BST Checklist for New RBTs: Instruct, Model, Rehearse, Feedback

A BST checklist for training new RBTs, with worked examples for pairing, prompting, and data collection, from a BCBA-led CEU.

Key takeaway

This is a printable BST checklist you can use during the first week with a new RBT, with three worked examples and a step most templates skip. Save the page, print it, or copy the checklist into your supervision notes before the first session.

Watch the full CEU recording

Dunder Mifflin’s Guide to Training and Onboarding: Lessons from The Office

Mellanie Page · 57 min
Watch on openceu.com →

This is a printable BST checklist you can use during the first week with a new RBT, with three worked examples and a step most templates skip. Save the page, print it, or copy the checklist into your supervision notes before the first session.

Behavior Skills Training (BST) is the gold-standard way to teach a new RBT a skill. Most teams know the four steps. Few teams write them down as a checklist a supervisor actually uses on a Tuesday morning. This page gives you that checklist and shows you how to fill it out for three of the highest-frequency skills a new RBT needs: pairing, data collection, and client engagement.

What BST is, in one short paragraph#

BST stands for Behavior Skills Training. It is a four-step teaching package: you tell the learner what to do, you show them, they try it, and you give them feedback. You repeat the loop until the learner can do the skill on their own. The package works for any teachable behavior, which is why it is the right tool for onboarding a new RBT. In Mellanie Page's CEU, she gets to the heart of it in a single line about why role-play tends to beat lecture:

Modeling techniques and role-playing. I can do the job perfectly. If you model, give me an opportunity to do it and give me that feedback and we wouldn't have to use any words.

That sentence is the whole package: model, rehearse, feedback. Add a clear instruction at the front, and you have BST.

The four steps as a checklist (instruct, model, rehearse, feedback)#

Print or copy this. Use one checklist per skill, per RBT, in the first week.

Step 1: Instruct. - [ ] Name the skill in plain language. No acronyms in the first sentence. - [ ] Write the operational definition. Who does what, where, when, and how. - [ ] State why the skill matters for the client and the team. - [ ] Share one non-example so the RBT knows what the skill is not.

Step 2: Model. - [ ] Show the skill in a real or near-real context (not a slide). - [ ] Narrate the key decisions out loud while you do it. - [ ] Model it twice if the skill has more than one step. - [ ] Pause and check for questions before moving on.

Step 3: Rehearse. - [ ] Have the RBT do the skill in a low-stakes practice setting. - [ ] Use a real prop, real materials, or a stand-in client. - [ ] Keep the trial short so you can give feedback fast. - [ ] Run at least three rehearsal trials before judging mastery.

Step 4: Feedback. - [ ] Lead with one thing they did well. - [ ] Name one specific behavior to change. - [ ] Show the change, do not just describe it. - [ ] Have them try again on the spot.

Mastery sign-off. - [ ] RBT performs the skill correctly on three consecutive trials with no coaching. - [ ] RBT can describe what they did and why in their own words. - [ ] Date and supervisor initials below.

Step zero most templates skip: define mastery in observable terms#

Most BST templates jump straight to "instruct." That skips the work that makes the rest of the checklist usable. Before you teach a skill, write down what mastery actually looks like. You need two things: an operational definition of the target behavior and a mastery criterion (how many trials, in what setting, with what level of help).

This is the same logic you apply to client programs. You would not run a tact program without an operational definition and a mastery criterion. Do not run RBT training that way either.

The other reason this step matters is that it lets you build in a "pause and check" moment between instruct and model. Mellanie names the gap clearly:

I often see a lot of lecture-based learning where you have an ongoing delivery of information and at key concepts or key principles, there is no opportunity to ask.

If you wrote the mastery definition in plain language, the RBT can ask the question that matters: "What counts and what does not?" That question, asked once at the front, saves you three weeks of correction later.

Worked example 1: BST for pairing#

Skill name (plain language): Pair with a new client so the client wants to be near you.

Operational definition: Sit within arm's reach of the client. Offer access to preferred items or activities at least once every two minutes. Do not place any demands during the pairing window. Match the client's pace and energy.

Mastery criterion: RBT runs a 10-minute pairing block with a stand-in client, offering at least 5 access opportunities, placing zero demands, and staying within arm's reach for the full block. Three consecutive trials.

Instruct: Read the definition aloud. Show one non-example (sitting across the room, offering one toy, then asking for it back).

Model: You run a 5-minute pairing block with the supervisor playing the client. Narrate every decision. "I am moving the truck closer because she looked at it. I am not asking her to clean up yet."

Rehearse: RBT runs a 5-minute block while you watch silently. No coaching during the block.

Feedback: One thing they did well, one specific change, then a second 5-minute block.

This is the skill that determines whether the rest of week one goes well. Get the BST loop right here and you save yourself a hundred small corrections later.

Worked example 2: BST for data collection#

Data collection is the skill that supervisors complain about most and train for least. The fix is to treat it as a BST target with the same rigor as a client program. Mellanie names the gap exactly:

Collect data. So, what's the point of collecting data? How often do I collect the data? ... We'll see maybe five learning opportunities an hour and we're wondering, why isn't this person collecting data? Well, did they know? Do they know which skills to collect data on? Do they know how to track data effectively?

Skill name: Record trial-by-trial data on a target program.

Operational definition: For each learning opportunity, the RBT (a) presents the SD, (b) waits the prompt delay, (c) marks the response as independent, prompted, or no response, and (d) moves to the next trial within 15 seconds. Records at least 10 trials per program per session.

Mastery criterion: RBT runs 10 trials on a mock program with 100% accurate scoring and a per-trial cycle under 15 seconds. Two consecutive sessions.

Instruct: Walk through the data sheet. Show the scoring codes. Show one non-example (writing notes in the margin instead of scoring trials).

Model: Run 5 trials yourself with the supervisor as the client. Narrate the scoring decision after each trial.

Rehearse: RBT runs 10 trials. You score the same trials independently. Compare.

Feedback: Walk through any trial where your scoring did not match theirs. Have them re-run two trials right away.

Worked example 3: BST for client engagement#

"Engage with the client" is the phrase supervisors use when they mean nine different things. The RBT cannot read your mind. Mellanie pushes on that exact problem:

Engaging with your clients. Does that mean I follow their lead? Do I do this only when they're following instructions or when they're not engaging in maladaptive behavior? Do I just do it all the time? And what does it mean?

Use BST to land one specific version of engagement, then teach the next version later.

Skill name: Follow the client's lead during free-play blocks.

Operational definition: During unstructured time, the RBT (a) stays within 6 feet of the client, (b) joins whatever activity the client chooses without redirecting, (c) responds to any client-initiated bid within 3 seconds, and (d) does not introduce a new activity unless the client disengages for more than 30 seconds.

Mastery criterion: RBT runs a 10-minute free-play block meeting all four criteria with 90% accuracy. Three consecutive trials.

Instruct, model, rehearse, feedback: Same loop as the other two examples. The point is that "engagement" is now one teachable, scorable behavior instead of a vague directive.

Print this page. Fill out one checklist per skill in the first week. Three skills is enough for week one: pairing, data collection, and one engagement target. Add more skills in week two once the first three are at mastery.

A note on time. A full BST loop on one skill takes 20 to 40 minutes including rehearsal trials. Block the time on your calendar before the RBT's first day. The most common failure mode is supervisors who plan to "find time" during session and then never do.

FAQ#

What are the four steps of BST in ABA? Instruct, model, rehearse, and feedback. You tell the learner what to do, you show them, they try it, and you give them specific feedback. You loop until they meet the mastery criterion you set up front.

How long does BST take for a new RBT skill? Plan 20 to 40 minutes per skill for the first pass, including at least three rehearsal trials. Simple skills like data collection on one program can land in 20 minutes. Skills with more steps, like running a discrete trial program with prompting and prompt fading, take longer and often need a second BST loop later in the week.

How do you know when an RBT has mastered a BST skill? Set the mastery criterion before you teach. A common standard is three consecutive trials with correct performance and no coaching, plus the RBT can describe what they did and why in their own words. Write the criterion at the top of the checklist so you do not move on too early.

Next step#

Print the checklist, pick the three skills your next new RBT needs in week one, and watch the source CEU for the full context on values, expectations, and connection that wrap around BST.