Competency-based supervision

Competencies must be observable

“Understands documentation” is hard to assess. “Identifies a late entry, preserves original content, records the current date and author, and routes claim impact to billing” names observable actions and boundaries.

Each competency should state conditions, materials, response, quality, time, allowed support, critical errors, and criterion. Define whether one critical safety error overrides an aggregate percentage.

Start with baseline evidence

Observe performance before prescribing training when safe. Baseline can show which steps are fluent, missing, or inaccessible. It also prevents teaching time from being spent on skills already demonstrated.

Use simulation when direct assessment could risk a client, breach privacy, or exceed the supervisee’s role. A supervisor may assess reasoning with fictional records before any live decision.

Train and reassess

Behavioral skills training can combine instructions, modeling, rehearsal, and feedback. Reassess under the same defined conditions, then vary examples and settings to test transfer.

Turner, Fischer, and Luiselli present a competency-based, ethical, and socially valid supervision model with bidirectional feedback. The paper proposed a practice model while noting limited data on supervision practices and trainee outcomes.

A fictional competency record

Mika practices routing nine fictional incident scenarios. The checklist covers immediate safety, clinical escalation, privacy review, payer notice, documentation, and role boundaries.

At baseline, Mika completes 7 of 9 scenarios without a critical error. After targeted modeling and rehearsal, Mika completes 9 of 9. These data show performance on the defined scenarios. They do not prove workplace generalization, client safety, or authorization to make every routed decision.

The supervisor schedules two novel probes and reviews any real incident only within appropriate authority and privacy safeguards.

Mastery criteria need defensible units

Percentage alone can hide omitted conditions. Report skills demonstrated divided by eligible skills, plus critical errors, prompts, opportunities, scenario types, and observer agreement when scores drive consequential decisions.

Sellers and colleagues recommend structured content and predetermined mastery criteria in individual supervision. Their guideline is a professional recommendation rather than a BACB handbook requirement.

Transfer and maintenance remain open questions

Role-play mastery may fail in a busy clinic, remote session, unfamiliar payer workflow, or emotionally difficult conversation. Probe representative conditions with ordinary tools. Check again after time passes or the procedure changes.

When transfer is weak, inspect environmental barriers: unavailable records, unclear authority, excessive workload, conflicting procedures, or inaccessible systems. Retraining cannot solve every system problem.

Current BACB sources define separate roles

The BACB supervision and training page distinguishes supervising ongoing RBT services, RBT assessment and training, BCaBA supervision, and supervised fieldwork. It links role-specific handbooks, packets, and curriculum.

The current Supervisor Training Curriculum Outline 2.0 covers ongoing services and trainee supervision. Curriculum scope does not replace the governing handbook, law, payer contract, or employer responsibility.

Supervision is bidirectional

The supervisee needs a safe way to report unclear feedback, conflicts, workload, access needs, boundary concerns, and mistakes. Supervisors should evaluate their own clarity, availability, competence, and response to feedback.

Document the next action and another performance opportunity. A signed form without observation or follow-up is weak competency evidence.

Use a competency map instead of a completion list

A competency map connects each important duty to its evidence. For a clinician learning to lead caregiver coaching, the map might include preparing accessible materials, explaining the purpose in plain language, modeling the step, inviting rehearsal, giving specific feedback, checking caregiver priorities, and documenting the agreed next action. Each element receives an operational definition, assessment method, mastery rule, assessor, and reassessment date.

The evidence should fit the skill. Knowledge can be sampled with explanation or scenario questions. Interviewing, observation, feedback, data interpretation, and safety responses require demonstration. A single score can hide a critical weakness, so retain component results when one step carries special risk.

Keep these states separate:

• Assigned: the duty appears in the person’s role.
• Trained: planned instruction and practice occurred.
• Demonstrated: direct evidence met the stated criterion.
• Generalized: the skill appeared under a relevant changed condition.
• Maintained: performance met the rule after the planned interval.
• Authorized: current law, credential, organization, payer, and case requirements permit the activity.

These labels prevent a course certificate from becoming a proxy for practice authority. They also make remediation more precise. When one component misses criterion, the supervisor can adjust instruction, opportunities, tools, workload, or environmental support and then sample that component again. Record who evaluated it and which version of the rubric was used.