this trial.3 A six-year follow-up of 121 of the children found that, taking the trial and the follow-up together, symptom severity on a calibrated severity score was lower in the PACT group (an effect of 0.55), though the difference at the follow-up visit itself was not significant, and there was no difference in language. The follow-up used a different measure from the trial's original primary outcome. The authors call it the first randomized trial of early intervention to show a long-term symptom reduction.14

ESDM. The first fair test had 48 children aged 18 to 30 months. Some got ESDM. The others were sent to the providers already in their area. Over two years, the ESDM children gained 17.6 points on a test of thinking skills. The other children gained 7.0 points. The ESDM children kept up with non-autistic children on everyday skills. The other children fell further behind. The people who made ESDM ran this study.5 A later study by the same team had 118 children aged 14 to 24 months. ESDM helped with language at two of three places. When all three places were counted together, ESDM still helped with language. It did not help with thinking, autism traits or everyday skills. The authors call it a partial repeat.6

The biggest study by an outside team found something different. In 2025, in France and Belgium, 180 children aged 19 to 36 months took part. Some got ESDM plus their usual care. The others got only their usual care. Someone who did not know the groups checked how the children changed. The ESDM children did not do much better on a test of overall development. They got fewer hours than in the earlier studies. The authors say ESDM plus usual care "cannot be universally recommended".7 One review added up 12 studies of 640 children. It found a small overall effect of 0.36, mostly in thinking and language. It found no real effect on autism traits, everyday skills, communication or repetitive behavior. One of its authors helped make ESDM, so it is not an outside review.8

Smaller studies by outside teams found some gains. One in Israel, with 51 children, found more progress in thinking, language and everyday communication. One in Japan, with 47 toddlers, found a small difference in overall development. Neither was a fair test.9,10

JASPER. The first fair test in this line looked at shared attention and pretend play.11 A review of 19 studies looked at JASPER. Most found more progress than a comparison group. The progress was in at least one of sharing attention, joining in, play or language. No factor predicted who improved. No study met all of a checklist of quality rules. Most of the tests were done close to the therapy. They were not done in a child's own day.2 In one study run by the makers, joining in led to more shared attention later. That predicted more language later.12 One long follow-up checked the children five years later. Eight out of ten were using spoken language. Early play skills predicted later language. For children who were talking at age eight, being in the shared attention or play group also predicted later language. It followed 40 of the 58 children.13

PACT. The first PACT study had 152 children aged 2 to almost 5 in the United Kingdom. It did not reduce autism traits on its main measure. Parents and children did communicate better. That included how parents answered their child. It also included how often the child started to communicate. The authors said PACT could not be recommended for reducing autism traits on its own.3 Six years later, 121 of the children were checked again. Across the study and the years after, their autism traits were lower on one measure. But the difference at that visit by itself was not strong enough to be sure. There was no difference in language. The authors call it the first fair test of early support to show a long-term change.14

Suggest a correction
Emerging

Each program's evidence comes mostly from trials run by its own developers. The programs are not compared with each other or with ABA in randomized trials, and the one large independent ESDM trial did not show a significant advantage on its primary outcome.

Word
Randomized controlled trial

A study that assigns people to treatments at random, so the groups are alike in the ways the researchers can measure and do not choose their own treatment.

Who ran the trial

A developer-run trial is one run by the people who created the program, which is common in this field. It can still be well run. The point is that the developer has an interest in the result, and an independent repeat is what tells careful work from the rest.5,6,3

Emerging

Most of these studies were run by the people who made the program. The programs have not been tested against each other, or against ABA, in a fair test. One big ESDM study run by an outside team found no clear benefit.

Word
Randomized controlled trial

A study where people are put into groups by chance. The groups are alike in the ways the researchers can measure. No one chooses their own group.

Who ran the study

A developer-run study is run by the people who made the program. This is common. It can still be a good study. But the makers want the program to work. So an outside repeat is what tells careful work from the rest.5,6,3

What the studies show Emerging

When the trials are pooled, the effects of the developmental and naturalistic approaches are small to moderate, depending on the outcome and the studies counted. In an update of the largest meta-analysis of early childhood autism intervention, limiting the results to randomized trials left developmental and NDBI approaches with small effects on social communication, adaptive behavior, language and play, with effect sizes between 0.16 and 0.35, and an effect of 0.38 on autism diagnostic characteristics. Behavioral interventions kept a moderate effect on social-emotional and challenging-behavior outcomes (0.58). When the authors also removed outcomes rated by the people who delivered the therapy, and studies at high risk of bias from unblinded assessment, only one significant effect remained. They report that adverse events across the studies were "poorly monitored, but possibly common".15

An NDBI-specific analysis from the same project, covering 27 studies, found effects on social communication, language, play and cognition, with confidence limited by how the outcomes were measured, and larger effects for skills a program teaches directly than for ones a child would have to carry into new settings.16 An independent meta-analysis of 27 studies found the largest effects on social engagement (0.65) and cognition (0.48), smaller effects on expressive language (0.32), play (0.23) and autism symptoms (0.38), and signs of publication bias in the language results.17 An overview of five reviews found positive effects on communication and language, cognition and adaptive behavior, and mixed effects on autism symptoms and repetitive behavior; its authors work in the behavior-analytic field.18

Two further analyses group the naturalistic approaches together with early intensive behavioral intervention as ABA-based. One mixed-methods review and meta-analysis of 16 randomized trials found improvements in adaptive behavior, language and joint attention at low to moderate certainty, and found that programs with more hours a week were associated with larger language gains in a subgroup comparison.19 A meta-analysis of 15 trials of early parent-mediated training reported a large overall effect (0.80), but with very high heterogeneity between studies and signs of publication bias, so the estimate is unstable.20

When many studies are added together, the effects are small to moderate. This depends on the outcome and on which studies are counted. In the largest review of early childhood autism support, the authors kept only fair tests. The developmental and naturalistic programs then had small effects. The effects were on communication, everyday skills, language and play. They were between 0.16 and 0.35, with a small effect of 0.38 on autism traits. Behavioral programs had a moderate effect on difficult behavior (0.58). The authors also removed tests rated by the people who gave the therapy. After that, only one effect was left. They write that bad events in the studies were "poorly monitored, but possibly common".15

Another analysis of 27 NDBI studies found effects on communication, language, play and thinking. The authors were not confident, because of how the studies measured results. The effects were bigger for skills a program teaches directly. They were smaller for skills a child must use in new places.16 A separate review of 27 studies found the biggest effects on joining in (0.65) and thinking (0.48). Smaller effects showed up for language (0.32), play (0.23) and autism traits (0.38). The authors found a sign that studies with good language results were more likely to be published.17 A review of five reviews found good effects on communication and language, thinking and everyday skills. It found mixed effects on autism traits and repetitive behavior. Its authors work in the behavior-analytic field.18

Two other analyses group the naturalistic programs with early intensive behavioral intervention. They call them all ABA-based. One review and analysis looked at 16 fair tests. It found improvements in everyday skills, language and shared attention. The certainty of the evidence was low to moderate. In one comparison, programs with more hours a week had bigger language gains.19 A review of 15 trials of parent training found a large overall effect (0.80). But the studies were very different from each other. There were also signs of publication bias, so the number may not be steady.20

Suggest a correction
Emerging

Pooled effects run from small to moderate, and are larger in the looser analyses. Across the field most trials were not randomized, and the measures often sit close to the therapy rather than in a child's daily life.

Emerging

They found small to moderate effects. Most of the studies were not fair tests. And the tests were often done close to the therapy, not in a child's own day.

What predicts who improves Emerging

Two child factors stand out. In a mega-analysis that pooled the individual records of 582 autistic children from 11 datasets in five countries, the age at which intervention started and the child's developmental level before it started predicted how much they changed. Children who started younger tended to improve more, and children who started with a higher developmental quotient grew faster on some measures. Sex and the total amount of intervention a child received were largely unrelated to change, but only about a fifth of the children, 121 of 582, were girls, so the sample is too small to show a difference by sex. Only one of the datasets came from a randomized trial, and the data were contributed voluntarily, so the authors caution that the results may not generalize.21

The developmental-level finding is not settled. An independent meta-regression of 202 samples found that a child's language and cognitive scores at entry did not significantly predict how much any kind of intervention helped, by outcome area; only cognitive scores predicted the effects of technology-based interventions. Its authors describe the result as exploratory.22

On how much therapy a child gets, the evidence no longer supports the advice to provide as many hours as possible. A meta-analysis of 144 studies found no significant positive link between any measure of how much intervention a child received, whether daily intensity, duration or cumulative hours, and how much the child improved within a given type of intervention. This shows no demonstrated added benefit from more hours, not that more hours do no good.23 The 2025 analysis that grouped the naturalistic and behavioral approaches together did find larger language gains from higher-intensity programs, so the question is open.19

The intervention a child receives may matter less than the child's starting profile. In a pooled sample of 707 preschoolers who received interventions for spoken language, the odds of learning to talk did not differ by which intervention they received. The children who stayed non-speaking had lower motor-imitation scores at the start, the children who did not go on to combine words had lower cognitive, social and adaptive scores and had received less intervention, and the age at which intervention began mattered differently depending on how much spoken language a child started with.24

Two things about the child stand out. One big analysis pooled the records of 582 autistic children from 11 sets of data in five countries. Children who started support younger tended to improve more. Children who started at a higher level of development grew faster in some ways. Only about a fifth of the children were girls. That is too few to show a difference between boys and girls. How many hours of support a child got did not change the result much either. Only one set of data came from a fair test. Families also chose to take part. So the authors warn the results may not apply to everyone.21

This finding about developmental level is not settled. An outside analysis looked at 202 groups. Children's language and thinking scores at the start did not predict how much any program helped. Only thinking scores predicted the effects of computer-based programs. The authors call this an early result.22

On how many hours a child should get, the old advice was to give as many hours as possible. The evidence does not support that any more. An analysis of 144 studies found no real link. More hours did not clearly add benefit. That is not the same as saying more hours do no good.23 But the 2025 analysis that grouped the programs together did find bigger language gains from more hours. So the question is open.19

Which program a child gets may matter less than how the child starts. One pooled group had 707 preschoolers. The chance of learning to talk did not depend on which program they got. The children who stayed non-speaking had lower imitation scores at the start. The children who did not go on to put words together had lower thinking, social and everyday scores. They had also got less support. And the age at which support began mattered differently depending on how much the child could already say.24

Suggest a correction
Emerging

These findings come from a few analyses that mostly reuse the same trials, and they disagree with each other. All the children are young, so the results say nothing about adults.

Emerging

These findings come from a few analyses. They mostly reuse the same studies, and they do not agree with each other. All the children were young, so the results say nothing about adults.

Different from ABA, or a rebranding? Emerging

What is not disputed. NDBIs and ABA are not separate traditions. The naturalistic programs use the same operant techniques as ABA, such as prompting, shaping and reinforcement, and the researchers who named the family write that NDBIs "represent the merging of applied behavioral and developmental sciences" and that "evidence-based NDBIs are based upon well-established principles of applied behavior analysis. Thus, they represent ABA treatment."1 The differences are in delivery and in goal. The naturalistic programs run in everyday settings and follow the child's lead, while the structured ABA of discrete trial training runs at a table and sets the pace itself.1,25

What the studies show. Almost no study puts a named naturalistic program and a named ABA or early intensive behavioral intervention program in a randomized trial against each other; none was found as of October 2026. A trial comparing ESDM with community early behavioral intervention is planned, run by researchers including ESDM developers, and has not yet reported.26 The evidence that exists is indirect. ABA clinicians interviewed about adopting NDBI strategies described having to unlearn discrete trial training, and named their own training and confidence as barriers.25 A review of 18 caregiver-mediated programs found that naturalistic and developmental programs share many coached strategies but differ in others, and often use the same words for different practices.4

Who holds which view. The developers of the naturalistic programs treat them as a distinct group with shared features. The Autistic Self Advocacy Network, in a white paper dated December 2021, answers the question the other way. Its section on the point is headed "'New ABA' is Still ABA"; it writes that "'new ABA' is no more acceptable or ethical than old ABA, even if it is less brutal", argues that "many practices termed 'new ABA' are not, in fact, ABA at all", and holds that an intervention which still uses ABA techniques at its core "is still ABA and still unethical, no matter what other methods it uses".27 Two behavior analysts, Jessica and Abraham Graber, writing in 2025, argue that ABA's recent reform has produced "distinct brands within the field" and warn that competing labels can split reform rather than advance it.28 Giacomo Vivanti, a researcher who works on these programs, wrote in 2024 to reconcile what he calls the "conflicting narratives" of early intervention and neurodiversity.29 In a 2026 survey of 33 autistic adults and 37 parents, the two groups agreed that support should be naturalistic, child-led and person-centered, while parents rated ABA-based intervention more positively and autistic adults were more concerned about harm.30

What would settle it. The question needs a trial that does not yet exist: one that assigns children at random to a named naturalistic program or a named behavioral program, measures outcomes with assessors who do not know which group a child was in, follows the children into adulthood, and reports which manual each group used. Until then, the two share their techniques and their lineage, and whether the differences in delivery matter for children's lives is untested.26

What is not disputed. NDBIs and ABA are not separate traditions. The naturalistic programs use the same behavioral methods as ABA, such as prompts and rewards. The researchers who named the family write that NDBIs "represent the merging of applied behavioral and developmental sciences". They also write that these programs "represent ABA treatment".1 The difference is in how they are run and what they aim at. The naturalistic programs happen in everyday places and follow the child. Older-style ABA with discrete trial training happens at a table and sets the pace itself.1,25

What the studies show. Almost no study puts a named naturalistic program against a named ABA program in a fair test. None was found as of October 2026. One such study is planned. It is run by people who include ESDM's makers, but the results are not out yet.26 Some other evidence is indirect. ABA staff were asked about trying NDBI methods. They said they had to unlearn discrete trial training. They said their own training and confidence were barriers.25 A review of 18 programs found that naturalistic and developmental programs share many ways of coaching but differ in others. They often use the same words for different things.4

Who holds which view. The makers of the naturalistic programs treat them as their own group with shared features. The Autistic Self Advocacy Network, in a paper dated December 2021, answers the other way. Its section on this is headed "'New ABA' is Still ABA". It writes that "'new ABA' is no more acceptable or ethical than old ABA, even if it is less brutal". It argues that "many practices termed 'new ABA' are not, in fact, ABA at all". It also says this. A program that still uses ABA methods at its core "is still ABA and still unethical".27 Two behavior analysts, Jessica and Abraham Graber, wrote in 2025 that ABA reform has made "distinct brands within the field". They warn that too many labels can split the reform.28 Giacomo Vivanti studies these programs. In 2024, he tried to bring together what he calls the "conflicting narratives" of early support and neurodiversity.29 In a 2026 survey, 33 autistic adults and 37 parents agreed on naturalistic, child-led, person-centered support. But parents rated ABA-based programs more positively, and autistic adults were more worried about harm.30

What would settle it. The question needs a study that does not exist yet. It would put children into a named naturalistic program or a named behavioral program by chance. It would check the results with people who do not know which group each child was in. It would follow the children into adulthood. And it would say which guide each group used. Until then, the two share their methods and their history. No one knows if the differences matter for children's lives.26

Suggest a correction
Emerging

The approaches overlap in technique and in research lineage, and almost no trial compares them head to head. The positions below are dated and attributed; the pedia describes them and does not take one.

See also
Emerging

The two kinds of program share their methods and their history. Almost no study compares them directly. The views below belong to named people and groups, with dates. The pedia describes them and does not pick a side.

See also

What is missing, and what adults say Emerging

Harms are the weakest part of the record. The largest meta-analysis found that adverse events were "poorly monitored, but possibly common" in the trials it covered, which leaves no way to weigh the benefits against the possible harms.15

The trials also stop at childhood. No study found as of October 2026 reports adult outcomes from a randomized trial of an NDBI, so how the gains, or the experience of these programs, play out over a life is not known from the research.15,21

Autistic adults, asked what early intervention should do, put well-being, adult support and reducing harmful behavior first, and do not endorse goals that reduce autistic traits. In the four quantitative studies inside a review of 19 studies of their views, the adults rated those priorities highest, and the review also named a cross-cutting theme of "Damage done" in the negative experiences they reported. The adults surveyed are self-selected, so the ranking is theirs and not a representative sample.31 In the 2026 survey, autistic adults and parents agreed on naturalistic, child-led and person-centered support, though the adults were more concerned than the parents about harm.30

Who benefits is uneven. In a secondary analysis of the ESDM trial, 73% of the children in the ESDM group and 57% in the comparison group made reliable improvement on a video-coded measure of social communication and repetitive behavior, the Brief Observation of Social Communication Change. That is one measure, not global improvement, and the analysis comes from the developers' own trial. The comparison group's gains depended on how a child started out: children with lower cognitive scores and higher autism severity scores at the start were less likely to meet the threshold for reliable change, while the ESDM group's gains did not depend on those factors.32 A researcher in the field notes that children with profound intellectual disability and minimal spoken language have been studied less over time, even though they have the greatest needs, and that access to early intervention is unequal.29

Harm is the weakest part of the record. The largest review found that bad events were "poorly monitored, but possibly common" in the studies it covered. So there is no way to weigh the good against the possible harm.15

The studies also stop at childhood. No study found as of October 2026 has followed children from a fair test of an NDBI into adulthood. So no one knows how the gains, or the experience, play out over a life.15,21

Autistic adults were asked what early support should do. In the four number-based studies inside a review of 19, the adults rated well-being, adult help and reducing harm highest. They did not rate reducing autism traits highly. The review also named a theme of "Damage done" in the bad experiences adults reported. These adults chose to take part, so they may not speak for everyone.31 In the 2026 survey, autistic adults and parents agreed on naturalistic, child-led and person-centered support. But the adults were more worried than the parents about harm.30

Who improves is uneven. In a second look at the ESDM study, some children improved reliably. That was 73% of the ESDM children and 57% of the comparison children. This was on one video-coded measure of social communication. It was not all-round improvement, and the analysis came from the makers' own study. For the comparison children, how much they improved depended on how they started. Children with lower thinking scores and more autism traits were less likely to reach the mark. The ESDM children's gains did not depend on those things.32 One researcher notes a gap. Children with a big intellectual disability and little or no speech have been studied less over time. They need the most help. He also notes that access to early support is not equal.29

Suggest a correction
Emerging

These are gaps and one dated absence claim. The best supported part is that harms were not tracked well; the rest rests on single reviews and one secondary analysis.

Emerging

These are gaps, and one claim about what has not been studied, dated. The best supported part is that harm was not tracked well. The rest comes from single reviews and one extra analysis.

In the community Community

Autistic writers rarely separate the naturalistic programs from ABA, and much of the writing treats "early intervention" as one thing. Kristy Forbes writes that early interventions are "deeply rooted in behaviorism", and that a child's own communication is "of no interest" to the early intervener, whose plan is shaped by neuro-normative standards.c1 Amy Pearson and Kieran Rose write that behavioral interventions and frameworks "frame 'autistic behaviour' ... as impaired, lacking functionality, and/or challenging".c2

Orion Kelly, an autistic writer, describes both his son's early intervention and his own later diagnosis, and writes that the therapy seemed to teach his son "how not to act or look Autistic".c3 What writers ask for instead is close to what the naturalistic programs claim to do. Jules Edwards favors "inclusive relationship-based interventions" that build relationships and meet a child where they are rather than demanding obedience.c4 The editors of Sincerely, Your Autistic Child write that the sources of strength for a good life were "almost never a particular kind of therapy or clinical intervention", and that acceptance, recognition and inclusion matter more.c5

Autistic writers rarely split these programs off from ABA. Much of the writing treats "early intervention" as one thing. Kristy Forbes writes that early interventions are "deeply rooted in behaviorism". She writes that a child's own communication is "of no interest" to the early intervener. Their plan follows what is seen as normal.c1 Amy Pearson and Kieran Rose write that behavioral programs treat autistic behavior as "impaired, lacking functionality, and/or challenging".c2

Orion Kelly is an autistic writer. He describes the early support his son had. He writes the therapy seemed to teach his son "how not to act or look Autistic".c3 What writers ask for is close to what these programs say they do. Jules Edwards favors "inclusive relationship-based interventions". These build relationships and meet a child where they are. They do not demand obedience.c4 The editors of Sincerely, Your Autistic Child wrote about the sources of strength for a good life. It was "almost never a particular kind of therapy or clinical intervention". They write that being accepted, seen as fully human and included matters more.c5

Suggest a correction
Community

How autistic writers describe early intervention and the naturalistic approaches, in their own words. Not a research finding.

Community

How autistic writers describe early support and these programs, in their own words. This is not a research finding.

From nonspeaking writers Nonspeaking

Not every account is negative. Rohit V. Dixit, who types to communicate, writes that the therapies he had as a child were hard but that "early childhood intervention made all the difference for me."c6

Not every account is bad. Rohit V. Dixit types to communicate. He writes that the therapies he had as a child were hard. But he also writes that "early childhood intervention made all the difference for me."c6

Suggest a correction
Nonspeaking

In their own words, written by typing, spelling, handwriting or AAC, some with a communication partner's support. Researchers dispute whether some partner-supported methods let the writer, not the partner, choose the words; the pedia reports these accounts as their writers' and does not settle that question here.

Nonspeaking

In their own words, written by typing, spelling, handwriting or AAC, some with a helper's support. Researchers disagree about whether some helper-supported methods let the writer, rather than the helper, choose the words. The pedia reports these accounts as their writers' and does not settle that question here.

Sources

Research

  1. 1Schreibman L, Dawson G, Stahmer AC, et al. Naturalistic Developmental Behavioral Interventions: Empirically Validated Treatments for Autism Spectrum Disorder. Journal of Autism and Developmental Disorders, 2015. https://doi.org/10.1007/s10803-015-2407-8
  2. 2Waddington H, Reynolds JE, Macaskill E, et al. The effects of JASPER intervention for children with autism spectrum disorder: A systematic review. Autism, 2021. https://doi.org/10.1177/13623613211019162
  3. 3Green J, Charman T, McConachie H, et al. Parent-mediated communication-focused treatment in children with autism (PACT): a randomised controlled trial. The Lancet, 2010. https://doi.org/10.1016/S0140-6736(10)60587-9
  4. 4Rajesh Kumar V, Choi L, Oram J, Binns A. Common and Differing Elements Across Caregiver-Mediated Social Communication Programs Supporting Autistic Children: A Scoping Review With Content Analysis. Autism, 2026. https://doi.org/10.1177/13623613261469913
  5. 5Dawson G, Rogers S, Munson J, et al. Randomized, Controlled Trial of an Intervention for Toddlers With Autism: The Early Start Denver Model. Pediatrics, 2010. https://doi.org/10.1542/peds.2009-0958
  6. 6Rogers SJ, Estes A, Lord C, et al. A Multisite Randomized Controlled Two-Phase Trial of the Early Start Denver Model Compared to Treatment as Usual. Journal of the American Academy of Child and Adolescent Psychiatry, 2019. https://doi.org/10.1016/j.jaac.2019.01.004
  7. 7Geoffray MM, Oreve MJ, Jurek L, et al. Early Start Denver Model effectiveness in young autistic children: a large multicentric randomised controlled trial in two European countries. BMJ Mental Health, 2025. https://doi.org/10.1136/bmjment-2024-301424
  8. 8Fuller EA, Oliver K, Vejnoska SF, Rogers SJ. The Effects of the Early Start Denver Model for Children with Autism Spectrum Disorder: A Meta-Analysis. Brain Sciences, 2020. https://doi.org/10.3390/brainsci10060368
  9. 9Sinai-Gavrilov Y, Gev T, Mor-Snir I, Vivanti G, Golan O. Integrating the Early Start Denver Model into Israeli community autism spectrum disorder preschools: Effectiveness and treatment response predictors. Autism, 2020. https://doi.org/10.1177/1362361320934221
  10. 10Tateno Y, Monden R, Kumagai K, Tateno M. Effectiveness of the Early Start Denver Model (ESDM) as an early intervention for young children with autism spectrum disorder in clinical settings in Japan: A non-randomized controlled trial. PCN Reports, 2026. https://doi.org/10.1002/pcn5.70340
  11. 11Kasari C, Freeman S, Paparella T. Joint attention and symbolic play in young children with autism: a randomized controlled intervention study. Journal of Child Psychology and Psychiatry, 2006. https://doi.org/10.1111/j.1469-7610.2005.01567.x
  12. 12Shih W, Shire S, Chang YC, Kasari C. Joint engagement is a potential mechanism leading to increased initiations of joint attention and downstream effects on language: JASPER early intervention for children with ASD. Journal of Child Psychology and Psychiatry, 2021. https://doi.org/10.1111/jcpp.13405
  13. 13Kasari C, Gulsrud A, Freeman S, Paparella T, Hellemann G. Longitudinal follow-up of children with autism receiving targeted interventions on joint attention and play. Journal of the American Academy of Child and Adolescent Psychiatry, 2012. https://doi.org/10.1016/j.jaac.2012.02.019
  14. 14Pickles A, Le Couteur A, Leadbitter K, et al. Parent-mediated social communication therapy for young children with autism (PACT): long-term follow-up of a randomised controlled trial. The Lancet, 2016. https://doi.org/10.1016/S0140-6736(16)31229-6
  15. 15Sandbank M, Bottema-Beutel K, Crowley LaPoint S, et al. Autism intervention meta-analysis of early childhood studies (Project AIM): updated systematic review and secondary analysis. BMJ, 2023. https://doi.org/10.1136/bmj-2023-076733
  16. 16Crank JE, Sandbank M, Dunham K, et al. Understanding the Effects of Naturalistic Developmental Behavioral Interventions: A Project AIM Meta-analysis. Autism Research, 2021. https://doi.org/10.1002/aur.2471
  17. 17Tiede G, Walton KM. Meta-analysis of naturalistic developmental behavioral interventions for young children with autism spectrum disorder. Autism, 2019. https://doi.org/10.1177/1362361319836371
  18. 18Song J, Reilly M, Reichow B. Overview of Meta-Analyses on Naturalistic Developmental Behavioral Interventions for Children with Autism Spectrum Disorder. Journal of Autism and Developmental Disorders, 2025. https://doi.org/10.1007/s10803-023-06198-x
  19. 19Han DG, Lee Y, Kim HS, et al. Effectiveness and experiences of early intensive behavioral and naturalistic developmental behavior interventions for autism spectrum disorders: a mixed-methods systematic review and meta-analysis. Child and Adolescent Psychiatry and Mental Health, 2025. https://doi.org/10.1186/s13034-025-00997-z
  20. 20Seo E, Ha Y, Jeon P, Yoo E. Early Parent-Mediated Training for Social-Communication Skills in Toddlers and Preschoolers With ASD: A Systematic Review and Meta-analysis. Journal of Autism and Developmental Disorders, 2025. https://doi.org/10.1007/s10803-025-07155-6
  21. 21Mandelli V, Busuoli EM, Godel M, et al. Predicting early intervention outcomes in autism via individual participant data mega-analysis. Molecular Autism, 2026. https://doi.org/10.1186/s13229-026-00736-x
  22. 22Sandbank M, Bottema-Beutel K, LaPoint SC, Feldman JI, Woynaroski T. The Influence of Language and Cognition on Intervention Effects in Young Autistic Children: A Meta-Regression Analysis. Journal of Speech, Language, and Hearing Research, 2026. https://doi.org/10.1044/2026_JSLHR-24-00391
  23. 23Sandbank M, Pustejovsky JE, Bottema-Beutel K, et al. Determining Associations Between Intervention Amount and Outcomes for Young Autistic Children: A Meta-Analysis. JAMA Pediatrics, 2024. https://doi.org/10.1001/jamapediatrics.2024.1832
  24. 24Vivanti G, Lombardo MV, Zitter A, et al. Proportion and Profile of Autistic Children Not Acquiring Spoken Language Despite Receiving Evidence-Based Early Interventions. Journal of Clinical Child and Adolescent Psychology, 2025. https://doi.org/10.1080/15374416.2025.2579286
  25. 25Pickard K, Islam N, Green N, et al. The Challenges Associated with Changing Practice: Barriers to Implementing Naturalistic Developmental Behavioral Interventions in ABA Settings. Behavior Analysis in Practice, 2024. https://doi.org/10.1007/s40617-024-01011-2
  26. 26Stahmer AC, Dufek S, Rogers SJ, Iosif AM. Study Protocol for a Cluster, Randomized, Controlled Community Effectiveness Trial of the Early Start Denver Model (ESDM) Compared to Community Early Behavioral Intervention (EBI) in Community Programs serving Young Autistic Children: Partnering for Autism: Learning more to improve Services (PALMS). BMC Psychology, 2024. https://doi.org/10.1186/s40359-024-02020-0
  27. 27Autistic Self Advocacy Network. For Whose Benefit? Evidence, Ethics, and Effectiveness of Autism Interventions. Autistic Self Advocacy Network, 2021. https://autisticadvocacy.org/policy/briefs/intervention-ethics/
  28. 28Graber J, Graber A. Applied Behavior Analysis at a Crossroads: Reform, Branding, and the Future of Behavior Analysis. Perspectives on Behavior Science, 2025. https://doi.org/10.1007/s40614-025-00462-4
  29. 29Vivanti G. Autism Early Intervention - Progress, Steps Backward, and the Reconciliation of Conflicting Narratives. Current Psychiatry Reports, 2024. https://doi.org/10.1007/s11920-024-01552-x
  30. 30D'Agostino SR, Bhana-Lopez N, Landon TJ, et al. Building on Common Ground: Autistic Adults' and Parents of Young Autistic Children's Perspectives of Early Behavioral Intervention Practices. Behavioral Sciences, 2026. https://doi.org/10.3390/bs16040591
  31. 31Nosova E, Sturrock A, Humphrey N, Leadbitter K. The Views of Autistic Adults on Early Autism Interventions: A Mixed-Methods Systematic Literature Review. Autism, 2026. https://doi.org/10.1177/13623613261460931
  32. 32Sterrett K, Rogers SJ, Estes A, Lord C, Kim SH. Probability of Intervention Benefit in the Early Start Denver Model. Journal of Autism and Developmental Disorders, 2025. https://doi.org/10.1007/s10803-025-07031-3

Community sources

  1. c1Ashburn M, Edwards J, Onaiwu M. I Will Die On This Hill: Autistic Adults, Autism Parents, and the Children Who Deserve a Better World. Jessica Kingsley Publishers, 2023. Chapter "Kristy Forbes".
  2. c2Pearson A, Rose K. Autistic Masking: Understanding Identity Management and the Role of Stigma. Pavilion Publishing and Media, 2023. Chapter "Behavioural interventions".
  3. c3Kelly O. Autism Feels ...: An Earthling's Guide. Dean Publishing, 2023. Chapter "Masking and early intervention therapies".
  4. c4Ashburn M, Edwards J, Onaiwu M. I Will Die On This Hill: Autistic Adults, Autism Parents, and the Children Who Deserve a Better World. Jessica Kingsley Publishers, 2023. Chapter "It Doesn't Have to Be This Way".
  5. c5Ballou EP, daVanport S, Onaiwu MG, eds. Sincerely, Your Autistic Child: What People on the Autism Spectrum Wish Their Parents Knew About Growing Up, Acceptance, and Identity. Beacon Press, 2021. Chapter "Letter from the Editors".
  6. c6Peña E, ed. Leaders Around Me: Autobiographies of Autistics who Type, Point, and Spell to Communicate. Self-published, 2019. Chapter "Rohit Dixit".