Behavior Assessment
Domain B is 8 of the 75 scored questions (11%) and has only three tasks — B.1, B.2 and B.3 — but it carries more scope-of-practice weight than any other technical domain. Look closely at the verbs: you conduct preference assessments, but you participate in skill assessments and you participate in components of functional assessment procedures. That difference is the whole domain. Candidates lose points here by answering as though the RBT were the person choosing the assessment, interpreting the results, or naming the function of a behavior — all of which belong to the supervising behavior analyst.
What this domain asks you to do (B.1–B.3)
Three tasks, three verbs, and the verbs are the answer key to most of the questions. "Conduct" in B.1 means you run the preference assessment yourself, from the protocol your supervisor gives you. "Participate in" in B.2 and B.3 means you are one part of a larger process: you administer the items you were trained to administer, you record what you see, you hand the data to your supervisor, and the supervisor selects the instrument, decides what the results mean, and writes the plan. Every time an option in a question has the RBT drawing a conclusion about a client's skill level, diagnosis, or behavioral function, that option is describing your supervisor's job.
Why preference assessments exist at all
A reinforcer is defined by its effect, not by its appeal: a stimulus is a reinforcer only if delivering it after a behavior makes that behavior more likely in the future. That definition creates a practical problem — before you can reinforce anything, you need a good guess at what will work, and asking a caregiver "what does he like?" produces a list that is often out of date, sometimes wrong, and never ranked. A preference assessment is the systematic way of producing that guess. It samples what the client actually chooses and engages with right now, today, before this session, because preferences shift with satiation, mood, illness and the time of day. That is why plans often call for a brief preference assessment at the start of every session rather than once a month. Remember the ceiling on all of this, though: a preference assessment identifies a highly preferred item; it does not prove the item is a reinforcer. Only the behavior data — does responding actually increase when the item follows it — can do that, and confirming a reinforcer effect from data is your supervisor's call.
The preference assessment formats you must be able to run (B.1)
Free operant: you make items freely available in the environment, let the client move among them, and record what they approach and how long they engage with each. Nothing is removed and no demands are placed, which is exactly why it is the format of choice for a client who becomes upset when items are taken away, and why it is often used with a client you barely know. It gives you duration of engagement rather than a clean ranking. Single stimulus (successive choice): one item is presented at a time and you record whether the client approaches it. It is the crudest format, and it is used for clients who cannot yet choose between two things. Paired stimulus (forced choice): two items at a time, every item eventually paired with every other, and you record which one is selected. It produces the most reliable ranked hierarchy and takes the longest. Multiple stimulus with replacement (MSW): an array of items, the client picks one, and after the pick the chosen item is put back and the array is re-presented. Multiple stimulus without replacement (MSWO): same array, but the chosen item is removed and the client picks again from what remains, producing a full ranking in one sweep — which is why MSWO is the format most often used for a quick pre-session check. Match the format to the client in front of you: free operant when removal triggers problem behavior, MSWO when you need a ranking fast, paired stimulus when the plan needs the most precise hierarchy and time allows.
Skill assessments and your role in them (B.2)
Skill assessments map what a client can already do and where the gaps are, so the supervisor can pick targets and sequence them. Three families show up in the outline's own examples. Curriculum-based assessments walk through a structured list of skills across areas such as requesting, imitation, play and academics, and the results feed directly into program targets. Developmental assessments compare a client's skills against typical developmental sequences. Social skills assessments look specifically at initiating, responding, sharing, and other interaction skills. Your part is administration and data, not selection and interpretation: your supervisor chooses the instrument and trains you on the items you will run; you present the items exactly as trained, score exactly what the client did, and avoid the two habits that ruin skill assessment data — prompting to help the client succeed, and scoring something as mastered because you have seen them do it before. If a client passes an item today that you know they usually fail, record today's performance and mention the discrepancy to your supervisor. And when a caregiver asks you what the results mean, or what level their child is at, that question goes to the supervisor.
Functional assessment: the three levels (B.3)
Functional assessment is the process of working out why a behavior happens — which consequence is maintaining it — and it comes in three levels of evidence. Indirect assessment gathers information from people who know the client: interviews, rating scales, questionnaires. It is fast and it gathers history, but nobody observed the behavior, so it produces hypotheses at best. Descriptive assessment is direct observation of the behavior as it naturally occurs, most often recorded as ABC data — the antecedent (what happened immediately before), the behavior (operationally defined), and the consequence (what happened immediately after). Scatterplots and ABC checklists are descriptive tools too. Descriptive data are real observations, but they show correlation only: you can see that the behavior tends to occur when demands are placed and tends to be followed by a break, which suggests escape but does not prove it. Functional analysis (FA) is the experimental level. Conditions are deliberately arranged and alternated — typically an attention condition, a demand or escape condition, an alone condition and a play or control condition, sometimes a tangible condition — so that one variable at a time is manipulated. Only an FA demonstrates a functional relation, which is why it sits at the top of the evidence ladder. It is also the level that most requires supervision: an FA involves deliberately evoking problem behavior, and you take part only when you have been trained on the specific protocol, only under your supervisor's direction, and only within the safety parameters written for that client.
Recording ABC data without interpreting it
ABC recording is the piece of functional assessment you will do most often, and it is graded on discipline rather than insight. The antecedent is what happened in the seconds immediately before the behavior — the instruction given, the item removed, the peer who walked past, the transition announced — described in observable terms. The behavior is written using the operational definition from the plan, not a summary word. The consequence is what happened immediately after: the break provided, the attention delivered, the item handed over, the demand repeated. Write all three as a stranger would see them on a video with the sound on. What you do not write is the function. "Antecedent: therapist presented a demand. Behavior: threw materials. Consequence: demand withdrawn for two minutes" is a good record. "Client was attention-seeking" and "behavior was clearly escape-maintained" are conclusions, and identifying the function of a specific client's behavior is the supervisor's determination made from the full data set, not a judgment you make in the moment. There is one exception worth knowing: task D.1 asks you to identify common functions of behavior as a category — that is knowledge you need in order to understand your plan, and it is different from declaring a particular client's function in a record or to a family.
Key facts and definitions for Domain B
This domain has fewer numbers than any other and more vocabulary that must be applied rather than recited. The list below is the applied core: know which format matches which client, and which level of functional assessment supports which kind of claim.
Worked scenario: which assessment, and whose job?
An RBT arrives for a session with Dev, a 6-year-old who has recently stopped engaging with the two toys listed as reinforcers in his program. Dev becomes upset when a preferred item is taken out of his hands. The plan directs the RBT to identify current reinforcers before running acquisition targets. Four options: interview Dev's mother about what he likes at home and use her top choice; run a paired-stimulus assessment with eight items so the ranking is as precise as possible; set out several items, let Dev move freely among them, and record what he approaches and how long he engages; contact the supervising BCBA to ask which items should be used today. All four are things real technicians do, which is the point. The interview is indirect information, and it is useful background — but it is second-hand and does not tell you what Dev will work for in this session, so it is the weakest basis for the decision. Paired stimulus produces the best hierarchy, but it requires repeatedly presenting and then removing items, which is precisely what evokes Dev's upset, and with eight items it consumes most of the session. Contacting the supervisor looks safe, and candidates over-choose it — but the plan already directs the RBT to identify current reinforcers, the RBT has been trained to conduct preference assessments, and this is squarely within B.1. Escalating here is unnecessary and costs Dev his session. The free-operant format is best: it places no demands, removes nothing, produces usable engagement data for the client in front of you today, and is exactly the format indicated when item removal evokes problem behavior. The general shape of this reasoning matters more than the answer — when the plan covers it and you are trained for it, you do it.
Exam traps in Domain B
Domain B traps come in two flavours: confusable assessment formats, and role confusion. The format traps are settled by a single detail in the stem. The role traps are settled by asking who, in this scenario, is being asked to draw a conclusion.
Key takeaways
Eight sentences that decide most Domain B items. Note how many of them are about role rather than technique — that is a true reflection of how the domain is tested.
Practice stays free. The full RBT (Registered Behavior Technician) study guide is the material itself, taught start to finish — a downloadable PDF + EPUB you keep.