Press Pause: What the RBT 40-Hour Training Was Really Meant to Do

The Seven Content Areas the BACB Requires in a 40-Hour Training

The BACB (the board that sets RBT rules) names seven content areas for the RBT (the frontline therapy staff credential) 40-hour training. Crystal Harms walked through each one on purpose. Her ask of the room was simple. Find where your program truly teaches an area, then find where it only mentions it in passing.

Area one is an intro to ABA (the science of behavior change). Harms wants new staff to know the history behind the job. The field grew out of concern for adults living in residential care. Neglect and abuse were the problem it was built to solve. Staff turnover ran high because people did not know how to do the work or how to answer hard behavior. Early work in California and years of parent advocacy brought us the rest of the way.

Data collection and graphing gets glossed over now. Software takes the data and draws the graphs for us. Harms has mixed feelings about that as a professor. She thinks the platforms have started to drive how we teach. Her example came from verbal behavior work (teaching language using behavior methods), where staff used to intersperse targets during a session. She sees that skill fading, and learners pick up new words more slowly because of it. Assisting with behavior assessments is thinner still. The exam touches ABC data (what happens before and after behavior) and preference assessments (finding what a learner likes most), while trainees rarely learn to help with the structured tools teams actually use.

Behavior-change work needs multiple exemplars (many different correct examples). Two behaviors can look the same and still need different responses based on function (the reason a behavior happens). Talking through those scenarios is not enough, because the real thing adds stress, noise, and questions. Documentation and reporting gets more time than it used to, since audits and payers hold up payment. Ethics often gets two or three examples and a quick interview, then the team calls it done.

The best way to learn that is to see it and to practice it.

From the talk — Crystal Harms, MEd, BCBA, LBA

Why 40 Hours of Video Became the Default

Harms described what she sees near her in South Jersey. Some groups now require the full 40 hours before they will hire you. The candidate finds a course and watches it alone, off the clock. That sidesteps state rules about paying for training you require. It looks like a savings. Harms argues the savings do not hold once you count turnover and rework. An attendee asked whether the hiring rule is even legal. She said labor law varies by state, and listing the course as a hiring requirement is often allowed. The riskier move is having a reported employee watch training off the clock. She keeps a business advisor for her own state and suggests you check yours.

Other groups do pay for the hours, which she respects. They sit the new hire at a computer in the clinic or the office. Some pay a lower rate than the clinical rate, and she finds that fair. The person still watches 40 hours of video. There is almost no live contact with a BCBA (the supervising certified behavior analyst) during any of it.

Then her inbox fills up. Students write to say they finished the hours on their own. They ask her to run their competency check for them. She has to explain why a stranger should not sign that form. The BACB has asked trainees to name who trained them and where the training happened. It wants a clear thread from training to assessment to exam.

One local leader told her about a plan he was proud of. New hires observe sessions for three days and work with no clients. Then the team runs the competency and signs them off. Harms had to take a breath before she answered him.

It's still not what we're talking about.

From the talk — Crystal Harms, MEd, BCBA, LBA

The Delivery Components the BACB Wants Beyond Video

The BACB also spells out how the training must be delivered. Video alone does not meet that bar. The outline calls for didactic instruction (telling someone how it works). It also calls for modeling with exemplars and nonexemplars (examples of what not to do). Add in vivo role-play with feedback (practice acting it out with peers) and interactive activities. Harms uses video content herself and pulls from three series she likes. Her point is that video is one tool in the box, not the whole box.

Lecture is the easy part, and video handles it well. Few adults learn much from hours of straight audio. Harms counts herself in that group and says so plainly. She watches trainees check out partway through a long block of content. That is not laziness, and it is not a character flaw. It is just how attention works over a long stretch of video. In her own program she never allows more than two hours of video in a day.

So she builds in moments to respond. Our own literature is clear on this point. More chances to respond with feedback produce better learning. She points to classrooms where every student holds a small whiteboard and answers every question. Choral responding (everyone answering out loud together) turns into verbal mush, while the whiteboard shows you each answer. Even a few minutes of back and forth after a session note counts as active instruction.

Nonexemplars deserve their own note here. With children we often avoid showing the wrong response on purpose. Adult learners need that contrast to sharpen a definition. Seeing what a behavior is not helps them see what it is. Role-play gives practice without the stress of the therapy floor, and she admits nobody cheers when she says the words. Most people tell her afterward that it helped, as long as the scenario is structured. Her praise activity is the simplest version of all this. Trainees pair up, one plays the child, and the other has one minute to praise every response with varied wording. She gives them a sheet with a hundred ways to say nice job and models it first. They laugh, and they notice how stale their own praise has become.

But it also could be something as simple as a BCBA or a very highly, highly qualified RVT.

From the talk — Crystal Harms, MEd, BCBA, LBA

Behavioral Skills Training and Its Active Ingredient

BST (four steps that teach a skill) is instruction, modeling, rehearsal, and feedback. Harms did not belabor the model, since most of the room could recite it. Her point was sharper than a review. This is our own science, and the evidence for it is strong. We use it with clients every day and then skip it with staff.

Modeling works best live, with a real client, when that is possible. Live modeling also produces errors in front of the trainee. Harms treats that as useful rather than embarrassing. Name the mistake out loud, then run the step again the right way. The trainer builds rapport and shows that a mistake is not the end of the world. Supervisors often tell her they showed the new person how. Showing is step two of four.

Rehearsal is the part that builds the skill, but only when feedback is tight. Harms coaches a five to one ratio with her trainers. That means five positive interactions for every correction you give. Praise has to be specific about the behavior, not a quick nice job. Corrective feedback needs a frame around it. Tell the person you know they want to do this well. Say the feedback is meant to support them. Some trainees have never heard feedback delivered that way in their lives. Repeat the frame every time you correct, not just the first time.

Exposure is not rehearsal, and the difference matters. Watching another video is exposure. Practicing the step and hearing exactly what to fix is rehearsal. Mastery is also not the finish line she wants. Harms wants trainees fluent (fast, easy, and accurate performance). That is the standard we set for our clients, so set it for staff too.

Modeling is not teaching.

From the talk — Crystal Harms, MEd, BCBA, LBA

Building a Treatment Integrity Checklist In House

A treatment integrity checklist (a step list for one protocol) turns a protocol into steps you can see. It is a task analysis (broken into small, observable steps) written for adults instead of learners. You use it while you teach, while you model, and while the trainee rehearses. Then you score it and give feedback against the same steps. It also lets you spread the check over days instead of one long exam.

Harms believes these are better built in house. The science is shared, but the settings are not. A home program, a clinic, and a residential adult program need different steps on the page. Setting, culture, and client needs all shape the list. So does how many people are in the room, and how much help the client needs. She guesses that is why the BACB stopped short of writing one list for everyone.

Pick your fidelity target (how closely steps were followed) before you train anyone. Most of the literature does not expect one hundred percent. Harms put the common range at 80 to 85 percent on this kind of checklist. That is a very different tool than the competency form. The form asks yes or no and then takes your initials. It does not tell you how well the person performed. It does not tell you what to teach next week.

She shared one checklist she built for continuous measurement (tracking every instance, not just samples) during the talk. She called it a first draft rather than a template to copy. It is headed into a small university clinic she is helping open. About six years of testing versions with a partner clinic sits behind it. Take her page as an example, then write the one your team actually needs.

I encourage you to think about that.

From the talk — Crystal Harms, MEd, BCBA, LBA

Where the Initial Competency Assessment Actually Fits

The initial competency assessment (the BACB sign-off before the exam) covers 19 tasks across five domains. Harms lined that up against the seven content areas. They do not match one to one. Some skills the BACB wants taught in training never get assessed at all.

That gap has two answers, and you need both. Teach past the test during the 40 hours. Then keep teaching after the credential through PDUs (ongoing learning units after credentialing). Transitions are her favorite example of a missing skill. Helping a learner move from one activity to the next matters at any age. It barely shows up on the assessment form.

Naturalistic teaching (teaching during play and daily routines) is another thin spot. So is a clear view of DTT (short, structured teaching trials). Trainees assume discrete trial means a desk and a stack of flashcards. It can happen on the floor during what looks like play. Teach the structure of the trial, not the furniture around it.

The timing rules are strict, so build around them. Training must span at least five days and no more than 180. It has to finish before the competency assessment begins. Do not put the same date on both, because that gets flagged. The assessment must land within 90 days of the exam. The assessor must be qualified. They must watch tasks with a real client, and the BACB has grown stricter there. Records stay for seven years, and both the trainer and the trainee can be audited. By assessment day, the basics should already be fluent. The assessment confirms skill. It does not build it.

I have a training plan that, uh, goes across 12 days.

From the talk — Crystal Harms, MEd, BCBA, LBA

A Redesign That Cut Hours and Raised Competence

Harms put a question to the room before her last slides. How many days does it take to turn a brand new hire into a competent RBT? Answers came back across a wide range, from about a month to well past 140 days. Most people landed in months, not days. She then pointed to a report she read in the trade press and heard at a conference in Boston. A provider worked with ABA Technologies on exactly this problem. Before the redesign, real competence took them about 136 to 139 days.

After the redesign, the training ran 66.5 hours across 10.5 days. Harms noted her own plan runs 12 days, which is close to the same shape. Sixty-day technician turnover came in at 44 percent. She was honest that 44 percent still sounds high to her. Many teams have seen worse numbers than that. They also reported lower training costs, which is the argument that reaches a C-suite.

Three changes drove the result. First, they matched each skill to its best teaching format instead of defaulting to video. Second, they re-sequenced the program so teaching, rehearsal, and feedback interweave instead of front-loading all the content. Third, they added competency checks along the way rather than one at the end. That third change is the checklist idea put into practice. Harms pointed listeners to Alice Betts at ABA Technologies for more of this work.

The results were not accidental. They were designed.

From the talk — Crystal Harms, MEd, BCBA, LBA

Mentor RBTs and Support After the Handoff

Harms ended on the constraint that no pep talk can fix. Good training takes more BCBA time than most companies can fund. She consults with several groups and sees the squeeze up close. The BCBA already carries a full clinical load. Training then gets pushed into whatever hours are left over, which is not a plan.

Large groups can hire a training director to carry that weight. That person writes the materials, the checklists, and the daily schedule in the background. The clinical team can then run training without giving up client time. Small and mid-sized groups cannot staff that role. Many of them grew on word of mouth and strong care, and the way insurance pays keeps capital tight.

Her answer for those teams is a layered staff. A senior BCBA with real leadership training owns the program. Below that sit lead RBTs and mentor RBTs. A lead RBT runs one client team day to day. A mentor RBT gets paid non-billable time to teach other RBTs. It also opens a path up for strong RBTs who do not plan to become a BCBA.

Support cannot stop at the handoff either. One attendee asked how to answer critics who say RBTs are not qualified for moment-to-moment calls. Harms did not dodge the question. The model trades some expertise for many more hours of contact with the science. A BCBA-only model could not deliver enough hours for meaningful progress. So the real fix is fast access to help. Her team keeps a clinician watching a protected chat during every session. Sometimes she just asks the person to get on the phone.

I'm heartbroken when I see situations in which there are RBTs alone.

From the talk — Crystal Harms, MEd, BCBA, LBA