Procedural Fidelity in ABA: What It Is and Why It Matters
Procedural fidelity is how closely an ABA plan is carried out as written. Learn why it matters, how to measure it, and where experts disagree.
Key takeaway
Procedural fidelity is how closely a treatment plan is carried out the way it was written. You may also hear it called procedural integrity or treatment fidelity.

Supervision Articles Deep Dive
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.
Experts who cover this
21 recordings on openceu cover procedural fidelity.
- Featured talk3.5 Supervision CEUs
Supervision Articles Deep Dive
Matt HarringtonThe Behaviorist Bookclub
Across two full episodes, Harrington argues procedural fidelity is universally agreed to matter but rarely tracked with real data: a survey of 140 BCBAs found…
- 1 CEUs (0.5 Learning, 0.5 Supervision)
No Man's Land: Is Your Data Real? Clock Time Without IOA and Procedural Integrity Is Unfalsifiable
Sean YocumHickory Learning Group
Procedural integrity (fidelity of IV delivery) is framed as the other non-negotiable pillar; without it, clinical outcomes become scientifically unfalsifiable…
- 1 Supervision CEU
Treatment Integrity Is Not a Checkbox: Building Meaningful Fidelity Systems
Dr. Jessica OsosCopper Consulting Group
The central topic of the talk: procedural fidelity is defined as the extent to which an intervention is implemented as designed, and the talk builds the…
- 0.5 Learning CEUs
Rad N Bad: Your Error Correction Sucks — Error-Correction Assessments, Trials to Mastery, and Learner-Informed Procedures
Sean Yocum, LBA & Michael CarreroHickory Learning Group
They highlight that error correction procedures are rarely checked for fidelity across staff, proposing an in-vivo spot-check (asking multiple staff how they…
- 1 Supervision CEU
Reflective Supervision and Client Outcomes
B. Kuereine GrayThe Behaviorist Bookclub
Cites Snowden, Liggett & Taylor (2017), finding reflective supervision improved quality and fidelity of intervention and adherence to evidence-based practice…
- 1 Supervision CEU
Teaching RBTs to Think Clinically: Training In-the-Moment Decision Making
Dr. Jessica Osos & Tauren KeelsCopper Consulting Group
Discusses how overly rigid checklist-based fidelity training breaks down because treatment plans can't anticipate every scenario; it limits generalization, can…
- 1 PDU
Mastering Your Time: Preparing for Service Delivery
Ashley AndrewInnovative Behavioral Therapy
Andrew emphasizes implementing protocols consistently and as written, while remaining flexible when sessions aren't going well, and communicating barriers to a…
- 1 PDU
Mastering Your Time: Preparing for Service Delivery
Ashley AndrewInnovative Behavioral Therapy
Andrew ties effective time management to implementing protocols consistently and as written while maintaining treatment integrity even when adapting in the…
- 1 Supervision CEU
Providing Fieldwork Supervision in a School Setting
Dr. Allison GohringRooted Consultation, LLC
Fidelity/implementation checks (including self-checks) are listed as core unrestricted supervisory activities, used to assess intervention effectiveness and…
- 1 Learning CEU
Quality Without Measurement is Just Marketing: A Critical Analysis of Quality Service Delivery
Kristen ByraUpskill ABA
Byra frames fidelity monitoring (regular fidelity checks on goal implementation, BST for new goals with every tech who works with a client) as one of five core…
- 2 Learning CEUs
Rad N Bad: Dosage by Tradition: What the Research Actually Says About ABA Hours, and the Contingencies Behind Them
Sean Yocum & Michael CarreroHickory Learning Group
The hosts identify treatment fidelity and social validity, alongside learning opportunities and reinforcement quality, as variables that actually determine…
- 1 Learning CEU
New Year, New Care Collab Goals
Matt HarringtonThe Behaviorist Bookclub
Recommended as the key data BCBAs can bring to OT partnerships: tracking how consistently caregivers implement OT homework (not just whether they did it) to…
- 1 Ethics CEU
Article Club: Cultural Awareness for Behavior Analysts
Pauli Catanese & Jasmine AbdelAchieve Behavior Partners
Discussed as compromised by poor cultural fit: when caregivers disagree with or deprioritize a plan, implementation fidelity drops and skills fail to…
- 0.5 Learning CEUs
Rad N Bad: 30–40 Hours of ABA — Dosage, Titration, and the Data Behind Service Intensity
Sean Yocum, LBA & Michael CarreroHickory Learning Group
The hosts challenge providers to prove that interventions are actually implemented as designed, arguing that fidelity is often claimed but rarely measured or…
- 1 Learning CEU
Rad N Bad: Parent Buy-In as a Clinical Variable — Caregiver Training, Generalization, and Least-Restrictive Service Design
Sean Yocum, LBA & Michael CarreroHickory Learning Group
The hosts insist that parents must be trained to a measured competency criterion, not just shown a skill once, using data checks (e.g., via the HiRasmus…
- 0.5 Learning CEUs
Rad N Bad: Your Prompting Hierarchy Sucks — Individualized Prompt Selection, Written Fading Rules, and Preventing Prompt Dependency
Sean Yocum, LBA & Michael CarreroHickory Learning Group
Mike identifies poor procedural integrity and fidelity of implementation as a key reason prompting hierarchies fail on the floor, since staff often default to…
- 1 Learning CEU
From Idea to IRB: Conducting Research in Your ABA Practice
Lynette WyckoffUpskill ABA
Patrick notes fidelity as one of three things to track during a study (alongside deviations and adverse events), explaining that a high-fidelity automated…
- 1 Supervision CEU
Toward an Evidence-based Supervision Model: Training Students in Independent Fieldwork
Dr. Allison Bowhers-PeutinWillow Mentorship & Consulting, LLC
Treatment fidelity of the supervisor trainings themselves was checked in a sample of sessions in both experiments, with high agreement (96% and 100%…
- 1 Supervision CEU
Building a Supervision System That Actually Works
Lauren DonovanThe Behaviorist Bookclub
Discussed briefly as a built-in component of the therapist observation form and something a supervisor checks for during direct observation.
- 1 Supervision CEU
Dunder Mifflin’s Guide to BCBA Supervision: Lessons from The Office
Mellanie PageThe ABA Collective, LLC.
Named as a core, measurable thing supervisors should track and give regular feedback on, tied to monthly integrity checks comparing how closely an RBT follows…
- 1 Learning CEU
Research to practice - extending past the pages
Matt HarringtonThe Behaviorist Bookclub
Named as a second parsimonious check before assuming a deeper clinical issue is present, and cited as a reason data or graphs may look poor even with a sound…
On this page · 15 sections▾
Procedural fidelity is how closely a treatment plan is carried out the way it was written. You may also hear it called procedural integrity or treatment fidelity. If a plan says to give a prompt on every trial, fidelity data show how often that prompt really happened.
This matters because a plan only works if people follow it. When fidelity is low, you cannot tell if the plan failed or if it was never really tried. Clinicians across many talks agree that fidelity protects both the client and the BCBA's own decisions. They do not always agree on how strictly staff should stick to the written steps.
What procedural fidelity means#
Dr. Jessica Osos gives the plainest definition. She notes that most people hear the term in their coursework and fieldwork. Yet it often gets lost once they start real practice. She traces the idea back to Gresham in 1989, and says Vollmer and colleagues carried it forward.
at its simplest, it's the extent to which an intervention is implemented as it was designed.
From the talk — Dr. Jessica Osos
In plain words, the plan is the recipe. Fidelity tells you how well the cook followed it. Osos uses a simple example. If your plan says to deliver a prompt on every trial, fidelity data tell you how often that prompt actually showed up.
The words "fidelity," "integrity," and "adherence" get used as if they mean the same thing. Most behavior analysts treat them that way. Still, some small differences exist between the terms. One editorial in the research section below suggests the field may want to settle on "procedural fidelity" as the main label. For this page, all three names point to the same core idea. Did the people running the plan do what the plan said?
Why fidelity protects your conclusions#
Osos gives three reasons fidelity matters. The first is internal validity. That is a research term for being able to say "this thing caused that change." When a client's behavior shifts, fidelity data let you claim the plan did it. Without that data, the change could come from anything.
Sean Yocum makes the same point with sharper words. He points to classic papers on the "technological" and "analytic" parts of ABA. Technological means the steps are fully described. Analytic means you can show a believable link between what you did and what changed.
When you fail to measure treatment integrity, clinical outcomes become scientifically unfalsifiable.
From the talk — Sean Yocum
"Unfalsifiable" means no one can prove it right or wrong. Yocum explains what this looks like on a flat graph. After six months with no fidelity data, you cannot tell a treatment failure from an implementation failure. A treatment failure means the plan does not work. An implementation failure means the plan was never delivered as written. He says the field often takes credit when kids improve. When kids plateau, it blames the diagnosis or asks for more hours.
Osos sums up her whole talk in one line: procedural fidelity is not a compliance checkbox.
It's how we protect our clients and our own conclusions and our own services.
From the talk — Dr. Jessica Osos
The gap between knowing and doing#
Matt Harrington digs into this topic in his supervision article series. His main point is simple. Everyone agrees fidelity matters, but very few people track it with real data. He points out that research journals now treat fidelity data as a must. In some journals, you cannot publish without it.
if you don't have procedural integrity data saying that this data corresponded to what you wrote, then how do we know that what you wrote was the actual thing that was done?
From the talk — Matt Harrington
Clinical work has no such rule. Harrington says he has never seen an insurance company require fidelity data. He has rarely seen a business or clinical director ask for it either. He reviewed a survey of practicing BCBAs. About 94% watched staff often, since weekly or monthly observation is part of billing. Feedback was common too.
90% provided feedback consistently. Almost all of it was in-the-moment vocal feedback. But only 20% used graphs to track their feedback.
From the talk — Matt Harrington
Harrington is honest that he works this way too. You see an error, you correct it, and you move on. He calls this the most effective way to change behavior, because it needs no prep work. The problem is that nothing gets recorded. He even guesses that the average BCBA's plan is followed correctly about 60% of the time at most.
Sean Yocum saw the same gap. He asked about 10 BCBAs at five companies if they tracked procedural integrity or IOA. IOA stands for interobserver agreement, which checks if two people record the same data. He admits his poll was informal. The answer was still the same across the board: no.
When the graph is flat, check fidelity first#
Many speakers say fidelity should be one of the first things you check when a plan stalls. Matthew Harrington lists three obvious reasons a plan is not working. First, it may not be function based, so it solves the wrong problem. Second, it may not carry over between home, school, and clinic. Third, fidelity may be missing.
The fidelity of the intervention is almost always the primary reason why interventions aren't consistent with the way we currently run ABA with our three tier model.
From the talk — Matthew Harrington
The "three tier model" is the common setup of a BCBA, an RBT, and a family. Each layer is a new chance for the plan to drift. In another talk, Harrington ties this to parsimony. Parsimony means checking the simple explanations before the complex ones. A supervisor may only see 20% to 30% of sessions. So you have to ask what is happening in the sessions you never see. He says no published intervention will work, and your graph will never look good, if there is no fidelity.
Michael Carrero frames fidelity data as a roadmap. It shows where you are and what to fix next.
Well, if you don't have procedural integrity, the first thing that you're going to do is X, right? And you're going to address procedural integrity.
From the talk — Mike Carrero
Carrero says this habit moves teams away from "winging it." It lets them look at programs through an objective lens. They can see who is doing their job and how well.
Not all errors are equal#
Fidelity is not just one number. The type of mistake matters too. Yocum reviews research on DRA, which stands for differential reinforcement of alternative behavior. In DRA, you reward a better behavior and stop rewarding the problem behavior.
there's like errors of commission where we're accidentally reinforcing problem behavior even part of the time. It degrades outcomes far worse than errors of omission and systematically break the functional relation.
From the talk — Sean Yocum
An error of commission means doing something you should not do, like rewarding the problem behavior. An error of omission means leaving out a step. Yocum says the first kind does far more damage. He also cites work showing that errors during discrete trial teaching can slow down or stall skill learning.
Error correction is a common place for drift. Carrero says no one reports fidelity data on error correction. He warns that staff may run the same written procedure in very different ways. Yocum offers a quick test. Pick one program and ask three staff, one at a time, what they do when the learner gets it wrong.
if you get three different answers, you don't have an error correction procedure, you have three
From the talk — Sean Yocum
His fix is to write the procedure into one clear sentence. Then he teaches it with BST at the next team meeting. BST stands for behavioral skills training. It means you explain, model, practice, and give feedback. Carrero sees the same risk with prompting. Staff tend to grab the prompt that works in the moment instead of following the written system. That habit can feed prompt dependency.
Low fidelity can cause harm#
Dr. Shane Spiker raises the stakes past slow progress. In his talk on crisis management, he says treatment is where clinicians often go wrong. Sometimes assessment misses key factors. Other times the plan is fine but people do not follow it well. Staff, caregivers, and other stakeholders then end up creating more problems.
people with poor treatment fidelity will re-traumatize. They will cause problems or your treatment itself could re-traumatize.
From the talk — Dr. Shane Spiker
Spiker says poor fidelity can trigger a person or bring on problem behavior that never needed to happen. This view changes how fidelity should feel to a supervisor. It is not only a data quality issue. It is a safety issue for the person being served.
This link between fidelity and harm fits Yocum's point about commission errors. When someone rewards problem behavior even part of the time, the plan can make things worse. A crisis may then look like a "hard case" when it is really a fidelity case. Spiker's message is that teams should look at their own actions before blaming the client.
How to measure it: direct observation and checklists#
Osos gives a clear method. She lists time, systems, and expectations as the biggest barriers. She notes that these are organization problems, which means they can be fixed.
Direct observation with component level checklists is going to be the recommended approach.
From the talk — Dr. Jessica Osos
A component level checklist breaks a procedure into small steps. The observer marks each step as done right or wrong. Osos also shares the checklist with new RBTs from the start of training. That way they know what is expected, and the first check is not a surprise. She points to ready-made templates from Morris and colleagues in 2024.
Other supervisors build fidelity into their own tools. Lauren Donovan uses a therapist observation form that has procedural integrity and IOA built in. She looks at a few graphs before she walks in. Then she watches the programs that seem to be struggling and checks for integrity.
Mellanie Page lists treatment fidelity among the clear things a supervisor should track and give feedback on.
Monthly integrity checks on learner plans and give them feedback on how closely they're following the plan.
From the talk — Mellanie Page
Page likes giving staff the rubric itself. Then they know exactly what she is judging. Dr. Allison Gohring lists fidelity checks among the "unrestricted" supervision activities in school fieldwork. These are tasks like assessing, measuring, and analyzing, rather than running sessions directly.
Building fidelity into supervision systems#
Kristen Byra, Ph.D., treats fidelity checks as a core part of service quality. She asks a pointed question of agencies that promise quality on their websites. Are they actually doing fidelity checks with every technician on every client? Her standard is a fidelity check on skill goals at least every two weeks.
New goals are introduced using BST with all the BTs and RBTs that work with a client.
From the talk — Kristen Byra, Ph.D., BCBA-D, LBA
Byra says most clinics fall short here. She blames the "musical chair" style of scheduling. One tech calls out sick, so another swaps in. Some kids end up with 15 RBTs. No BCBA can do BST with all of them. In her view, fidelity is a staffing problem as much as a training problem.
B. Kuereine Gray adds another angle. Her talk centers on how reflective supervision affects treatment fidelity, service quality, and implementation. Reflective supervision is a style where the supervisor and staff talk through the work together, not just check boxes. Gray treats fidelity as one of the outcomes good supervision should move.
Research shows fidelity checks also apply to training supervisors. Dr. Allison Bowhers-Peutin reported fidelity checks on the BST workshops in her own study. In one experiment, fidelity was checked in 33% of trainings with 96% agreement. In the next, it was checked in 50% with 100% agreement.
Fidelity at home with caregivers#
Fidelity is not only for paid staff. Caregivers run much of a child's plan. Yocum says his team trains parents to a set level of skill, not just a single show and tell. A skill is not mastered until the parent meets that level. RBTs even run data checks on the parents.
You got to maintain high treatment fidelity.
From the talk — Sean Yocum
Yocum argues that parent-led models need the same or even higher standards than staff models. They also need more supervision and real family buy-in.
Matt Harrington applies the same idea to work with occupational therapists. When an OT assigns homework, he wants to know how often the family does it and how well.
If we're doing something 50% of the time and there's no impact on behavior, that makes sense. If we're doing something 100% of the time and there's no impact, then we might have an intervention mismatch.
From the talk — Matt Harrington
His top tip is to help caregivers self-track their fidelity. That way the OT can decide based on data, not on a story.
Jasmine Abdel names one reason home fidelity drops. She lists it among the most common breakdowns in practice.
the low implementation fidelity because of the poor cultural fit
From the talk — Jasmine Abdel
When parents do not agree with a plan, or it is not their priority, they are less likely to carry it out. Skills then fail to carry over.
Fidelity, dosage, and accountability#
Fidelity also comes up in debates about how many hours of ABA a client needs. Michael Carrero and Sean Yocum argue that clock time alone is not the treatment. What happens in those hours is what counts. They list learning opportunities as one key factor. Carrero then asks if treatment fidelity should be on the list too. Yocum agrees and adds social validity, meaning whether the change matters to the person.
Carrero also pushes back on how fidelity gets used in hours requests. He says providers should stop "weaponizing" fidelity to justify very high hours. Some agencies claim only their techs can run a plan correctly. He wants proof.
Show me your procedural integrity and your IOA data.
From the talk — Michael Carrero
He guesses most clinicians in private insurance models could not show this data. He wants providers to report fidelity and IOA in their progress reports. That would show the data are valid and accurate. It would also show the plan is the real reason a child is or is not improving. In his view, a claim of high fidelity without data is just a claim.
The RBT's role in session#
Ashley Andrew speaks to front-line staff directly. Her advice is to run protocols consistently and as written. If you follow the plan and the data improve, that is great. If you follow the plan and the data get worse, your supervisor needs to know you did it as written.
Follow those established protocols and communicate any barriers or concerns to your supervisor.
From the talk — Ashley Andrew
Andrew does not ask RBTs to act like robots. She tells them to stay flexible when a session goes badly. Swap in new toys. Go back to getting the learner happy, relaxed, and engaged. She warns that staff who are too black and white often see more problem behavior. The key is to adapt while still keeping treatment integrity.
When the same thing keeps failing, she says to stop repeating it. Ask the BCBA to model the protocol. Then you can tweak or change it under their direction. In her view, good fidelity includes honest reporting back to the team, not just quiet compliance.
Fidelity in practice-based research#
Patrick Baga discusses fidelity in research done inside an ABA practice. He lists it as one of three things to track during a study, next to deviations and adverse events.
Baga makes a simple point. You will always have a plan, and then reality will happen. You need some process to track the gap between the two. Many studies use an IOA step or something like it to confirm that what was promised is really happening.
He also notes an exception. His team used an automated system that does things exactly as prescribed. Because of that, they did not need to build in separate fidelity checks. This shows that the need for fidelity checks depends on who or what delivers the plan. Human delivery invites drift. Automated delivery may not. Still, most ABA plans are run by people, so most teams will need some form of fidelity data.
Where experts disagree: checklists or clinical judgment#
The speakers agree fidelity matters. They split on how tightly staff should be held to the written steps.
Osos champions component checklists as the best way to measure fidelity. Yocum and Carrero want every procedure written as one clear version and taught the same way to all staff. In this view, drift is the enemy and the written plan is the standard.
The presenters of a talk on teaching RBTs to think clinically see a risk in that approach. They ask why checklist trainings break down. Their answer is that no plan can predict everything a client will do. Plans are written for a perfect world with stable conditions. When training focuses only on following the written program, it limits how well a tech's skills generalize. A tech may freeze when something new happens. The presenters also want techs to reinforce good behavior as it occurs, even when it is not listed in the treatment plan.
They give an example. A tech named Jordan has a checklist for naturalistic teaching. It says to follow the learner's lead and reward appropriate responses. Then the learner disengages. The checklist does not help Jordan tell if this is a motivation issue, a skill gap, or something in the setting.
Ashley Andrew sits between these views. She tells staff to stay flexible but keep treatment integrity, and to report problems instead of improvising alone. The tension is real. One side guards against drift. The other guards against rigid staff who cannot respond to the child in front of them.
What the research says#
Recent articles back up many points the speakers make.
An editorial introducing a special issue on procedural fidelity notes that the three terms (fidelity, integrity, and adherence) are mostly used as synonyms. It suggests there may be merit to adopting "procedural fidelity" as the shared term (St. Peter, 2023).
One study found that fidelity may drop a lot when no supervisor is watching. Four staff used video self-monitoring, meaning they watched and scored their own session videos, and their fidelity went up and stayed high (Paden & Carroll, 2024).
Another study found that errors in delivering consequences lowered correct responding for about half of participants, even after they had mastered the task. Reinforcing errors was the more harmful error type (Jones et al., 2026).
A fourth study warns that fidelity and data errors could lead teams to change treatment too soon. The authors urge regular checks of data accuracy as part of procedural integrity (Halbur et al., 2024).
FAQ#
What is the difference between procedural fidelity and treatment integrity? In most behavior analysis talks, the two terms mean the same thing. Both describe how closely a plan is carried out as written. You may also hear "treatment fidelity" or "treatment adherence." Some writers see small differences between the terms. In daily practice, though, clinicians use them as one idea. The key question is always the same: did people follow the plan?
How do you measure procedural fidelity in ABA? Dr. Jessica Osos recommends direct observation with a component level checklist. You break the procedure into steps and mark each one as right or wrong. Then you divide correct steps by total steps to get a percentage. Supervisors like Mellanie Page run these checks monthly. Kristen Byra suggests checking skill goals at least every two weeks.
Why does procedural fidelity matter if the client is making progress? Without fidelity data, you cannot say the plan caused the progress. Something else, like time or maturity, could explain it. Fidelity data also help when progress stops. They tell you if the plan failed or if it was never run as written. That changes what you fix next.
To see a full system for building checks your team will actually use, watch Treatment Integrity Is Not a Checkbox: Building Meaningful Fidelity Systems.
Turn this topic into a CEU
You just studied this. Now get credit for it.
Watch Supervision Articles Deep Dive with Matt Harrington and earn 3.5 free BCBA CEUs. Detailed certificate, delivered the moment you finish.
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.









