Built by operators who ran ABA clinics.

Proof the plan ran the way it was written.

Build fidelity checklists from your own protocol, score them item by item against real sessions, and get the result back to the technician — so Wilma® turns treatment integrity from an assumption into something you can show.

not a generic checklist
Your protocol
scored, not pass/fail
Per item
with the person observed
Shared
per staff member over time
Trended

What is treatment fidelity software?

Treatment integrity, recorded as the session runs.

Treatment fidelity software — treatment integrity software, if that is your team’s term — records whether an intervention was delivered the way the plan specified. It is the question sitting underneath every outcome graph: this data was produced by something, and fidelity is how you establish that the something was the programme you wrote.

The failure mode is a checkbox. A single tick on a session note satisfies the requirement and tells nobody which component was missed, never reaches the technician, and cannot be aggregated into a training need. In Wilma the checklist is built from your protocol, scored item by item against the session it describes, released to the staff member it is about, and rolled up so the item that keeps scoring low becomes the thing you coach.

Fidelity is half of the data-quality question. It establishes that the plan was delivered as written; interobserver agreement establishes that the measurement was reliable. Both belong on the same record as the supervision trail they were gathered alongside.

Build the checklist once. Run it against real sessions.

Treatment fidelity — treatment integrity, if that is the term your team uses — is only measurable against a written statement of what should have happened. Build that statement as a template, with the items and categories your protocol actually contains, and run it against sessions as they happen.

Templates you define

The items reflect your protocol, not a vendor’s idea of one.

Grouped by category

Items sit in the categories your team already thinks in.

Reused across staff

One template, applied consistently to everyone delivering it.

Scored per item

Not a single pass/fail that hides which step was missed.

Filled in against the session, not from memory on Friday.

A fidelity checklist completed three days later measures recall. The observer completes it against the session it belongs to, so what you end up holding is a record of the delivery rather than a reconstruction of it.

Attached to the session

The checklist belongs to a session and a staff member.

Narrative alongside the score

Free-text notes where a number does not carry the point.

In person or over video

A remote observation is still an observation.

Sits with the data

On the same record as the session data and the supervision trail.

The observation reaches the person it is about.

Fidelity data that never gets back to the technician changes nothing about the next session — it just documents that the drift was noticed. Completed checklists are shared with the staff member deliberately, and whether that happened is itself tracked.

Shared on purpose

Finished checklists are released to the staff member, not left sitting.

Acknowledgement tracked

Whether it was seen, answered, or declined is on the record.

Nothing shared half-finished

A checklist has to be complete before it can be shown.

Outstanding work is visible

Shown-but-not-filed and awaiting-answer are states you can chase.

One weak session is noise. The same weak item is a training need.

A single observation tells you about one afternoon. The value is in what repeats — and the item that scores lowest across a technician’s last dozen observations is a far better coaching target than whatever went wrong most recently.

Scored per item over time

Each checklist item carries its own score across a date range.

Weakest items surfaced

The lowest-scoring items come to the top, where coaching should start.

Per staff member

A rollup for the person, across every observation of them.

Notes carried through

The narrative from each observation stays attached to the item.

Proof the plan was delivered, not just written.

An authorization review asks two questions about your data. Was it measured reliably, and was the programme actually run as written? Fidelity answers the second. Without it, a treatment plan is a description of intent and the outcome data has nothing to attribute itself to.

On the record, at the time

Captured against the session rather than assembled for the review.

Per-item detail

Show which components were delivered, not an overall impression.

Pairs with IOA

Fidelity covers the delivery; IOA covers the measurement.

Defensible trail

Who observed, what they saw, and whether it reached the technician.

Buyer's Checklist

Evaluating treatment fidelity software? Demand every line.

FAQ

Questions clinical directors ask about treatment fidelity.

What is treatment fidelity software?

Is treatment fidelity the same as treatment integrity?

How is this different from IOA?

Can we use our own checklist?

Does the technician see the result?

Can I see whether someone is improving?

Why do funders care about fidelity?

Where does it live?

How much does it cost?

Is Wilma HIPAA compliant?

Customer Story · Clinical

Why a clinical team left "click-after-click" software behind

How fewer clicks for notes, data, and supervision gave a BCBA their day back.

See a fidelity checklist run end to end.

Thirty minutes. Bring one of your protocols — we’ll build it as a checklist, score it against a session, and show you what the technician receives afterwards.

Choosing a fidelity tool

How to evaluate treatment fidelity software

Treatment fidelity is easy to implement badly, because a checkbox on a session note technically satisfies the requirement. What separates a fidelity system from a fidelity gesture is whether anything happens after the box is ticked.

  • 1Is the checklist yours?A fixed vendor checklist measures adherence to a generic model of ABA delivery. Your protocols specify things that list does not contain, and omit things it insists on, so the score drifts away from what your clinical leadership actually cares about — and the staff being scored notice immediately.Red flag: A fixed set of fidelity items with no way to add your own, or "customisation" that means renaming existing ones.
  • 2Per item, or one overall score?A single fidelity percentage tells you a session was 80% right and nothing about which fifth was wrong. Coaching needs the item. An aggregate that cannot be decomposed is a number for a report rather than an input to supervision.Red flag: One fidelity field per session, with the detail living in a free-text note if anywhere.
  • 3Does the feedback reach the technician?This is the step most implementations skip, and it is the only one that changes a future session. If the completed observation lives in a supervisor’s report, the person whose delivery it describes may never see it, and the drift it identified continues on schedule.Red flag: No mechanism to release an observation to the staff member, and no record of whether they saw it.
  • 4Can you see what is outstanding?Observations that were completed but never shared, or shared but never acknowledged, are the ones that quietly accumulate. A system that tracks only completed-and-closed work will report healthy fidelity coverage while a queue of unreleased observations sits behind it.Red flag: A completion count with no visibility of what was shown, awaiting an answer, or declined.
  • 5Does it aggregate across observations?One observation is an anecdote. The training need is the item that scores low repeatedly, across sessions and sometimes across staff — and if a pattern spans several technicians it is usually a training gap rather than an individual performance problem.Red flag: Observations you can only read one at a time, with no per-item view over a date range.
  • 6Does it sit with the data it explains?Fidelity exists to qualify outcome data. Kept in a separate system it becomes a parallel record nobody consults at the moment the outcome data is challenged, which is the only moment it was ever needed.Red flag: Fidelity in a separate quality-assurance tool with no link to the session, the data, or supervision.

Keep reading

Related guides on running an ABA practice — the same operation, from a different angle.