Approach FL-CBA preparation as decision practice: for every vignette, name the dimensional measure that fits the behavior, the design that fits the research question and constraints, the procedure that fits the hypothesized function, and the supervision level that fits the supervisee's role. Write a one-sentence justification for each choice. This builds a selection-and-justification habit that ties your notes across scope areas into one coherent preparation framework.
Choosing a measurement procedure that fits the behavior, not the habit
Match the dimensional measure to the behavior's structure and the decision the data must support: rate for discrete responses, duration for extended episodes, latency for response to a prompt, and interval methods only when continuous recording is impractical.
Start by classifying the target behavior: is it discrete and brief (a raised hand), extended (an out-of-seat episode lasting minutes), or a response class with variable forms? Discrete, brief responses support rate or count per session. Extended episodes support duration, either total duration per session or duration per occurrence. When the question is how quickly a client responds to instruction, latency is the measure. Anchoring selection in this classification guards against the core error of the concept: counting a behavior whose real risk lies in its duration.
Worked scenario: a consultant records 'frequency of tantrums' for a client whose episodes range from 40 seconds to 25 minutes. Over two weeks, counts look flat, so the team concludes behavior is stable. Plotting duration per occurrence reveals episodes lengthening sharply. The mistake was treating a duration problem as a count problem; the better decision defines episodes operationally, records duration and rate together, and reviews both before judging stability. This matters because treatment decisions built on the wrong dimension can miss a worsening pattern entirely.
- Rate or count: discrete responses of brief, similar duration (hits, mands, worksheet problems).
- Duration: behaviors that persist for meaningful lengths of time (on-task engagement, tantrum episodes).
- Latency: time from an antecedent, such as a demand or instruction, to response onset.
- Partial or whole-interval recording: continuous observation is impractical and an estimate is acceptable; partial intervals tend to overestimate and whole intervals underestimate total occurrence.
Picking a single-case design when a reversal is off the table
Use a comparison table of design logic: reversal designs demonstrate control by withdrawing and reintroducing treatment; multiple baselines stagger treatment across tiers; alternating treatments compare conditions rapidly; changing-criterion designs shape a target gradually.
Worked scenario: a consultant plans to evaluate functional communication training for escape-maintained hand-biting using an ABAB reversal, expecting to remove the treatment and return to baseline conditions. The plausible mistake is the design choice itself: withdrawing an effective treatment for dangerous self-injury prolongs exposure to risk, and skills established through functional communication training may not revert cleanly, undermining the demonstration of control. The better decision is a multiple baseline across settings or stimuli, introducing treatment in staggered tiers so each tier serves as its own control. The same experimental logic is demonstrated without deliberately removing protection from harm, and the demonstration survives even when the skill is irreversible.
Compare that with a question about which of two prompting procedures works faster. An alternating treatments design suits it because conditions can be counterbalanced within sessions and the procedures are independent. The decision rule: ask what the demonstration requires. If control must be shown by treatment removal, a reversal may fit when the behavior is not dangerous and is likely reversible; if removal is unsafe or the skill will not unlearn, stagger treatment across tiers; if you are comparing conditions head to head, alternate them. Rehearse assigning each design to a constraint rather than to a topic.
| Design | Demonstrates experimental control by | Fits best when | Key caution |
|---|---|---|---|
| Reversal (ABAB) | Withdrawing then reintroducing treatment | Behavior is safe to leave untreated and likely reversible | Irreversibility and ethical risk with dangerous behavior |
| Multiple baseline | Staggering treatment across behaviors, settings, or people | Withdrawal is unsafe, impractical, or the skill is irreversible | Requires independent, functionally similar tiers |
| Alternating treatments | Rapidly comparing two or more conditions | The question is a direct comparison of procedures | Potential carryover between counterbalanced conditions |
| Changing criterion | Stepwise shifts in the performance criterion | Shaping a gradual increase or decrease in one response dimension | Criterion changes must be neither too large nor too frequent |
Reading graphs like an analyst: level, trend, variability, and overlap
Visual analysis evaluates level, trend, and variability within each phase, then compares adjacent phases for immediacy of change, magnitude, and consistency across similar phase changes; a functional relation requires a convincing pattern across all of it.
Train yourself to describe a graph in a fixed order: first the stability of the baseline (is variability low enough to predict?), then the direction of trend, then the change in level at the phase boundary, and finally whether the same change repeats across phase transitions or tiers. This ordered description prevents the error of leaping to the most visually dramatic feature. Overlap between baseline and treatment points is a useful supplement, but a large overlap in a variable baseline is interpreted differently from the same overlap in a stable one.
Apply the ordered description to a paper exercise: sketch a multiple baseline across two settings where treatment produces an immediate, large level change in Setting A and a delayed, modest change in Setting B. An analysis reporting only 'the treatment worked' misses what the graph shows: control is demonstrated in A, but the delayed change in B warrants questions about tier independence or treatment integrity there. Expected observation: your written analysis distinguishes the convincing demonstration from the weaker tier and names what would resolve the ambiguity, such as integrity data from Setting B.
Matching behavior-change procedures to function instead of topography
Select procedures from the hypothesized function of the behavior: reinforcement-based interventions such as functional communication training and noncontingent reinforcement directly address the maintaining contingency, while topography-based choices ignore why the behavior pays off.
Scenario: a behavior plan targets property destruction that assessment suggests is escape-maintained, but the plan centers on planned ignoring and a token system for 'calm hands.' The plausible mistake is applying an attention-based logic to an escape problem: planned ignoring removes attention the behavior does not seek, and the token system competes with a contingency the assessment did not implicate. The better decision pairs functional communication training to teach an acceptable escape response with demand fading and, where appropriate, noncontingent escape or scheduled breaks, plus extinction of the destructive response. Two behaviors that look identical can require opposite procedures depending on function, so procedure selection follows the assessment summary, not the behavior's appearance.
Extend this reasoning to antecedent strategies. For escape-maintained behavior, supports such as behavioral momentum, choice among tasks, or interspersing easy demands reduce the aversiveness that sets up the behavior. For attention-maintained behavior, noncontingent attention on a dense schedule reduces motivation to emit the response. When reviewing any plan, ask one question of each component: which contingency does this component alter, and is that the contingency the assessment identified? Components answering neither part are candidates for revision regardless of how familiar they are.
Drawing supervision boundaries between RBTs and BCaBAs
Delegate implementation-level tasks to RBTs under close, ongoing oversight, design- and analysis-level tasks to BCaBAs who remain under your supervision, and retain decisions that require independent-level judgment, documenting training and fidelity for each delegation.
Worked scenario: an FL-CBA supervises both an RBT and a BCaBA. The plausible mistake is a role swap in both directions: asking the RBT to independently modify a behavior plan mid-week when data look poor, while asking the BCaBA only to collect data and file paperwork. The RBT role is defined by close, ongoing supervision and paraprofessional-level practice, so protocol modification exceeds it; the BCaBA is an undergraduate-level professional who practices under supervision and can carry substantial design and analysis work, so underusing that role wastes qualified capacity. The better decision trains the RBT to implement protocols as written and report data promptly, and tasks the BCaBA with drafting the measurement system and assessment summaries, with the FL-CBA reviewing and approving functional analyses and behavior-plan decisions. Matching delegation to role protects clients and keeps oversight documented.
Build a delegation habit with a two-question check before assigning any task: first, does this task require the supervisee to make independent-level decisions about assessment, treatment selection, or plan change? Second, what documentation shows the supervisee was trained on this task and observed performing it to criterion? Tasks failing the first question belong with the FL-CBA or with a BCaBA under documented review. Tasks passing both can be assigned with a fidelity-check plan. Rehearse this check on a dozen task descriptions until assigning them becomes fast and consistent.
Applying a structured ethics decision model to practice dilemmas
Use a repeatable sequence: state the problem behavior-analytically, identify the relevant requirements in the BACB Ethics Code and applicable law, weigh risks and benefits for the client, consult qualified colleagues, choose the least restrictive effective option, and document the process.
A structured model earns its keep when two principles appear to conflict. Scenario: a team requests a report on a client by tomorrow, but your data are incomplete and a summary now would overstate treatment effects. The plausible mistake is resolving the tension with hedged language that still implies conclusions the data do not support, which conflicts with the requirement to be accurate and truthful in professional communications. The better decision is to state what the current data do and do not support, deliver an interim factual summary on schedule, and commit to a complete analysis once the planned observations finish. Timeliness and accuracy are both honored, and the written record shows the reasoning.
Practice the model on paper dilemmas until the sequence is automatic, because under time pressure an unstructured response resolves the conflict in favor of whichever demand is more visible; a written sequence makes the trade-off explicit. Good practice dilemmas pair a deadline with incomplete data, a stakeholder request with a scope-of-practice boundary, or a supervisor instruction with a documentation obligation. For each, write four sentences: the problem stated behavior-analytically, the requirement implicated, the options generated, and the option selected with its justification. Self-check by confirming every option considered was implementable, and that the selected option is the least restrictive among effective candidates.
A decision-drill exercise, study sequence, and readiness checks
Run weekly decision drills on short vignettes, cycling through the six scope areas; score yourself against a rubric that rewards justified selections, not recognition of terms; track which decision types still require notes.
The drill: write or find a one-paragraph vignette containing a target behavior, a setting, a hypothesized function, and a supervisee. In ten minutes, produce five one-sentence decisions: the dimensional measure and why; the recording method and its known estimation direction; the design and the constraint that rules out alternatives; the procedure and the contingency it alters; and the delegation, with the supervisee's role. Expected observations from honest scoring: measure and design sentences are the easiest to leave vague, so a passing sentence must name the specific option and one justification, not 'use appropriate data collection.' A useful rubric: 5 of 5 specific and justified is a milestone for mastery of that vignette type; 3 or 4 means drill that area again; reusing the same justification across areas signals term-matching rather than decision-making.
An adaptable sequence: first, build the decision rules from the sections above into your own one-page decision sheet; second, apply the drill across all six scope areas, rotating ethics and systems vignettes in with clinical ones, since organizational behavior management questions reward the same selection logic applied to staff performance and systems; third, drill only the decision types that missed the rubric, plus a full mixed set. Readiness checks: you can justify every design choice from the table against a named constraint; you can delegate ten mixed tasks correctly in under five minutes; your graph descriptions follow the fixed order without prompting; and your drill sheets consistently hit the rubric milestone. Administrative details about the credential itself, including application requirements, belong with the BACB rather than with any study guide.
- Drill output: five specific, justified one-sentence decisions per vignette.
- Rubric: 5/5 justified marks the area as mastered for that vignette type; 3/5 flags a re-drill.
- Sequence stage one: build a personal one-page decision sheet; stage two: rotate drills across all six scope areas; stage three: drill missed decision types plus a mixed set.
- Final self-checks: design table justified by constraint, delegation drill under five minutes, ordered graph descriptions, consistent rubric milestones.
References and further reading
Use these references to explore the concepts and check the latest information from the relevant organizations.