QABA QASP-S common mistakes
QABA QASP-S · Understand the traps and how to avoid them.
When a scenario describes a client who lacks a skill (e.g., no way to ask for a break, no way to recruit attention), the candidate selects a consequence-based reductive tactic such as extinction, response cost, or time-out as the primary answer.
Reductive procedures are the most salient and most heavily drilled tactics, so they are retrieved first. The stem's emphasis on a missing repertoire ('never learned to ask,' 'has no other way to') gets read as a behavior excess rather than a skill deficit.
What to do instead
Before answering, label the problem out loud as either a skill deficit (behavior never learned) or a performance deficit (behavior exists but is not occurring under these conditions). If it is a skill deficit, the correct option must include teaching a functionally equivalent replacement; only select reductive tactics when the replacement is already in the repertoire and the stem is asking about reducing an unsafe excess.
The candidate picks an answer containing a definition such as 'tantrums when frustrated,' 'gets anxious,' or 'is aggressive' as an acceptable target behavior definition or as an adequate description in a data sheet.
Everyday language is fluent and feels descriptive, so it slips past as 'good enough.' Candidates do not run the dead-man's test or check for observability and measurability, because those checks feel like extra steps under time pressure.
What to do instead
Apply two quick screens to any definition in an answer choice: (1) Could a dead person meet it? If yes, reject. (2) Could two observers with no shared history agree on occurrence/non-occurrence in real time? If not, reject. The correct option will name observable movement or product (e.g., 'says "no" above conversational volume,' 'leaves seat more than 1 m for 5+ s').
For a behavior that occurs in long, variable episodes (e.g., crying bouts lasting 2–20 minutes), the candidate selects frequency or event recording, or selects partial-interval recording when the goal is an accurate estimate of total duration.
Frequency/event recording is the first method taught and is mentally tagged as 'the standard method.' Candidates match on the word 'behavior' instead of matching the measurement system to the dimension of behavior the goal actually targets.
What to do instead
Identify the dimension the goal names before looking at the options: how often → count/frequency or rate; how long → duration (total or per occurrence); how intense → magnitude; whether it happened at all in a period → interval. If the stem mentions long or variable episodes, duration is the answer; if the stem asks about over- or under-estimating, use the interval rule (partial interval overestimates, whole interval underestimates, momentary time sampling is closest to true duration).
In a contingency description, the candidate mislabels the procedure - calling removal of a task after a behavior 'negative punishment,' calling a reprimand 'positive reinforcement because attention is added,' or calling a token loss 'extinction.'
The terms positive/negative get mapped to good/bad and punishment/reinforcement get mapped to the unpleasantness of the event, rather than to the technical definitions of stimulus presentation/removal and future frequency change.
What to do instead
Parse every contingency in a fixed order: (1) What is the consequence - is a stimulus added or removed? (2) What happens to the future frequency of that behavior - increase or decrease? Answer (1) gives positive/negative, answer (2) gives reinforcement/punishment. If the future frequency is not stated or implied, the stem is asking you to identify the procedure by its formal definition, not by how it feels to the client.
When a technician's data show a supervision problem (missed sessions, poor fidelity, late paperwork), the candidate selects 'retrain the technician' or 'review the protocol again' as the fix.
Training is the tool candidates are most confident in, and retraining feels like a safe, supportive answer. The can't-do vs. won't-do distinction is not spontaneously applied, so a performance problem is misdiagnosed as a knowledge problem.
What to do instead
Run the performance diagnostic first: does the technician perform the skill correctly when directly observed or prompted? If yes, the deficit is performance, and the correct answer is a performance-management fix - clear expectations, feedback, self-monitoring, scheduled checks, or rearranged contingencies. Choose retraining only when the technician cannot perform the skill even with prompts, or has never been trained on it.
In a scope-of-competence item (e.g., a referral for a client with a feeding disorder, a court-ordered case, or a co-occurring psychiatric condition), the candidate selects 'accept the case and seek supervision or continuing education while working.'
The value of access to services and the desire not to abandon a client pull hard, and 'seek supervision' sounds like a responsible-sounding compromise. Candidates do not check whether supervision can make the task within competence before the work begins.
What to do instead
Test each option against the sequence: identify the limits of competence → obtain the needed training or supervised experience → then provide the service, or refer/coordinate. An option that has you delivering the service first and gaining competence on the job is the distractor; the correct option either declines/refer or arranges qualified oversight and training before the service starts.
Presented with a request for information (a grandparent asking how the program is going, a teacher asking for the behavior plan, a school asking for the client's records), the candidate selects the option that shares the information because the requester is involved or well-meaning.
The stem is written warmly, and the candidate evaluates whether sharing is kind and helpful rather than whether it is authorized. Consent and release-of-information considerations get skipped when the requester sounds sympathetic.
What to do instead
Before choosing, ask: is this person on the signed release, and is this the type of information the release covers? If not, the correct answer is to decline to share until consent is obtained, and to tell the requester what the process is. Warmth of the requester is irrelevant; authorization is the only thing that decides the item.
When a caregiver asks the candidate's character a question about diagnosis, prognosis, medication effects, or whether the child 'has autism,' the candidate selects the option that answers the question directly with the requested information.
Caregivers' questions feel like a service opportunity, and the candidate knows the answer is roughly what the caregiver suspects, so answering feels like collaboration. The boundary between behavior-analytic scope and medical/diagnostic scope is not checked.
What to do instead
Screen the content of the question: is the answer a behavior-analytic statement about observable behavior, or is it a diagnosis, prognosis, or medical judgment? If it is the latter, the correct option refers to or coordinates with the qualified professional (and supports the caregiver in raising the question there), rather than providing the information oneself.
For a scenario involving suspected abuse, neglect, or an immediate safety risk, the candidate selects the option that reports to the supervisor, the family, or the agency's internal contact as the complete response.
Internal chain-of-command reporting is what candidates practice most, and it feels like escalation. Candidates conflate 'tell my supervisor' with 'fulfilled my legal obligation,' and the internal action is listed as a plausible distractor.
What to do instead
Separate the two duties: internal notification (supervisor/agency) is good practice, but mandated reporting to the designated external authority is the non-delegable legal duty. When a question asks what must be done, choose the option that includes reporting to the designated protective/legal authority; choose internal notification only when the stem asks what to do in addition or regarding a non-mandated practice concern.
The candidate treats interobserver agreement and treatment integrity as interchangeable, selecting 'collect IOA data' when the stem describes a plan that may not be implemented as written, or selecting 'check fidelity' when the stem asks whether two observers record the same behavior the same way.
Both involve a second person checking someone against a standard, so the surface similarity is strong. The candidate matches on the setup (a second observer, a checklist) rather than on what is being compared.
What to do instead
Name the two variables being compared in the stem. If it is behavior data compared across observers → IOA. If it is the staff member's implementation compared to the written procedure → treatment integrity/fidelity. Then confirm the answer choice uses the matching term; the correct option will not swap the labels.