A few years ago I said something at a state conference that some university professors weren't thrilled about:
Just because a behavior plan is technically sound doesn't mean it will have the outcome you expect.
A plan can pass every rubric in a graduate course and still fail, because the people who have to carry it out aren't comfortable, confident, or able to do it the way it's written.
I stand by that. And when a plan fails, our first instinct is usually to look at the person implementing it, not at the plan.
The old question: was it implemented as designed?
Traditional procedural fidelity asks one question: was the plan implemented as designed? It's an important question. I don't think it's the whole question.
I think of treatment integrity as fidelity plus fit. Fit asks: could this plan realistically be implemented by this person in this setting?
One useful definition describes contextual fit as the extent to which a plan's procedures are consistent with the knowledge, values, skills, resources, and administrative support of the people expected to implement it.
Five pieces, and any one can sink a plan
Knowledge. Do they know how to do the procedure and why it works?
Values. Does it fit what they believe is appropriate and likely to work? If it goes against their core beliefs, they can have every skill and resource and still not do it.
Skills. Can they carry it out in real time, with real demands on them?
Resources. Do they have the time, space, materials, and staffing?
Administrative support. Is the organization set up so that doing it is worth doing?
We don't measure fidelity as often as we think
While preparing this topic I fell down a research rabbit hole, and I'd like to be corrected if I'm wrong. In one review of single-case studies published roughly between 2017 and 2021, only about half measured procedural fidelity at all, and fewer than one in five of those checked whether the measurement itself was reliable.
The reasons people give are understandable: early behavior analysis used equipment that either worked or didn't, journals haven't required fidelity data, we assume trained staff do it right, and measuring it well takes a second observer.
If we are not consistently measuring whether plans are implemented in research, it's no surprise we have an uneven picture in practice.
Technical adequacy and contextual fit pull in different directions
Research from school-based behavior support shows a trade-off:
When a behavior specialist designs a plan alone, it tends to score highest on technical adequacy (what a university rubric rewards) and lower on contextual fit.
When a team designs a plan without a specialist, contextual fit is higher but technical adequacy often drops by a lot.
When the team and the specialist build the plan together, you get closer to both.
A plan that is technically sound and a plan that is practically implementable are not automatically the same plan.
Fit isn't lowering the bar
This is the point I most want clinicians to hear. Contextual fit does not replace effectiveness. A feasible plan that is technically weak fails just as surely as a strong plan nobody can carry out. Both can basically be thrown in the trash, in my opinion.
Fit is also measurable, and it's already part of our ethical standards. Our ethics code asks us to select interventions that best meet the diverse needs, context, and resources of clients and stakeholders, alongside scientific evidence, not instead of it.
Two resources worth a look: Horner and colleagues developed a validated 16-item self-assessment of contextual fit, and Wolf's classic work on social validity asks whether goals are socially significant, procedures are socially appropriate, and outcomes matter in ways people actually notice.
What to do differently
Add a fit check before you finalize any plan. Look at the environment, the implementer, and the organization. Do it while writing the plan, not after it fails.
Ask what the implementer would accept. In the research I share in my trainings, paraprofessionals who rated a plan as more acceptable implemented it far more often, and challenging behavior dropped accordingly.
Look at the system before the person. Personnel issues are real. I would address them only after ruling out a contextual fit problem.
Measure fidelity with feasible tools. If the check is too heavy to repeat, it won't be done.
This doesn't mean everything is on you. If you've built a plan that fits and someone still chooses not to follow it, that isn't a failure on your part. But if you're doing 100% of the work, that's where burnout starts.
If you remember one thing: treat integrity as fidelity plus fit, and check the fit before you blame the person.
Want help building plans and systems that hold up in real homes, schools, and clinics? Learn about Clinical and Agency Consulting or try the free fit check in Clinical Tools. I offer a free 15-minute consultation if you want to talk it through.
Casey Gressmen, MBA, MEd, BCBA, LBA, is a licensed special education teacher and board certified behavior analyst and the founder of Phoenix Rising Behavior and Academic Consulting.
