Operational Definition in ABA: What It Is and Why It Matters

An operational definition describes behavior so clearly that any two observers would score it the same way. Learn how ABA experts write and use them.

Key takeaway

An operational definition describes a behavior so clearly that any two people watching would score it the same way. It says what the behavior looks like, what it sounds like, and how it plays out over time.

Dunder Mifflin’s Guide to Training and Onboarding: Lessons from The Office
Watch the full CEU recording

Dunder Mifflin’s Guide to Training and Onboarding: Lessons from The Office

Mellanie Page · 1 CEU · 57 min
Watch on openceu.com →

Free learning path on this topic

Supervision, Done Right

Four supervision CEUs that build on each other. What reflective supervision is, how to build a supervision system, how to assess supervisee competence instead of counting hours, and how supervision affects provider outcomes. Each course is free and issues its own certificate.

Start the path →4 courses · 4 supervision CEUs · Watch at your own pace

Experts who cover this

6 recordings on openceu cover operational definition.

An operational definition describes a behavior so clearly that any two people watching would score it the same way. It says what the behavior looks like, what it sounds like, and how it plays out over time.

In applied behavior analysis (ABA), this is the first step before you measure anything. Speakers from several OpenCEU talks show how one clear definition shapes your work. It affects your data, your treatment plans, your staff training, and even your own bias. Each section below draws on their own words.

What makes a definition operational#

Sean Yocum traces the idea back to the start of the field. He points to Baer, Wolf, and Risley in 1968. Their very first dimension of ABA is "behavioral." That means we measure what people actually do. We do not measure what someone thinks they did, or what they remember doing.

Yocum then names the standard that came next. Hawkins and colleagues set it in 1977.

Hawkins et al., 1977, established that a behavioral definition has to be objective, clear, and complete. So two independent observers score the exact same response.

From the talk — Sean Yocum

In plain words, a definition is operational when it removes guesswork. A loose word leaves room for opinion. Mackenzie Sandler gives a good example of this. She notes that a note saying someone was "aggressive" is not the same as "physical aggression." The first word is a judgment. The second can be pinned down to actions you can see and count.

So the test is simple. Hand your definition to a new person. Ask them to watch the same session as you. If you both mark the same moments, the definition works. If you argue about what counted, it still needs work.

The Triple Cs: clear, concise, and complete#

Mike Carrero co-hosted that talk with Yocum. He teaches a short rule to students and to RBTs. An RBT is a registered behavior technician, the staff member who often works one on one with the learner.

Triple Cs for definitions. Clear, concise, and complete.

From the talk — Mike Carrero

Carrero explains what "complete" really means. You know the timing of the behavior from start to finish. You know what it looks like and what it sounds like. You know "every nitty-gritty aspect of it."

Then he names the part he sees as most important. The people using the definition have to agree on it.

if there's subjectivity in what we're about to observe objectively, now we have a problem.

From the talk — Mike Carrero

He asks listeners to picture a common moment. You read a definition and think, "Hmm, what about this?" You are not sure whether to count it or skip it. Carrero says that small moment matters. It shows the science has already started to slip.

Notice that Carrero's list uses "concise" where Hawkins used "objective." The two versions still point to the same goal. Both want a definition that two people can use the same way, every time.

Why your data depend on it#

Yocum frames the operational definition as the base of the whole graph. In research terms, the learner's behavior is the dependent variable. That is the thing we measure to see if treatment is working.

He cites Horner and colleagues from 2005. Their quality standards for single-case research set a clear bar. Data must be measured again and again. They must also be backed by IOA of at least 80% across at least 20% of sessions in every condition. IOA stands for interobserver agreement. It is a check of how often two observers record the same thing.

Here is the link to definitions. Two people can only agree if they are counting the same behavior. A fuzzy definition makes good IOA almost impossible.

Yocum then walks into a real clinic at 4:50 p.m. He describes one RBT or a tired caregiver. They are juggling 12 active programs on a tablet app with 40 drop-down menus. Every graph needs a data point before anyone goes home. In that rush, a vague definition gets even harder to apply.

Carrero adds a fair note. He says they are not calling people incompetent. The system has flaws that set traps. No one is acting with bad intent. A clear definition is one way to make those traps smaller.

Operational definitions in treatment plan goals#

Dr. Jessica Osos looks at definitions through the lens of medical necessity. Medical necessity is the standard insurance payers use to decide whether to fund a service. Osos says each goal must be measurable, observable, and criterion based. The goal has to describe what the behavior looks like, how it will be measured, and when it should occur.

She lists the parts every strong ABA goal needs. First is an operational definition, so anyone can recognize the behavior. Then comes the baseline, which is how the client performs before treatment. Next is the measurement system, such as frequency, duration, or percentage. After that come mastery criteria, a time frame, the conditions, and a plan for generalization and maintenance. Generalization means the skill shows up in new places. Maintenance means it lasts over time.

So each goal is including an observable, measurable definition. It has clear mastery criteria. There's a plan for how you're going to check for generalization or confirm that generalization occurred. And then there's also maintenance expectations.

From the talk — Dr. Jessica Osos

Osos also stresses stating the baseline in the goal itself. A mand is a request. Her sample reads like this: during assessment, Alex manded with two-word vocal phrases with 50% independence. She adds that each goal needs a clear reason tied to a functional impairment. A mand for a break, for example, can replace problem behavior when tasks show up. You can see her full case examples in Writing Defensible ABA Treatment Plans - Part 2: Applying the Medical Necessity Framework & Case Examples.

Labels versus definitions#

Mackenzie Sandler comes at this topic from the angle of unconscious bias. She recalls her early years as a new BCBA, or Board Certified Behavior Analyst. She sat through many appeals and peer reviews with insurance. She also had many talks with families about their child's plan.

Families often expected the plan to show their child's worst day in their worst moment. Everything in it sounded bad. It was not strength based. Sandler thinks that is the wrong focus. What matters is a definition that others can repeat and understand.

I like to really make sure that the definition is replicable and everyone can identify what it means.

From the talk — Mackenzie Sandler

Sandler says she does not really care what a behavior is called. She gives a clear case. One team may write "tantrum behavior." Another may write "dysregulated behavior." The actual operational definition might be exactly the same. Only the label changes.

Her point is that labels carry weight. A word like "violent" shapes how people see a learner before they ever meet them. A replicable definition keeps the focus on actions. It also makes the plan easier for a family to read without feeling judged.

Using definitions to check your own reactions#

Sandler takes the idea one step further. She applies it to the adults in the room, not just the learner. She describes strong feelings that come up at work. Maybe you feel frustrated every time a parent asks for something. Maybe a tech did not take data on a goal again.

She warns that these big feelings open the door to bias. When you are upset, unconscious biases can take over your choices. That matters when you make clinical or supervision decisions.

Her fix is to name and define the behavior. That could be your own behavior, the learner's behavior, or a parent's behavior.

The more we can operationally define it, then you can break it down. You can actually identify it, um, and use your behavior analysis, right?

From the talk — Mackenzie Sandler

Next, she asks, "What assumptions am I making?" Sandler then points the finger at herself. She had just assumed a family wanted motor stereotypy to stop so their child would not look autistic. Motor stereotypy means repeated body movements, like hand flapping. She admits she may be wrong. The family may fear bullying. They may not understand the behavior. Or they may have seen another child get hurt. Defining the behavior first gives room for those other reasons.

Clear expectations for staff#

Mellanie Page uses clips from the TV show The Office to make her point about training. In the clips, staff keep breaking rules that were never spelled out. Page says the lesson is to turn vague hopes into clear expectations.

Most failures and shortcomings of new team members are due to lack of clear expectations.

From the talk — Mellanie Page

She describes a common scene. A supervisor is asked what they told a new RBT to do. The answer is often "whatever we train them on." Or it is "they should already know." Page says that is just not sufficient. The most useful thing a leader can do, early and often, is set clear expectations.

That means naming who will do the behavior and what the behavior is. It also means where it happens, when it happens, and how it should be done. Page admits this can be a painstaking process. But she sees it as the same skill we already use with clients.

Policies and procedures should be a set of operational definitions and task analyses of desired behavior.

From the talk — Mellanie Page

A task analysis breaks a skill into small, ordered steps. Paired with an operational definition, it tells staff exactly what success looks like.

Show what it is not#

Page adds a detail that many people skip. Training should leave no room for doubt about what should be done. It should also make clear what should not be done. She pauses to stress the word "shouldn't."

We often tell people what to do. We forget to show non-examples, which are cases that do not count. Page found the non-examples in her clips helpful for this reason. Seeing the wrong version makes the right version easier to spot.

She also talks about why this matters. Page notes that we are all biased by our own understanding of information. Our view might be X. Someone else might see Y or Z. Without a shared definition, two people can read the same rule and do two different things. Clear expectations put the whole team on the same page.

For anyone who thinks this level of detail is overkill, Page has an answer. She asks them to think of real examples where a vague rule went wrong. Her clips are full of them. The same is true on a clinic floor. A definition with both examples and non-examples saves time later, because fewer people have to guess.

Shared terms across disciplines#

In a separate talk, Sean Yocum turns to teamwork across fields. A learner may see an OT, a PT, a speech therapist, and a teacher. OT means occupational therapy. PT means physical therapy. Each field has its own words.

I think that's a, that's a hard one that we have to just operationally define your terms versus my because OT versus PT versus speech versus school versus all of the domains, we all use the same words, but they mean very different things.

From the talk — Sean Yocum

Yocum uses "functional communication" as one example. Several fields use that phrase. But each one may mean something a little different by it.

His answer is not to argue about who is right. He says it is about what will help the individual. The team can take a piece of one definition and a piece of another. Then they can build one shared definition that supports every area.

This fits his wider view of goals. In the same talk, he says he should be able to justify any goal he writes. If he cannot explain why it matters years down the road, it should not be a goal. A shared, operational definition makes that kind of explanation possible across the whole team.

A gap in the research base#

B. Kuereine Gray brings the idea into supervision research. Gray speaks about reflective supervision, a style of supervision built around guided reflection. Gray notes it has been linked to less burnout. It has also been linked to more independent decision making, more clinical skill, and better outcomes.

But Gray also names barriers. Research on reflective supervision in ABA is still limited. Resources are limited too. And there is a problem with how the research defines what happens.

a lot of the research relies heavily on theoretical definitions of what should happen. Not operational. definitions of observable behavior.

From the talk — B. Kuereine Gray

A theoretical definition describes an idea. An operational definition describes actions you can see. Gray also flags mentalistic terms as another issue. Mentalistic terms describe inner states, like thoughts or feelings, instead of behavior.

Gray adds that our field has some bias against trusting research from close fields, such as psychology and social work. ABA asks clinicians to define client behavior with care. Gray's point suggests the same standard should apply to how we study our own supervision practices.

What the research says#

Outside ABA clinics, researchers face the same problem. When a concept has many operational definitions, results become hard to compare.

A 2026 mouse study on behavioral persistence makes this point directly (Cao et al., 2026). Persistence means keeping up goal-directed action despite obstacles. The authors note that persistence research is split across fields with different operational definitions. To find persistence periods in their data, they built a data-driven model instead of using arbitrary frequency cutoffs.

A 2022 study raises the same concern for immediacy, which is how quickly behavior changes after an intervention starts. The authors found many different conceptual and operational definitions in use. Their suggestion is to test several operational definitions at once. If the results agree across definitions, you can trust the immediate effect more.

Both studies echo the speakers above. A definition is not a small detail. It decides what your data can and cannot tell you.

FAQ#

What is an operational definition in ABA? It is a description of a behavior in terms of what a person actually does. It is clear and complete enough that two independent observers would score the same response. Sean Yocum ties this standard to Hawkins and colleagues in 1977.

What makes an operational definition good? Mike Carrero teaches the Triple Cs: clear, concise, and complete. A good definition covers what the behavior looks like, what it sounds like, and its timing. Most of all, the people using it must agree on what counts.

Do operational definitions only apply to client behavior? No. Mellanie Page uses them to set clear expectations for staff, with examples and non-examples. Mackenzie Sandler uses them to check bias in her own reactions. Dr. Jessica Osos builds them into every treatment plan goal.

Turn this topic into a CEU

You just studied this. Now get credit for it.

Watch Dunder Mifflin’s Guide to Training and Onboarding: Lessons from The Office with Mellanie Page and earn 1 free BCBA CEU. Detailed certificate, delivered the moment you finish.

Watch and earn the CEU →Free account · No card · detailed cert

Keep going with a free learning path

Supervision, Done Right

Four supervision CEUs that build on each other. What reflective supervision is, how to build a supervision system, how to assess supervisee competence instead of counting hours, and how supervision affects provider outcomes. Each course is free and issues its own certificate.

Start the path →4 courses · 4 supervision CEUs · Watch at your own pace