Personnel Supervision and Management
Supervising and managing the people who deliver services, one task at a time: the relationship, training, performance problems and supervision data.
- Concepts
- 7
- Sections
- 5
- Diagrams
- 5
In this chapter
Why supervision matters
In most services, the people who deliver sessions day to day are not the people who designed the plan. Supervision is the link between the two, and its quality reaches the client.
Figure 1
How supervision reaches the client
Supervision
What the supervisor does
Observes, measures, gives feedback, practices with the supervisee
Staff performance
A proximal outcome to measure
Procedures carried out as written: higher treatment integrity
Client outcomes
What it is for
Measured progress toward goals the client values
Other possible benefits to assess: skill use across work settings, preparation of future supervisors, satisfaction and retention.
Why behavior-analytic supervision matters
Staff and trainees often implement client plans, so their work affects the services clients receive. Supervision that includes observation, measurement, practice and useful feedback can improve skills and procedural integrity. Client progress must still be measured: accurate implementation of an unsuitable plan will not guarantee benefit. Supportive, useful supervision may also contribute to satisfaction and retention alongside workload, pay, management and other conditions; it is not a sufficient explanation for staff staying or leaving. Evaluate what supervision does and what follows it, rather than treating logged hours as an outcome.
Example
Two clinics record the same supervision hours. One mainly reviews timesheets and answers questions by email. The other also observes sessions, scores implementation and rehearses missed steps with feedback. The second provides more direct opportunities to identify and improve implementation. Compare staff performance and client outcomes over time before concluding which arrangement helps or why.
Common confusion
Supervision is not paperwork kept to satisfy a requirement, and it is not mainly a way to catch and document errors. Its effects are not confined to staff: implementation quality is what the client experiences. More hours do not mean more benefit when those hours contain no observation, feedback or practice.
Key points (4)
- Supervision that observes implementation and gives feedback raises treatment integrity, which supports client progress; supervision is a route to client outcomes, not only a staff matter.
- Supervision builds supervisees' skills and helps them carry those skills into the settings where they work.
- Skill-focused supervision with regular feedback is associated with staff retention; turnover is costly and disrupts services, and pay is not its only cause.
- The benefit comes from what happens in supervision (observation, data, feedback, skill building), not from the hours logged or the paperwork signed; it also develops future supervisors and consistent practice.
The supervisory relationship
A good supervisory relationship is not a matter of personality. It rests on a structure both people can rely on, and on making sure every supervisee gets a fair share of it.
Figure 2
A supervision feedback cycle
Repeats each supervision period
Agree on expectations
Written, observable, with a criterion (contract)
Observe
Watch and score the work itself
Give feedback
Soon, specific: what went well, what to change
Practice
Rehearse the change before the next session
Ask for feedback
How is supervision working? Adjust it
-
Agree on expectations
Written, observable, with a criterion (contract)
-
Observe
Watch and score the work itself
-
Give feedback
Soon, specific: what went well, what to change
-
Practice
Rehearse the change before the next session
-
Ask for feedback
How is supervision working? Adjust it
- Repeats each supervision period
Building an effective supervisory relationship
The relationship starts with a written supervision contract, agreed before supervision begins and revised when circumstances change. It sets out roles, the meeting schedule, expectations, how performance will be evaluated and how feedback will be given and received. Expectations describe observable performance with a criterion, and they are shared before anyone is evaluated against them. Feedback is provided promptly enough to guide the next relevant opportunity, is specific to behavior, identifies what went well and what to change, and offers a chance to practice the change. Feedback also travels upward: the supervisor asks how supervision is working, asks clarifying questions and adjusts. Rapport and trust grow from that reliability, with meetings kept, commitments followed through and the supervisor modeling the conduct they expect from others.
Example
Instead of “be more engaged with clients”, a supervisor writes “during play sessions, comment on what the client is doing at least once a minute” and shows what that sounds like. After the next session she says: “You did it steadily during the block play; it dropped off at snack time. Let’s practice a few comments for snack right now.” Before they finish, she asks what would make their meetings more useful.
Common confusion
Friendliness does not replace structure: a pleasant supervisor with no contract, no criteria and no feedback routine leaves the supervisee guessing. A contract signed once and never reopened is a formality. Feedback that only lists errors, arrives weeks later or stays general (“good job”, “be more professional”) gives the supervisee nothing specific to repeat or change.
Key points (5)
- A written supervision contract, agreed before supervision starts and revised when circumstances change, sets out roles, schedule, expectations, evaluation and how feedback will work.
- Clear expectations describe performance in observable, measurable terms with a criterion, and are shared before the supervisee is evaluated on them.
- Effective feedback is timely, specific and about behavior; it names what was done correctly as well as what to change, and gives a chance to practice the change.
- Feedback runs in both directions: the supervisor asks for feedback on the supervision, asks clarifying questions and adjusts.
- Rapport and trust come from reliable structure (kept meetings, follow-through, the supervisor modeling expected conduct), not from informality that replaces structure.
Equity in supervision practices
Equitable supervision gives each supervisee access to observation, feedback and development opportunities in proportion to their needs, distributed by criteria everyone can see and checked with data on how supervision is actually spread. Equity is not identical treatment: removing a barrier may call for a different schedule, format, language or accessibility arrangement. What stays constant is the performance standard; arrangements change to help people reach it, not to lower it. Evaluating everyone against the same observable criteria and recorded data limits the room for bias and for impressions such as “fit”. Supervisees also need a structured, safe way to raise concerns about fairness, and the supervisor answers with changes and checks whether that route is being used.
Example
Reviewing her records, a supervisor finds that the two trainees who share her daytime hours were observed twice as often as the trainee who works evenings. She adds monthly evening observations and the option to review recorded sessions, and from then on tracks the number of observations per trainee each month.
Common confusion
The same arrangement for everyone does not guarantee equal access: a fixed morning meeting excludes a supervisee scheduled to work at another site at that same time. Removing a barrier is not the same as waiving a standard. And equity cannot be read from good intentions or from a team total; it is checked person by person, against need.
Key points (4)
- Equitable supervision gives each supervisee access to observation, feedback and development opportunities in line with need, allocated by transparent criteria and checked with data on how supervision is distributed.
- Equity may require different arrangements (schedule, format, language, accessibility) to remove barriers, while keeping the same performance standards.
- Evaluating every supervisee against the same observable criteria and recorded data reduces the influence of bias and impressions such as 'fit'.
- Supervisees need a structured, safe way to raise concerns about fairness, and the supervisor responds with changes and checks whether the route is used.
Supervision goals and training
Supervision goals are chosen the same way client goals are: from a direct look at what the person does now. The procedures that move staff performance are behavioral procedures, applied to work behavior.
Figure 3
Behavioral skills training, in order
-
Instructions
Describe the skill and why it matters
-
Modeling
Show it done correctly
-
Rehearsal
The learner performs it
-
Feedback
Praise correct steps, correct the rest
Choosing supervision goals from assessment
Supervision goals start from a direct assessment of the supervisee’s skills: observing them at work, competency checks with a clear criterion and samples of their work, such as data sheets or written programs. Self-ratings, résumés and credentials describe what someone has studied or believes, not what they perform accurately. Among skills below criterion, priority goes to those with the greatest effect on the safety and progress of current clients and to skills that other skills depend on, rather than to a fixed list worked through in order. The supervisee’s environment (setting, caseload, resources, working alone or in a team) shapes which skills are needed and feasible. Cultural variables such as language, values and learning history are explored with the supervisee, not assumed from group membership. Goals are written in observable, measurable terms with a mastery criterion.
Example
A new supervisee reports confidence with preference assessments. Observation shows accurate presentation but frequent recording errors, and her work takes place in busy family homes. Her first goals address recording and session setup. For recording, she scores practice videos against an independently verified reference record and must record at least 95% of scored opportunities correctly across three practice sessions, followed by direct checks in the home setting. Agreement with one supervisor alone would not establish accuracy.
Common confusion
A credential, a degree or a confident self-rating is not a performance assessment. A standard sequence used for every supervisee ignores what this person already does well and what their setting demands. And cultural variables are learned about from the person, not guessed from a group they belong to.
Key points (5)
- Goals start from a direct assessment of the supervisee's skills (observation, competency checks, work samples); self-ratings, résumés and credentials do not show what the person performs accurately.
- Among skills below criterion, priority goes to those with the greatest impact on current clients' safety and progress and to prerequisites for other skills, not to a fixed sequence.
- Cultural variables (language, values, learning history) are explored with the supervisee and shape which goals are relevant and how they are taught; they are not assumed from group membership.
- The supervisee's environment (setting, caseload, resources, working alone or in a team) determines which skills are needed and feasible.
- Supervision goals are written in observable, measurable terms with a mastery criterion.
Performance management procedures
Performance management applies behavioral procedures to the work behavior of staff and supervisees. Behavioral skills training (BST) includes instructions, modeling, rehearsal and performance feedback. A typical initial sequence describes the skill and its purpose, demonstrates it, provides practice, and gives feedback on that performance. Practice and feedback repeat to a defined criterion, with additional instruction or modeling when needed and checks in the relevant work setting. Antecedent procedures, such as task clarification, checklists, prompts and goal setting, specify what to do before or during the work. Consequence procedures include timely, behavior-specific feedback tailored to the person and reinforcement contingent on performance; whether a consequence actually functions as a reinforcer for a given person is tested, not assumed. Lowering response effort helps when effort is the barrier. Delivery can be adapted in language, format and order while the effective components stay in place.
Example
To teach a new technician a discrete-trial program, the supervisor hands over a one-page description of the steps, runs three trials with a colleague playing the learner, then has the technician run trials while she scores every step and gives feedback after each set. They repeat until the technician scores all steps correctly on two sets in a row, then run one more check with the actual client.
Common confusion
Instructions or a demonstration alone do not constitute the complete BST package and do not demonstrate competent performance. A quiz measures answers about a procedure rather than implementation of it. Preserve the four components: the usual initial sequence is instructions, modeling, rehearsal and feedback, while corrective cycles can return to instruction or demonstration as needed. A consequence is identified as a reinforcer by its effect on the staff behavior, not by its name or intended value.
Key points (5)
- Behavioral skills training combines instructions, modeling, rehearsal and feedback, repeated until a performance criterion is met, ideally including the setting where the skill is used.
- Antecedent procedures such as task clarification, checklists, prompts and goal setting specify what to do before or while the behavior occurs.
- Consequence procedures include timely, specific, individual feedback and reinforcement contingent on performance; whether a consequence works as a reinforcer for a given staff member is checked, not assumed.
- Manipulating response effort (putting materials at hand, simplifying a form) can increase desired performance when effort is the barrier.
- Culturally responsive delivery adapts language, format and order of presentation to the supervisees while keeping the components that make the procedure effective.
Diagnosing performance problems
When work is not getting done as expected, the question is not who to blame but what in the work environment is producing the result. The logic resembles a functional assessment, applied to job performance rather than to a client’s behavior.
Figure 4
Diagnosing a performance problem before choosing a fix
Has independent, accurate performance been demonstrated under clear, accessible conditions?
Not yet demonstrated
Check access, instructions, resources and prerequisites; assess further and train documented gaps
Demonstrated in those conditions
Compare usual work conditions, effort and consequences; check fluency or other remaining gaps
-
Antecedents and information
- Ask
- Are the expectation and the instructions clear?
- Matching fix
- Task clarification, checklists, prompts, goal setting
-
Equipment and processes
- Ask
- Are tools, materials and workflow in place and workable?
- Matching fix
- Repair the materials or the process; lower the effort
-
Knowledge and skills
- Ask
- Has observation shown the person can do it?
- Matching fix
- Training to a criterion
-
Consequences
- Ask
- What follows doing it well, poorly or not at all?
- Matching fix
- Feedback, contingent reinforcement, remove competing contingencies
Function-based assessment of performance problems
Assess a defined performance problem before selecting a remedy. Examine expectations and cues, resources and workflow, relevant skills, and the consequences and effort associated with competing actions. Use interviews to generate hypotheses and verify relevant information with observation and work products. Independent success under clear, accessible conditions demonstrates skills under those conditions; prompted success does not demonstrate independence, and nonperformance alone does not establish inability. Skill gaps and environmental barriers can coexist. Select changes matched to the supported barriers and measure their effects before treating a diagnostic hypothesis as confirmed.
Example
Session notes arrive late. Under clear instructions with a usable template, staff independently produce accurate notes in about eight minutes. Observation also shows duplicate logins and no useful feedback about timeliness. The supervisor simplifies access and adds timely, specific feedback, then measures accuracy and on-time completion. This tests the proposed barriers instead of assuming that another writing workshop or a label such as low motivation explains the problem.
Common confusion
Do not default to retraining when existing skills are demonstrated, but do not rule out focused practice or instruction if only part of a task is fluent. Check environmental barriers and skill needs together. Labels such as unmotivated or careless do not identify controlling variables. Diagnostic interview results are hypotheses, and improvement after a matched change still needs adequate measurement and, for a causal claim, experimental control.
Key points (4)
- Before intervening, a performance diagnostic examines antecedents and information, equipment and processes, knowledge and skills, and consequences for the performance.
- A skill deficit (cannot do it even with prompts, time and an incentive) calls for training; a performance deficit (can do it but does not) calls for changing antecedents or consequences.
- The intervention matches the variables identified, including competing contingencies such as high effort, missing consequences or punishment of the desired performance.
- Interview reports are checked against direct observation and permanent products before an intervention is chosen.
Judging supervision by its data
Supervision is an intervention on staff behavior, so it is evaluated the way any intervention is: with repeated measurement, a baseline and graphs.
Figure 5
Judging supervision from a supervisee’s data
↑ Treatment integrity (%)
Observations →
- Baseline
- BST + feedback each session
- Weekly feedback
Data-based decisions about supervision
Whether supervision is working is judged by measured supervisee performance, such as treatment integrity or accuracy of data collection, and by client outcomes, not by how satisfied people feel or how much supervision was delivered. Judging a change requires a baseline and repeated measurement of the same person, graphed and examined through visual analysis, just like client data. If the data show no improvement, the supervisor revises the supervisory practice instead of concluding that the supervisee is the problem. If they show improvement, maintenance is checked as support is faded, because performance that holds only while the supervisor is in the room has not yet been maintained. Social validity ratings from supervisees add information about acceptability; they supplement performance and outcome data and do not replace them.
Example
A supervisor graphs a technician’s integrity on a prompting procedure: about 60% across four baseline observations, then about 90% after two weeks of BST with feedback after every session. She switches to weekly feedback and keeps scoring; integrity stays above 85% for a month, and the technician’s ratings say the feedback was useful and not stressful.
Common confusion
High satisfaction ratings, plenty of supervision hours or a plan followed to the letter do not show that supervision changed performance. A single before-and-after comparison without a baseline is weak evidence. And when performance does not improve, the first question is what the supervisor should do differently.
Key points (4)
- The efficacy of supervision is judged by measured supervisee performance (such as treatment integrity) and client outcomes, not by satisfaction or by the amount of supervision delivered.
- Judging a change requires a baseline and repeated measurement, graphed and examined with visual analysis.
- When data show no improvement, the supervisor revises the supervisory practice rather than attributing the failure to the supervisee; when they show improvement, maintenance is checked as support is faded.
- Social validity ratings from supervisees supplement, and do not replace, performance and outcome data.
Keep going
Practice the whole chapter in the course
Original questions on every task of area I, with explanations and a record of which tasks you have mastered.
Try a sample free, without an account