Modeling as a Teaching Procedure in ABA

In ABA, modeling is a teaching procedure where you show a learner the exact behavior you want them to do. It is not about fashion or making models. It means the teacher does the skill first, so the learner can see it and copy it.

Modeling can be a word, an action, or a whole routine. A therapist might say a word out loud or act out a play step. A trainer might show a new staff member how to run a program. This page covers how speakers at OpenCEU events use modeling, where it shines, and where it falls short.

What modeling looks like in a session#

Michelle Vences gives a clear picture of modeling with young learners. She describes naming things out loud instead of quizzing the child. Naming things is called "tacting" in ABA. A tact is simply a label for something the learner sees, hears, or touches.

Vences explains that she does not need to ask "what is that?" over and over. She does not ask "is that a butterfly?" Instead, she just says "butterfly." The child hears the word tied to the real thing in front of them. No question is asked, so there is no wrong answer to give.

do a lot of modeling i love modeling i love tacting things labeling things for them

From the talk — Michelle Vences

This is modeling at its simplest. The adult shows the behavior, which here is saying the label, and the child gets a chance to pick it up. When the child copies a spoken word, that copying is close to what ABA calls an echoic. Modeling is the broader idea. It covers any shown behavior, spoken or physical.

Why children learn by watching#

Vences ties modeling to how babies learn in everyday life. She points to a simple example. A baby pretends to talk on the phone. Why? Because the baby has watched a parent do it many times.

In her words, that is "literally observation." The child is watching and learning through the process. No one sat the baby down and taught phone talking as a drill. The baby saw it, and then did it.

This matters for ABA teams. It means modeling is not a strange clinical trick. It is built on a way of learning that children already use. A therapist who names objects or acts out a routine taps into that same habit. The learner watches, then copies.

It also shifts the goal of a session. The therapist is not only asking for answers. They are giving the child lots of clear examples to learn from. Over time, those examples can become the child's own words and actions.

Modeling as a low-demand approach#

Vences also makes a point about the setting. She asks listeners to think about a child's full day. The child may sit in a loud cafeteria. Then they move to a noisy gym. People place demands on them for eight hours straight.

A "demand" in ABA is any request or instruction a learner is asked to follow. After a day full of them, the child arrives at therapy. Then the therapist expects them to follow even more instructions right away.

Vences says teams need different approaches here. They should consider how loud the space is and how demanding it feels. They should look for ways to teach that do not sound like a demand.

Modeling fits that need well. Saying "butterfly" asks nothing of the child. It gives them the word without pressure. For a tired or overloaded learner, that can make a session feel safer. It keeps learning going while lowering the number of questions the child has to face.

Modeling is not teaching on its own#

Crystal Harms brings a very different angle. Her talk focuses on training staff, like RBTs. An RBT is a Registered Behavior Technician, the frontline staff member who works directly with clients. Harms talks about BST, or behavioral skills training. BST is a step-by-step way to teach a skill to another person.

Modeling is one step inside BST. Harms is firm that it is not the whole job.

Modeling is not teaching.

From the talk — Crystal Harms

She says she tells her students this so often they "probably want to throw things" at her. Her supervisees will say, "oh, but I showed them." Her answer is kind but direct. Great, you did step two. Now keep going.

Harms names the real core of BST. She calls rehearsal with feedback BST's "active ingredient." Rehearsal means the trainee practices the skill. Feedback means the trainer tells them what went well and what to fix. Just watching someone, which she calls "exposure," is not enough.

From showing a skill to fluency#

Harms explains why the later steps matter so much. The goal is not just that a trainee knows a skill. The goal is fluency. Fluency means a skill flows easily, quickly, and correctly without much effort.

She links this to clients too. Teams want clients to be fluent in their skills. They want those skills to flow easily. The same standard should apply to staff.

Harms also talks about corrective feedback. That is feedback that tells a person what to change. She notes that some trainees may never have received feedback this way before. It depends on their learning history, meaning their past experiences. So a trainer may need to say the same thing again and again.

She finds that people react better to correction when she sets it up well and repeats it. That kind of patient, repeated practice is what turns a model into a mastered skill. If you want the full picture of her approach, see Press Pause: What the RBT 40-Hour Training Was Really Meant to Do.

Modeling, tacting, and matrix training#

Dr. Jessica Osos looks at tacting in a more structured way. Her research used matrix training. Matrix training is a way to set up targets in a grid. You teach some combinations, and the learner may figure out the rest on their own.

Osos explains that she replicated a 2019 study by Frampton and colleagues. To "replicate" means to repeat a study to see if the results hold up. That earlier study taught learners to name both the color and the shape of an item.

she specifically taught color shape tacting skills and so she taught one diagonal in the color shape and it led to high levels of correct responding for those untrained targets

From the talk — Dr. Jessica Osos

"Untrained targets" are the combinations nobody taught directly. Teaching just one diagonal of the grid led learners to name many of the others correctly. That saves a lot of teaching time.

Osos also notes that matrix training has wide uses. It has been used in pretend play. It has been used for academics like spelling, telling time, and building sentences.

Teaching over telehealth#

The other half of Osos's work is about telehealth. Telehealth means delivering services over a video call instead of in person. Her study taught these tacting skills through a screen.

Osos explains that COVID-19 led to a big rise in telehealth, especially in ABA and autism services. Many places have since gone back to face-to-face work. Even so, she says telehealth is still important.

She gives a few reasons. Some families simply prefer it. Some clients live in rural areas, far from a clinic. In other cases, it may just be the most fitting option for that family.

This connects back to modeling. When a therapist is on a screen, they cannot guide a child's hands. What they can do is show and say things clearly. Naming items, showing examples, and setting up structured targets all still work through video. Osos's research shows that useful labeling skills can be built in this format.

Two views that fit together#

At first, Vences and Harms may seem to disagree. Vences says she loves modeling. Harms says modeling is not teaching. But they are talking about different learners in different settings.

Vences is talking about young children in a busy, demanding day. For them, modeling lowers pressure. It gives clear examples without asking a question each time.

Harms is talking about training adults to do a job well. For staff, a single demonstration is not enough. They need to practice and get feedback until the skill is fluent.

Put together, the lesson is clear. Modeling is a strong first step. It shows the learner what the target looks like. But most skills also need practice, feedback, and time before they are solid. Osos's matrix work adds one more idea. Smart setup of targets can stretch a small amount of teaching into many new skills.

What the research says#

Published research on modeling itself is thin in this collection. The one article source is a 2024 review of state licensure laws (Morris, Donovan, & Switzer, 2024). It is not a study of modeling as a teaching procedure.

That review looked at the APBA Model Act from 2018. APBA is the Association of Professional Behavior Analysts. The Model Act is sample wording states can use when writing their own laws.

The Model Act describes ABA as designing and testing changes to instruction and the environment. The goal is socially significant gains in behavior. That means changes that truly matter in a person's life.

It also lists the tools ABA uses. These include antecedent stimuli, which are cues that come before a behavior. They also include positive reinforcement and other procedures. These tools help people develop new behaviors and use them in the right settings.

The Model Act does not mention modeling by name. A model could be described as one kind of antecedent cue. It comes before a learner's response and shows what to do. That link, though, is not something the review itself draws.

The same act also lists what ABA does not include. Diagnosis, psychological testing, and psychotherapy fall outside it.

FAQ#

What is modeling in ABA?

Modeling is when a teacher shows the learner the behavior first. The learner watches and then tries to copy it. It can be a spoken word, an action, or a full routine.

Is modeling enough to teach a skill?

Usually not on its own. Crystal Harms stresses that modeling is just one step of behavioral skills training. Rehearsal with feedback is what builds real fluency.

How is modeling different from asking a child to label things?

Asking places a demand on the child. Modeling just gives the answer. Michelle Vences says "butterfly" instead of asking "what is that?" This keeps pressure low while the child still hears the word.