DRAGENwiki · Article

Naturalistic Teaching

Sources checked. Clinical review pending.

Naturalistic teaching is a family of ways to teach skills inside everyday play and routines, starting from what the learner already wants, so that using the new skill gets the learner what they were after.

Naturalistic teaching in 60 seconds

  • It happens inside real life. Teaching is built into play, meals, routines and the places a learner already spends time. [2]
  • It starts from the learner's interest. The adult follows the learner's lead and teaches around what the learner chose. [1]
  • The reward fits the moment. Asking for the ball gets the ball. The reward is the thing the learner was already after. [1]
  • It is still ABA. Each teaching moment has a cue, a response and a consequence, and progress is measured. [1]
  • It is recognized as an evidence-based practice for autistic children, youth and young adults by a national review. [2]

Jump to: What it is · Where it came from · The core moves · What stays the same · What the research says · Naturalistic teaching and DTT · Choice, assent and the learner's lead · Who does it well · An example · Myths · Words to know · Need a PDF? · More from DRAGEN · Sources

What it is

A national review of autism research calls this practice naturalistic intervention. It describes techniques "embedded in typical activities and/or routines in which the learner participates." [2]

The review says these practices build skills "based on learners' interests." [2] They are used during everyday routines at home or in the classroom. [2]

You will see three names.

  • Naturalistic teaching. The everyday name. This page uses it.
  • Naturalistic intervention (NI). The name in the national review. [2]
  • Naturalistic developmental behavioral interventions (NDBIs). The name researchers use for programs that blend behavioral science with the science of child development. [1]

Schreibman and colleagues describe NDBIs as programs that "involve shared control between child and therapist" and "utilize natural contingencies." [1] In plain words: the learner helps steer, and what happens after a response grows out of the activity itself.

Named programs. Researchers list many programs in this family, including incidental teaching, pivotal response treatment (PRT), the Early Start Denver Model (ESDM), Enhanced Milieu Teaching, reciprocal imitation training (RIT), Project ImPACT, JASPER, SCERTS and Early Achievements. [1] Some focus on one area, such as social communication (JASPER, RIT). Others are comprehensive and reach communication, thinking, motor and daily living skills (ESDM). [1]

The national review names three manualized programs under naturalistic intervention: JASPER, Milieu Teaching and PRT. PRT used to be counted as its own practice; the review now groups it under naturalistic intervention. [2]

Where it came from

Discrete trial teaching (DTT) became popular with families of autistic children. By the mid- to late-1980s, research showed that highly structured teaching such as DTT, while effective, sometimes led to four problems: [1]

  1. Skills that did not carry over to new places and people.
  2. Escape and avoidance behavior during teaching.
  3. A lack of spontaneity.
  4. Too much dependence on prompts.

Researchers turned to ways of teaching inside the learner's own activities. One of them, incidental teaching, goes back to work by Hart and Risley in 1968 and 1975. [1]

A turning-point study. In 1987, Koegel, O'Dell and Koegel worked with nonverbal autistic children. They started with a traditional clinic format, where the therapist presented the instructions, prompts and rewards for correct answers. Then they changed four things: [6]

  • Teaching items were functional and varied.
  • Rewards were natural.
  • Tries at talking were rewarded, not only correct words.
  • Teaching happened inside a natural back-and-forth.

These changes "resulted in broadly generalized treatment gains." [6]

The core moves

Schreibman and colleagues list features that NDBIs share. [1] Here they are in plain words.

  • Follow the learner's lead. Teaching starts from what the learner is already interested in. This is "also referred to as following the child's lead or interest, or child choice." [1]
  • Set up the space. The adult arranges things so the learner has a reason to communicate. Examples include keeping a favorite thing in sight but out of reach, using a toy that needs help to work, waiting with an expectant look, and playfully changing a familiar routine. [1]
  • Use natural rewards. The reward is part of what the learner wanted. If a learner copies a play action with a favorite toy, they keep playing with that toy. They do not get a token or a different toy for it. [1]
  • Reward the try. Many programs use "loose shaping or reinforcing attempts." A good try counts, even when it is not perfect yet. [1]
  • Mix easy and new. Skills the learner already knows are mixed in with new ones, which helps keep motivation up. [1]
  • Prompt, then fade. Adults still give planned help and fade it. Planned prompting, rewards and ongoing data are what set NDBIs apart from developmental programs that do not use ABA principles. [1]
  • Take turns. The learner and adult share control of the play. The authors note this practice has had limited study on its own. [1]
  • Model a little ahead. The adult models words or play slightly beyond what the learner does now. [1]
  • Copy the learner. The adult imitates the learner's play, sounds or actions. Researchers call this contingent imitation, mirroring or reciprocal imitation. [1]
  • Widen attention. Use many different materials, places and people, so the learner does not depend on one small detail to respond. [1]

What stays the same

Naturalistic teaching is not "no structure." Schreibman and colleagues list what it shares with all good ABA. [1]

  • The three-part contingency. Every teaching moment has an antecedent, a response and a consequence. The authors write: "Thus, they represent ABA treatment." [1] See three-term contingency.
  • Manuals. NDBIs are written down, so they can be taught and checked. A manual is "necessary, but not sufficient." Staff also need coaching and feedback. [1]
  • Fidelity. Teams check that the program is run the way it was written. [1] See procedural fidelity.
  • Individual goals. Goals are set for each learner. [1]
  • Ongoing measurement. Data can be taken trial by trial, by intervals, with short probes, or with skill checks every month or quarter. [1]

What the research says

The national review. Naturalistic intervention meets the review's bar for an evidence-based practice. [2] Studies reached many outcome areas, including communication, social skills, joint attention, play, thinking skills, school readiness, academic skills, daily living skills, challenging behavior, motor skills and mental health. [2]

A 2019 meta-analysis. Tiede and Walton pooled 27 group-design studies of NDBIs. [4]

  • Small, significant effects on expressive language, autism symptoms and play.
  • Larger effects on social engagement and thinking skills.
  • Only marginal effects on joint attention and on understanding language.
  • For joint attention, the size of the gains depended on how many hours professionals were involved.
  • Signs of publication bias for language outcomes, which means language gains may look bigger than they are.

A 2023 review of early childhood studies. Sandbank and colleagues reviewed 252 studies with 13,304 children under age 8. [3]

  • In randomized controlled trials, NDBIs showed significant effects on adaptive behavior, language, play, social communication and measures of autism characteristics. [3]
  • Nearly two thirds of all outcomes in the review were at risk of detection bias, a kind of bias in how outcomes are measured. [3]
  • When outcomes at high risk of detection bias were left out, only one NDBI effect remained: on measures of autism characteristics. [3]
  • The authors concluded that NDBIs improve core challenges of autism, "particularly difficulties with social communication." [3]
  • They also warned: "Adverse events were poorly monitored, but possibly common." [3] This was a problem across the studies they reviewed, not only for NDBIs.

An older review. In 2001, Delprato reviewed 10 controlled studies that compared normalized (naturalistic) language teaching with discrete-trial teaching. [5]

  • In all eight studies that measured language, normalized teaching was more effective. [5]
  • In both studies that looked at parents' feelings during teaching, normalized teaching led to more positive affect. [5]

What this means for a family. The evidence is real and growing, but it is not a promise for any one child. Ask your team how they will measure progress, and what they will watch for if something is not going well.

Naturalistic teaching and DTT

Delprato's review gives a clear side-by-side. [5]

  • The discrete-trial approach: highly structured, direct teaching sessions; the teacher starts each trial; artificial rewards; and response shaping. [5]
  • The normalized approach: loosely structured, indirect teaching in everyday situations; the child starts; natural rewards; and liberal rules for giving rewards. [5]

These are two ends of a range, and NDBIs themselves blend behavioral and developmental ideas. [1] At DRAGEN, the question is not which camp to join. A learner's plan can use both, chosen for that learner and checked with data. See discrete trial teaching.

Choice, assent and the learner's lead

This section is DRAGEN's practice standard, built on DRAGEN's four pillars: autonomy, beneficence, nonmaleficence and justice.

  • Shared control is the point. Naturalistic teaching already asks the adult to follow the learner. Keep it that way when the plan gets busy.
  • Keep the setup playful. Holding a toy out of reach should feel like a game. If it upsets the learner, stop and rethink the plan. A learner never has to earn back basic comfort, food or care.
  • "No" is information. A learner who walks away, says "no" or picks something else is telling you something. That is not "noncompliance," and it is never a treatment target.
  • Keep safety separate from dissent. If someone could get hurt, keep everyone safe first. A learner's "no" is not a safety problem.
  • Teach what matters to the learner and family. Choose goals that make life better for the learner, such as asking, playing and connecting with people, not goals that only make a learner look less autistic.

Assent never replaces the consent a parent or guardian gives. See assent and assent withdrawal.

Who does it well

  • Trained adults, with coaching. A manual alone is not enough. Staff need coaching and feedback. [1] See behavioral skills training.
  • Parents. Researchers recommend teaching parents these strategies, so skills carry over to home and the learner gets more chances to learn every day. [1] See parent training.
  • Professional time matters. In one meta-analysis, gains in joint attention depended on how many hours professionals were involved. [4]
  • A BCBA who plans and reviews. At DRAGEN, a BCBA reviews every clinical decision, including which naturalistic program fits a learner.

An example

This is a made-up example. It is not a real child.

Learner A is three and loves bubbles. The team's goal is for Learner A to ask for things with a word or a word-like sound.

  1. The setup. The adult blows bubbles once, then holds the wand and waits, smiling.
  2. The try. Learner A reaches and says "buh."
  3. The model. The adult says "bubbles," one small step ahead of Learner A.
  4. The reward. The adult blows bubbles right away. "Buh" was a good try, so it counts.
  5. Another chance, later. At snack, the crackers are in a jar Learner A cannot open. Learner A hands over the jar. The adult waits, then models "open." Learner A says "oh," and the adult opens it.

After a few minutes, Learner A walks over to the trains. The adult follows and plays trains. That is information, not a problem. New chances to ask come up inside the train game.

The adult notes each chance: did Learner A ask alone, or after a model? Today, Learner A asked alone in 6 of 10 chances. Next week, the team plans to check whether Learner A asks for bubbles at a grandparent's house, too.

Common myths

  • "Naturalistic teaching is just playing." Each teaching moment still has a cue, a response and a consequence, and progress is measured. [1]
  • "It is not ABA." Researchers describe NDBIs as built on the three-part contingency and write that "they represent ABA treatment." [1]
  • "It has no structure." NDBIs use manuals, fidelity checks, individual goals and ongoing data. [1]
  • "It is proven best for every child." When outcomes at high risk of measurement bias are left out, most effects are no longer significant, and harms have not been tracked well. [3] No study can promise results for one child.
  • "Following the lead means teaching nothing." The adult still prompts, fades help and rewards new skills, inside what the learner chose. [1]

Words to know

  • Naturalistic intervention (NI): the national review's name for teaching built into everyday routines and activities. [2]
  • NDBI: naturalistic developmental behavioral intervention, a program that blends behavioral and developmental science. [1]
  • Child-initiated teaching episode: a teaching moment that starts from what the learner chose. [1]
  • Environmental arrangement: setting up materials so the learner has a reason to communicate. [1]
  • Natural reinforcement: a reward that is part of what the learner wanted. [1]
  • Loose shaping: rewarding good tries, not only perfect answers. [1]
  • Shared control: the learner and the adult both steer the activity. [1]
  • Contingent imitation: the adult copies the learner's actions or sounds. [1]
  • Fidelity: how closely a program is run the way it was written. [1]

Need a PDF?

Create it here, for free: the DRAGEN app. Use it at your next session.

Or print one we already made, also free:

More from DRAGEN

Related DRAGENwiki pages: incidental teaching · natural environment teaching · pivotal response treatment · discrete trial teaching · prompting and fading · shaping · generalization · pivotal behavior · imitation · parent training · behavioral skills training · assent · communication and choice · mand · preference assessment · positive reinforcement · motivating operations · procedural fidelity

What remains unknown

  • Which program works best for which learner. As of 2015, no large randomized trial had compared DTT with an NDBI, or two NDBIs with each other. [1]
  • Which core moves matter most. Some, such as balanced turns, have had little study on their own. [1]
  • How often harms happen. Adverse events were poorly tracked in the early intervention studies reviewed. [3]
  • How big the language gains really are. One meta-analysis found signs of publication bias for language outcomes. [4]
  • This page is general information, not medical advice. It was last checked on November 12, 2026.

Sources

  1. Schreibman, L., Dawson, G., Stahmer, A. C., Landa, R., Rogers, S. J., McGee, G. G., Kasari, C., Ingersoll, B., Kaiser, A. P., Bruinsma, Y., McNerney, E., Wetherby, A., & Halladay, A. (2015). Naturalistic developmental behavioral interventions: Empirically validated treatments for autism spectrum disorder. Journal of Autism and Developmental Disorders, 45(8), 2411–2428. https://doi.org/10.1007/s10803-015-2407-8 (full text: https://pmc.ncbi.nlm.nih.gov/articles/PMC4513196/)
  2. Steinbrenner, J. R., Hume, K., Odom, S. L., Morin, K. L., Nowell, S. W., Tomaszewski, B., Szendrey, S., McIntyre, N. S., Yücesoy-Özkan, S., & Savage, M. N. (2020). Evidence-based practices for children, youth, and young adults with autism. The University of North Carolina at Chapel Hill, Frank Porter Graham Child Development Institute, National Clearinghouse on Autism Evidence and Practice Review Team. https://ncaep.fpg.unc.edu/wp-content/uploads/EBP-Report-2020.pdf
  3. Sandbank, M., Bottema-Beutel, K., Crowley LaPoint, S., Feldman, J. I., Barrett, D. J., Caldwell, N., Dunham, K., Crank, J., Albarran, S., & Woynaroski, T. (2023). Autism intervention meta-analysis of early childhood studies (Project AIM): Updated systematic review and secondary analysis. BMJ, 383, e076733. https://doi.org/10.1136/bmj-2023-076733 (full text: https://pmc.ncbi.nlm.nih.gov/articles/PMC10644209/)
  4. Tiede, G., & Walton, K. M. (2019). Meta-analysis of naturalistic developmental behavioral interventions for young children with autism spectrum disorder. Autism, 23(8), 2080–2095. https://doi.org/10.1177/1362361319836371 (https://pubmed.ncbi.nlm.nih.gov/31018655/)
  5. Delprato, D. J. (2001). Comparisons of discrete-trial and normalized behavioral language intervention for young children with autism. Journal of Autism and Developmental Disorders, 31(3), 315–325. https://doi.org/10.1023/a:1010747303957 (https://pubmed.ncbi.nlm.nih.gov/11518484/)
  6. Koegel, R. L., O'Dell, M. C., & Koegel, L. K. (1987). A natural language teaching paradigm for nonverbal autistic children. Journal of Autism and Developmental Disorders, 17(2), 187–200. https://doi.org/10.1007/BF01495055 (https://pubmed.ncbi.nlm.nih.gov/3610995/)

Pages that link here