Onboarding, Assessing Supervisee Skills, and Selecting Supervision Goals
Key Takeaways
- Supervision goals should be derived from assessment of the supervisee's actual repertoire, not from supervisor convenience or supervisee preference alone.
- Useful assessment sources include direct observation, permanent products, role-play, interview/self-rating, and client/staff outcome data; self-report alone is insufficient when client welfare is at stake.
- Strong supervision goals are observable, measurable, socially significant, criterion-referenced, and matched to risk, setting, and current competence.
- Goal selection prioritizes safety and ethics first, builds prerequisites before complex independent decisions, and front-loads unrestricted activities to avoid last-minute hour problems.
- A fieldwork trap is approving broad activities without verifying the supervisee can perform the underlying component skills.
Onboarding and Initial Skill Assessment
Strong supervision starts with onboarding the supervisee to the setting, expectations, and systems, and then assessing the supervisee's current skills. Treat supervisee skill assessment exactly like any behavior-analytic assessment: define the relevant performances, gather baseline data, and select goals based on risk, social significance, and context.
The assessment answers a simple question with serious implications: what can this person already do reliably, and what must be built before they handle it independently? Skipping this step is the supervisory equivalent of writing a treatment plan without an assessment. You would be guessing, and a wrong guess in supervision can mean a client receives services from someone not yet ready to deliver them.
Onboarding should also establish the logistics that make assessment possible: how observations will be scheduled, where permanent products live, how feedback is delivered, and how progress is recorded. A supervisee who understands from day one that they will be observed running real procedures, scored against a checklist, behaves differently than one who expects supervision to be a weekly chat. Setting that expectation early prevents the supervisee from experiencing the first direct observation as a surprise "gotcha," and it makes the resulting performance data a fair, anticipated sample rather than an ambush.
Multiple Assessment Sources
No single source tells the whole story. Self-report is useful for goals, barriers, and history, but it should never be the only source when client welfare or fieldwork verification is on the line, supervisees may over- or under-estimate their own skills.
| Source | Best use |
|---|---|
| Direct observation | Treatment integrity, feedback delivery, caregiver coaching, in-vivo skills |
| Permanent product | Session notes, graphs, FBA summaries, written protocols, behavior plans |
| Role-play / rehearsal | Rare, risky, or not-yet-available skills (e.g., crisis response, a procedure no current client needs) |
| Interview / self-rating | Goals, barriers, confidence, learning history, preferences |
| Client and staff outcome data | Whether the supervisee's behavior actually affects client progress |
The gold standard is direct observation of the actual performance, supplemented by permanent products and outcome data. Role-play fills the gap when a skill is too risky or rare to wait for a natural opportunity.
From Broad Competency to Observable Goal
A weak goal says the supervisee will "understand assessment" or "improve clinical skills." Those cannot be observed, scored, or known to be met. A strong goal is operationally defined with a criterion:
"The supervisee will write operational definitions that pass a peer-review checklist at 90% accuracy across three separate cases."
This tells everyone what to observe (written definitions), how to score it (the checklist), and when to advance (90% across three cases). Strong supervision goals share the same properties as good client goals: observable, measurable, socially significant, and criterion-referenced.
The exam frequently presents a vague goal and several measurable alternatives; the keyed answer is the one that specifies the behavior, the measurement, and the mastery criterion. "Understand," "appreciate," "be familiar with," and "know" are red-flag verbs that signal an unmeasurable goal.
Goal Selection Priorities
Not all goals are equal, and the order matters. Sequence supervision goals using these priorities:
- Address safety, ethics, and high-risk implementation first. A supervisee running a restraint or feeding protocol must be competent there before anything optional.
- Build prerequisite skills before complex, independent decision-making. Operational definitions before FBA; integrity on a written plan before designing one.
- Include unrestricted activities early enough to avoid a last-minute fieldwork-hour problem at the end of the experience.
- Choose goals the supervisor is competent to teach and evaluate. You cannot assess what you cannot perform.
- Reassess and re-prioritize when performance data or setting demands change.
This priority list is itself a frequent exam target: when two goals compete, the higher-risk, more ethically loaded performance is taught and verified first.
The Component-Skill Trap
A recurring fieldwork trap is approving a broad activity without verifying the component skills. "The supervisee may now conduct functional analyses independently" sounds like progress, but an FA is a chain of component skills: arranging conditions, ensuring safety, manipulating MOs, collecting and graphing data, and interpreting results. If the supervisee cannot graph or cannot safely run an escape condition, independent FA authorization is premature and risky.
The behavior-analytic move is to task-analyze the broad competency into components, assess each, teach the missing ones with behavioral skills training, and only then expand the supervisee's scope. When an item offers "approve the broad activity" versus "verify components first," the latter protects the client and is the defensible answer.
The same logic governs how you fade your own involvement. As a supervisee demonstrates competence on components, the supervisor systematically reduces prompts and direct oversight, mirroring prompt fading with a client. Removing support too quickly (declaring independence before the data support it) risks the client; removing it too slowly fosters dependence and wastes the supervisee's growth.
The decision to expand scope is therefore a data-based one: scope grows when, and only when, the performance data on each component meet criterion. "They have enough hours now" is never the reason to expand scope; "the integrity data on each component meet criterion" is. This keeps the entire supervisory experience anchored to observable performance from onboarding through sign-off.
A supervisor needs to determine whether a supervisee can correctly implement a discrete-trial teaching protocol with clients. Which assessment source provides the strongest evidence?
Which of the following is the strongest supervision goal?
A supervisee is eager to start designing behavior-reduction plans independently, but observation shows they cannot yet write accurate operational definitions. How should the supervisor sequence goals?
Why is self-report an insufficient sole basis for verifying supervisee competence in high-risk procedures?