Supervision by Design: Applying Instructional Technology to Fieldwork Training

Why Two Trainees With the Same 2,000 Hours Are Not the Same

Dr. Allyson Wharam opened with a quick poll. Two trainees finish the same 2,000 hours of fieldwork (supervised hours before certification). How sure are you that both are ready to practice alone as a BCBA (board certified behavior analyst)? The chat filled with ones. Almost nobody rated it high. Hours count time. They do not count skill.

She then showed four sample trainees. Each one logged the same hours. The skills they walked away with looked nothing alike. Some spread is normal and expected. The worry is the wide gap. One trainee leaves ready for the job. Another leaves without the skills the job needs.

She grounded that in work she cited from the field. A 2016 special issue on supervision argued that fast growth makes good supervision more urgent. Many new supervisors were never taught how to supervise. A 2019 survey by Sellers and colleagues found only 27 percent of supervisors used any graph of trainee skills. Cohen and Kodak found that only 33 percent of BCBA directors felt their trainees sufficiently demonstrated hard skills. A third study looked at unrestricted hours and found wide swings in what trainees got to practice. Left to chance, what a trainee practices tracks caseload and billable hours.

What Instructional Design Adds to Supervision

Wharam defined instructional design (planning how people learn) in one line. It is the intentional, systematic process of building learning that leads to real behavior change. Two words carry the weight: intentional and systematic. Training often drifts into content dumping. You hand over a pile of information. That information may or may not change what someone does.

She asked the room what they already do that counts as design. Answers came fast. Writing objectives in behavioral terms. Starting from the assessment. Task analysis (breaking a skill into steps). Shaping and scaffolding (support that fades over time). Planning for generalization (using a skill in new places). Collecting data, graphing it, and checking social validity (whether people find it worthwhile).

Her point was simple. You already own these tools. Most analysts use them well for client programs. Far fewer carry them into staff training and fieldwork. That carryover is not automatic. It depends on your own learning history as a supervisor.

So you'll notice the bolded words here are intentional and systematic.

From the talk — Dr. Allyson Wharam

ADDIE as a Loop You Run at Three Levels

ADDIE (a five step design loop) stands for analyze, design, develop, implement, and evaluate. The chart looks like a straight line. The work is not. You circle back as you go, and evaluation feeds the next round.

In most workplace projects, the analyze step asks whether training is even the right fix. Fieldwork is different. The gap is built in. Someone shows up without the skills, and your job is to close that gap on purpose.

The part that is easy to miss is the layering. You design the fieldwork program for any trainee who walks in. You then design one trainee's path through it. Inside that, you design a single learning experience. Same process, three levels, running at once.

Analyze: Outcomes, Learners, and Setting

The analyze step asks what fieldwork should prepare a trainee to do. What skills does independent practice require? Who is actually entering your program? Wharam described the range she sees. One trainee has fifteen years in the field. The next is fresh out of college. Another taught special education for twenty years. Work history, age, culture, and prior learning all shape how they approach fieldwork.

Setting matters just as much. A school based trainee practices different things than one in an early intervention clinic. Adult services and organizational work shift it again. Billing rules and caseloads can block access to whole skill areas. Those limits belong in the plan, not in the excuses later.

Some outcomes still hold for everyone. Her example: anyone finishing fieldwork should be able to read and analyze a graph. She also warned against borrowing the test content outline (the exam topic list) as your curriculum. It was never built for that. Nothing on it covers building therapeutic relationships, and that is clearly something a trainee needs. Watch the same trap with topics. Preference assessment (finding what a learner likes) is a topic, not an outcome. The outcome is different. A trainee should pick the right assessment for a client. They should run it, or train someone else to run it. Then they should read the results and use them to make a clinical call.

Analysis takes a really long time.

From the talk — Dr. Allyson Wharam

Design: Backward Design, Sequence, and Evidence

Backward design (planning from the end result) asks three questions in order. What should they be able to do? What will show you they can do it? What experiences build it? Wharam leans on behavioral skills training (tell, show, practice, feedback). She also uses talk aloud problem solving (saying your thinking out loud). These are her two workhorses.

Then comes sequence. Nobody runs a functional behavior assessment (why a behavior keeps happening) on day one. The prerequisites are not there. Her scope and sequence (the order skills get taught) starts with foundations, measurement, ethical practice, and skill acquisition. Later blocks cover skills assessment, behavior assessment and planning, and supervision monitoring. That order is not random. It follows component composite analysis (small skills build bigger ones). Data collection and graph reading show up inside goal writing, assessment, and progress monitoring. Teach them early and the complex work gets easier to teach.

Evidence deserves the same care. She listed permanent products (work you can review later). Examples include a written report, a graph, a plan, or a completed data sheet. Add direct observation of the trainee running the procedure. Add the rationale behind the choice. Seeing the right pick once is not enough. Ask them to walk you through the decision so you can hear the clinical thinking. One more axis matters: move from taught and supported practice toward independent performance. Her culminating projects are all done with a real client.

Because once you are a BCBA, you don't have your supervisor with you.

From the talk — Dr. Allyson Wharam

Develop and Implement: Build a Toolkit, Then Tailor It

Develop is where the plan turns into usable things. Sample data sheets. A short article to read before the first attempt. Scaffolds for running behavioral skills training. Quick references on assessment types. Templates that make a trainee's reasoning visible. A fidelity checklist (did they follow each step) so you know the skill was done right. Rubrics and job aids live here too.

She also flagged background knowledge. Some trainees arrive without the basic concepts, or they learned them but not to fluency. Some trainees need basic terms taught first. Response class (behaviors that serve the same purpose) is one example. Topography (the physical shape of a behavior) is another. Build those pieces once and reuse them. You still do the tailoring in the moment, but you are not starting from zero every time.

Implement is where the individual work happens. Her example: a trainee can name the types of preference assessment and has watched several. They have never picked one for a client, and they lean hard on supervisor prompts. Skip the introduction they do not need. Spend the time on clinical decision making and on prompt fading (slowly giving less help). Wait before you answer. Ask what they think first. A curriculum is a toolkit, not a replacement for your judgment about the person in front of you.

It depends on the trainee, whether or not you need them.

From the talk — Dr. Allyson Wharam

Evaluate: Make Progress Measured and Visible

Evaluation sits last on the chart and runs the whole time in practice. Look at what the trainee can do, not just what they covered. Can they give examples, describe it, spot it, and perform it? Can they recombine it into harder work? Does it hold up in messier cases and more independent conditions?

If you oversee several supervisors, check their delivery too. Is everyone running a similar experience? Is the system easy enough to actually use? Ask trainees and supervisors how it feels to use. Look at client outcomes. Then hunt for gaps, bottlenecks, and steps that are not earning their keep.

Two habits she flagged as common problems. Progress gets judged by feel, with no real measure behind it. And progress stays in the supervisor's head. Share the map. A trainee who can see what is still unlearned can ask for those opportunities. Portfolios help here. Have trainees collect the products they build across fieldwork. It becomes evidence for you and a toolkit for them.

So like any instruction, we're using our data to inform decisions.

From the talk — Dr. Allyson Wharam

Choosing Technology: Fit First, Then Affordances and SAMR

Technology is broader than software. A timer counts. So does a pen or a specific form. The test is fit. Start from the goal, the learner, the setting, the activity, and the assessment. She walked through TPACK (tech, teaching, and content together). How you teach a procedure differs from how you teach a principle. Procedures call for models and fidelity checklists. Principles call for visuals, explainer videos, readings, and quizzing. Design tasks call for talk aloud problem solving, worked examples, and templates.

Then weigh affordances (what a tool makes possible) against constraints (what a tool makes harder). Her example was observing a trainee run a procedure. In person you give feedback on the spot and move around the room. You also cannot replay it, the trainee never watches themselves, and people react to your presence. Remote adds scheduling and travel relief plus replay for both of you. It costs you immediate feedback and a clear view. So moderate the trade. Record the live session so the trainee can watch it back with you. She mentioned using a platform that let her tag video with feedback at an exact moment. Not immediate, but tightly tied to the performance.

SAMR (four levels of tech change) gives one last lens. Substitution changes nothing. Augmentation improves the task a little, like a clickable digital version of a paper curriculum. Modification redesigns it, like feedback embedded right in the video. Redefinition enables work you simply could not do before, such as collaboration across time and distance. You do not always need the top rung. You need the rung that serves the outcome.

Not just picking something that seems cool.

From the talk — Dr. Allyson Wharam