Discrete Trial Training: How DTT Works and Where It Fits
Discrete trial training breaks skills into small, repeated teaching trials. Learn the parts of a trial, where DTT shines, and where it falls short.
Key takeaway
Discrete trial training (DTT) is a teaching method that breaks a skill into small parts. Each part is taught through short, repeated practice turns called trials.
Rad N Bad: Beyond the Blue Square — Demystifying Discrete Trial Training, Stimulus Control, and Generalization for RBTs
Free learning path on this topic
Caregiver Collaboration and Parent Training
Leanne Page on why caregivers disengage and how to tell a skill gap from a relationship problem. Then Matt Harrington's seven habits for parent training and a workflow for setting collaboration goals with schools, pediatricians and other partners. Four CEUs, all free.
Experts who cover this
12 recordings on openceu cover discrete trial training.
- Featured talk0.5 PDUs
Rad N Bad: Beyond the Blue Square — Demystifying Discrete Trial Training, Stimulus Control, and Generalization for RBTs
Sean Yocum & Michael CarreroHickory Learning Group
The whole episode is a pros/cons breakdown of discrete trial training: its definition as breaking behavior into discrete teaching units with repeated trials…
- 1.5 CEUs (1 Learning, 0.5 Supervision)
Rad N Bad: The Middle Ground of NET — Escaping the Babysitting Trap with Rigorous Natural Environment Teaching
Sean Yocum, LBA & Michael CarreroHickory Learning Group
The hosts debate whether the field's flight from discrete trial training is warranted, concluding that DTT itself is a powerful technology for building rapid…
- 0.5 Learning CEUs
Rad N Bad: DTT Is Not a Permanent Residence — Integrating Discrete-Trial and Naturalistic Teaching for Generalization
Sean Yocum, LBA & Michael CarreroHickory Learning Group
The hosts define DTT as breaking behavior into discrete teaching units with repeated trials of prompts, responses, and reinforcement, and debate it as a…
- 1 Supervision CEU
Article Club: Treatment Integrity Errors During Discrete Trial Instruction
Pauli Catanese & Sam MuellerAchieve Behavior Partners
Core teaching procedure analyzed throughout, with its components (SD, prompt, response, consequence, intertrial interval) and the effect of errors in each…
- 0.5 Learning CEUs
Rad N Bad: Traditional vs. New ABA — Neutral Science, Values-Driven Application, and Social Validity
Sean Yocum, LBA & Michael CarreroHickory Learning Group
Sean describes traditional ABA's core teaching method as a rigid, repeated antecedent-behavior-consequence trial structure delivered at a table, originally…
- 2 Learning CEUs
Intersection of Cultural Ontogeny, Ecological FBA and the Behavior Instructional Plan
Louis SandlerHickory Learning Group
DTT is critiqued as not developmentally appropriate (tracing its 'bastardized' origin to Catherine Maurice's developmental treatment model) and as failing to…
- 1 PDU
Beyond Table Work
Noah McAmisABA Collective / Camber
Contrasted with NET as the structured, table-based teaching method most RBTs learn first; presented as complementary rather than competing with NET.
- 1 Learning CEU
Press Pause: What the RBT 40-Hour Training Was Really Meant to Do
Crystal HarmsBowsprit Consulting LLC
Harms clarifies that discrete trial training does not require a desk and flashcards and can occur during floor play, and stresses that RBT trainees should…
- 1 Learning CEU
Multiple Exemplar Instruction (MEI): Training for Generalization
Dr. Jessica OsosCopper Consulting Group
Described as the traditional approach of teaching one operant at a time to mastery before moving on; effective for skill acquisition but does not reliably…
- 1 Learning CEU
It’s Complicated: Teaching Relationship Safety That Reflects Real Life
Patricia "Tricia" Lund, CSE, LBA & Carolyn Broner, M.S. Ed.The Behaviorist Bookclub
Broner points out that the stories/videos/role-play delivery format used across nearly all body-safety curricula is essentially DTT and BST applied to safety…
- 1 Learning CEU
Child Development Deep Dive: Early Childhood (2-5 year olds)
Kelly BrzakThe Behaviorist Bookclub
Kelly warns against rigid DTT expectations for toddlers (e.g., sitting at a table, strict point-to-point correspondence) and contrasts it with more…
- 1 PDU
Mastering Your Time: Preparing for Service Delivery
Ashley AndrewInnovative Behavioral Therapy
Andrew notes that fast-paced DTT can make real-time data collection difficult if a learner can't sustain independent engagement during breaks, suggesting a…
On this page · 15 sections▾
Discrete trial training (DTT) is a teaching method that breaks a skill into small parts. Each part is taught through short, repeated practice turns called trials. Each trial has a clear start, a learner response, and a clear ending, such as praise or a correction.
DTT is one of the most studied tools in applied behavior analysis (ABA). It is often used with young children with autism to teach language, play, and school skills. It is also one of the most debated tools in the field. Clinicians agree it works. They argue about where it belongs and how much of it to use. They also argue about how to keep skills from getting stuck at the table.
How a single discrete trial works#
A discrete trial is one small learning turn. The word "discrete" means separate. Each trial stands on its own, with a clear beginning and end. Sean Yocum, a BCBA, puts the basic idea simply.
It's basically breaking behavior into discrete, hence using the word, discrete. teaching units and then basically repeated trials with prompts and responses and reinforcement, right? From the talk — Sean Yocum
Sam Mueller lists the parts of each trial. First comes the SD, which stands for discriminative stimulus. In plain words, that is the instruction or cue, like "touch dog." Next comes a controlling prompt, which is a hint strong enough that the learner will get it right. Then the child responds. Then the adult gives a consequence. Last comes a short pause before the next trial, called the intertrial interval.
Some of the components include an SD, an instruction, a controlling prompt, the child's response, a consequence, and an intertrial interval. From the talk — Sam Mueller
Mueller notes that DTT is also called discrete trial instruction (DTI). It can teach many kinds of skills. Examples include receptive labeling, which means pointing to an item when it is named. Another example is expressive tasks, which means saying the name out loud. As Mueller puts it, this is "stuff we do all day" in many clinics.
The traditional picture of DTT#
Many people picture DTT the way it looked in early autism programs. Yocum describes that "traditional ABA" model as starting in the 1960s. Sessions happened at a table. The room was quiet and free of distractions. The therapist controlled the setting.
So therapist gives that instruction as the antecedent, child gives a response to behavior, therapist gives some sort of reward or correction as the consequence. From the talk — Sean Yocum
That sequence is often called the ABC chain. A stands for antecedent, or what comes before. B is the behavior. C is the consequence. Yocum says the earliest models ran this chain "rapidly and intensely." Programs often called for 20, 30, or even 40 hours a week. Teams believed that heavy dosage was needed for lasting change.
Early programs also leaned on tangible rewards. These were things a child could hold or eat, like snacks, tokens, or small toys. Yocum adds that there was a strong early focus on compliance. That means teaching a child to follow simple adult instructions quickly and correctly. This history matters. A lot of today's criticism of DTT is really criticism of this older, rigid style of delivery.
Why DTT is not the same thing as ABA#
A common mistake is to treat DTT and ABA as one and the same. On the Rad N Bad podcast, Yocum calls this out directly. ABA is the science. DTT is only one method inside that science. When a field mixes up one procedure with the whole science, he says, "technical drift" starts to take over.
His co-host, Michael Carrero, compares this mix-up to how a young child learns words. A child with a family dog may first call every four-legged animal with ears and a tail a "dog." That is an overgeneralization, which means stretching a label too far. Calling all of ABA "DTT" makes the same kind of error.
The hosts also point to a 1980 article by Hayes. It warned that ABA can drift into "narrow technical ritual" when it loses touch with its core ideas. In practice, that ritual can look like this: the table, the chair, the card, the prompt. Then the snack, the data sheet, over and over. Yocum frames the better question this way. The real issue is not "does DTT work?" It is what it works for, under what conditions, at what cost, and toward what goals.
What DTT does well#
The case for DTT is strong. Yocum calls it one of the clearest and easiest to analyze teaching tools in the field.
DTE is one of the clearest, most analyzable teaching technologies that we have within our field. From the talk — Sean Yocum
He lists the features that make it work. DTT offers tight stimulus control, which means the learner responds to the right cue and not to random hints. It packs in many learning chances in a short time. Prompts are clean and easy to fade. Errors get fixed fast. The learner also contacts reinforcement often, so trying feels worth it.
Yocum argues this structure helps one group most. That group is early learners and learners with limited attending, imitation, or discrimination skills. For them, he says, structure "can be exactly what makes learning possible in the first place." His co-host adds one firm condition. It works "if and only if done correctly."
The evidence base is wide. Yocum cites reviews finding that DTT meets evidence-based practice criteria across 58 single-case design studies. It has been used for communication, adaptive skills, play, school readiness, social skills, and job skills. Dr. Jessica Osos agrees on the core point.
We have tons and tons of research to show that DTT is effective From the talk — Dr. Jessica Osos
Osos notes that teaching small sets of targets, such as three at a time, to mastery is well supported during skill acquisition. Skill acquisition simply means the stage when a learner first picks up a new skill.
Treatment integrity: getting each step right#
Because DTT has so many small parts, each one has to be done well. This is called treatment integrity. Mueller defines it as the degree to which a treatment is carried out with precision, the way it was written.
Mueller reviews research showing that lower integrity can make teaching less effective. Two parts stand out. The first is delivering the controlling prompt when it is needed. The second is withholding information, which means not giving away the answer through extra hints. When staff skip or blur these steps, learners may pick up skills more slowly. They may also learn to wait for a hint instead of responding on their own.
This is one reason DTT is so visible in staff training. Each trial is short and has clear steps, so a supervisor can watch and score it. Mueller's talk ties these errors to supervision and to ways of improving procedural fidelity. Procedural fidelity is another name for doing the steps as planned. For a closer look at how specific errors affect learning, see Article Club: Treatment Integrity Errors During Discrete Trial Instruction.
Yocum makes a related point about why integrity problems show up so clearly in DTT. At a desk, bad timing, slow pacing, or long pauses are easy to see. That visibility is a strength. It lets supervisors spot and fix problems early.
DTT does not need a desk#
Many new staff think DTT only happens at a table with flashcards. Crystal Harms, a BCBA who trains Registered Behavior Technicians (RBTs), wants trainees to unlearn that idea.
Her point is simple: DTT does not require a desk or flashcards. Harms explains that DTT can happen on the floor during what looks like play. What makes it DTT is the structure, not the furniture. She wants RBTs to spot the ABC pattern anywhere it happens. That includes natural environment teaching (NET), which means teaching during everyday activities and play. She also notes that the RBT competency assessment says little about applying this skill in natural settings.
Kelly Brzak makes the same point from the other side.
Really good DTT is an NET, excuse me, NET is DTT embedded in the natural environment. From the talk — Kelly Brzak
Brzak stresses that each trial should still have the same parts to be effective, no matter the format. A cue, a response, and a consequence still matter during play. She also suggests looking around the child's home for teaching targets. She urges teams to get buy-in from technicians and families when choosing targets.
When skills get stuck at the table#
The biggest known weakness of DTT is generalization. That means using a skill in new places, with new people, or in new ways. A skill learned at the table may not show up anywhere else.
Osos describes one version of this problem. Traditional DTT often teaches one operant at a time. An operant is a type of language use. Examples include a tact, which means naming something you see, and a mand, which means asking for something you want. A third example is a listener response, which means pointing when something is named.
Traditional discrete trial or DTT approaches often focus on teaching one operant as a... ...at a time and mastering it to a criterion before moving on. From the talk — Dr. Jessica Osos
The catch, Osos says, is that this does not always transfer across operants. Clinicians often see learners who "know the word" but cannot use it in flexible ways. If a word is taught as a tact, the child may only learn it as a tact. The team then has to teach it again as a mand, and again as a listener skill.
The Rad N Bad hosts describe a similar pattern in daily practice. A learner may do well with a paper flashcard at a desk. Then, two feet onto the clinic floor, the whole skill falls apart. That is not a reason to drop DTT. It is a reason to plan for generalization from the start.
DTT inside multiple exemplar instruction#
One way to build more flexible skills is multiple exemplar instruction (MEI). In MEI, the teacher rapidly switches between different ways of using the same concept. For example, a child might name a dog, then ask for the dog toy, then point to the dog. Osos is clear that MEI does not replace DTT. It runs on it.
MEI, you need to use DTT within MEI to teach and you have to use the same strategies that you use during DTT when you're using MEI From the talk — Dr. Jessica Osos
Osos says every MEI trial should follow the typical DTT trial format. First, gain the learner's attention. Then present the SD. The prompt may come between the SD and the response. Then the child responds. Then comes the consequence, either reinforcement or error correction.
What changes in MEI is the order and mix of trials, not the trial itself. Osos also explains that MEI and DTT serve different goals. Isolated operant teaching with DTT is strong for early acquisition. MEI becomes necessary when the goal is generative learning. That means new language that shows up without direct teaching, especially for learners who do not yet link operants on their own. The two can be used side by side or combined.
The same logic outside autism programs#
DTT's basic logic shows up in places people may not expect. Dr. Carolyn Broner studies body safety programs for children. She notes that almost all of them share one delivery model. Stories, videos, games, and puppets introduce rules. Role play and behavioral rehearsal give children practice using them.
When you look at the instructional logic underlying all of these, it should look familiar. It's essentially DDT and BST applied to safety content. From the talk — Carolyn Broner
BST stands for behavioral skills training. It is a teaching package of instruction, modeling, rehearsal, and feedback. Broner says the structure itself is sound. The evidence shows it can teach a rule.
Her concern is the generalization assumption. That is the belief that a child learns and rehearses a rule in a structured setting. The belief also assumes the child will then use that rule in real life. For many young children, that leap is not automatic. It is harder when a situation is emotionally overwhelming. It is harder when it involves someone familiar. It is also harder when danger does not look obvious. This echoes the generalization concerns raised in clinical DTT.
Teaching in the moments between trials#
Noah McAmis trains RBTs and focuses on what happens around the table. He points out that most RBTs learned DTT first.
We learned DTT first. You learn to set up the table, run the trials, take the data. From the talk — Noah McAmis
What often goes untrained, McAmis says, is the "in-between." That is the dead time or reinforcer time between trials. Staff wait for the timer, clean up, and reset. Added up, that can be a big chunk of every session.
McAmis describes a session where the program was run, the data was taken, and on paper it looked like a successful session. Yet three real chances to engage with the child went right past the RBT. He frames this as a perception problem, not a knowledge problem. The RBT knew the skills. The RBT just saw teaching as something that only happens in one place, with one set of materials, in one format.
McAmis says the in-between time is where some of the most valuable pairing can happen. Pairing means building a positive bond so the adult becomes linked with good things. That bond, he says, is what earns cooperation later at the table.
DTT with toddlers and young children#
Brzak's work on early childhood (ages two to five) raises a caution about how DTT goals get written for very young kids. She shares several goals she sees as poor fits for this age. One reads like this.
Client will complete DTT at a table for 80% of the duration of therapy sessions across 100% of therapy sessions. From the talk — Kelly Brzak
Her response is direct: "Please, please, please play with these kids." She notes that crying, whining, throwing items, and avoiding tasks are common and age appropriate in toddlers. They do not always mean a child is being difficult.
Brzak also shares advice from Mark Mullady, a BCBA and a dad. He warned against overly strict rules on point-to-point correspondence for small children who are learning to talk. Point-to-point correspondence means requiring an exact match between the model and the child's response. Even five-year-olds may have a lisp or a stutter, or mix up sounds. Holding them to exact matches can be unfair and, in his words, sometimes "cruel."
Brzak's practical tips include reducing questions and demands. Fewer demands give young children fewer reasons to avoid or ignore the adult.
Where experts disagree: how much table time is too much#
The sources agree on a lot. DTT works. Trials need the right parts. Generalization must be planned. But they push in different directions on one key question. Has the field moved too far away from structured table work, or not far enough?
Yocum argues the field has overcorrected.
The pendulum swung so hard away from discrete trial training that clinicians are terrified to even look at a flashcard. From the talk — Sean Yocum
In his view, many programs fled to naturalistic teaching that is not done well and is barely teaching at all. He offers a pointed take on where the harm really came from.
DTT never traumatized anyone. Terrible, robotic, disconnected practitioners and people running bad DTT are what caused the problems From the talk — Sean Yocum
Yocum says the field turned a powerful tool into "a joyless assembly line." Then, instead of fixing delivery, it "threw the baby out with the bathwater." He goes further, calling it a "hot take" that NET can let weak clinicians hide. At a desk, poor timing and pacing are easy to see.
Brzak reads the risk the other way, at least for toddlers. Her concern is not too little table time. It is goals that keep very young children seated for most of every session. Her answer is more play, fewer demands, and gentler standards for early speech.
McAmis lands in the middle. He says plainly that some clients and targets need the table. Learning to sit, tolerate demands, and work through something not fun are real skills. They are usually built through structured practice. He warns RBTs not to tell their BCBA they are "done with the table." Yet his whole course pushes staff to teach beyond it.
Table work matters. Structured instruction matters. From the talk — Noah McAmis
These views can fit together, but the tension is real. The disagreement is mostly about emphasis and age. It is also about whether the bigger risk today is rigid DTT or sloppy NET.
Using DTT as a ladder, not a home#
Carrero offers a clear rule for judging when DTT is being used well. If DTT builds a skill and then fades out, it is good ABA. If a learner gets stuck in it, it is not. He compares DTT to a ladder. It is powerful and important, but it is meant to be climbed off of onto something else.
Carrero also explains why teams use DTT at all.
We're sterilizing the environment in order to contrive learning opportunities for future reference point. From the talk — Michael Carrero
In plain words, DTT removes distractions on purpose so a skill can be taught cleanly. Then the skill is moved into real life. Carrero notes that this is useful for teaching isolated skills before a learner can safely practice in certain places. He adds that people do a level of DTT all the time. Think of how someone practices scales before playing a full song on guitar.
The hosts' closing advice is to not confuse the procedure with the purpose. The goal is independence. The hosts put it bluntly: build independence, not robots. A clinician who keeps that goal in view will use DTT when structure helps. They will blend it with naturalistic teaching. They will fade it once the skill can live outside the session.
What the research says#
Published research backs up many points the speakers make. A scoping review describes DTT as highly structured, teacher-led teaching. It breaks complex skills into simpler target responses and teaches them through fast-paced trials (Frank‐Crawford et al., 2024). The review also notes the method goes by several other names, including discrete-trial training and discrete-trial instruction.
On staff training, studies support the integrity concerns Mueller raises. One study used behavioral skills training with new staff who had never used DTT. Correct teaching rose from about 58 to 70 percent at baseline to 96 to 99 percent after training. Those gains held 30 days later (Clayton & Headley, 2019).
Research has also tested DTT in new formats. One study compared DTT delivered by telehealth with in-person DTT for children with autism learning to label occupations. It found little difference in how well children learned (Lindgren et al., 2023). Another study compared traditional table-top DTT with embedded DTT, where targets are taught during an activity. Accuracy was similar across both formats for one child with autism, and the study also tracked problem behavior (Haq & Aranki, 2019). This supports the speakers' point that trial structure matters more than the table itself.
FAQ#
What is the difference between DTT and NET? DTT is often run at a table with planned, repeated trials. NET teaches during play and daily routines. As Brzak notes, good NET still uses the same trial parts, just inside the natural environment.
Is DTT the same as ABA? No. ABA is the science of behavior. DTT is one teaching method within that science. Many other methods exist, and DTT is best used alongside them.
Why do skills learned in DTT sometimes not generalize? DTT often teaches one skill in one setting with one set of materials. Without planning, the learner may not use the skill in new places or new ways. Teams use methods like MEI and naturalistic teaching to help skills spread.
Turn this topic into a CEU
You just studied this. Now get credit for it.
Watch Rad N Bad: Beyond the Blue Square — Demystifying Discrete Trial Training, Stimulus Control, and Generalization for RBTs with Sean Yocum & Michael Carrero and earn 0.5 free BCBA CEUs. Detailed certificate, delivered the moment you finish.
Keep going with a free learning path
Caregiver Collaboration and Parent Training
Leanne Page on why caregivers disengage and how to tell a skill gap from a relationship problem. Then Matt Harrington's seven habits for parent training and a workflow for setting collaboration goals with schools, pediatricians and other partners. Four CEUs, all free.
Want the primary literature? Read the Discrete Trial Training research roundup on our sister site, Behaviorist Book Club.


