Connection-Based Behavior Intervention: Sessions Without Demands

Two cases pull this idea into focus. Valerie sat with a student every day for rapport sessions before consent for a practical functional assessment came through, just watching what he liked and how he wanted to interact. Alana worked with David, a 30-year-old, and was told by Dr. Megan DeLeon to back up and build a connected relationship first by running a quiet choice assessment around what he picked when no one was asking. Both used joint activity routines borrowed from the Early Start Denver Model (ESDM), and both moved through the same three-phase progression: connection, then skill, then independence. A Board Certified Behavior Analyst (BCBA) who builds a session this way is not skipping ABA. They are sequencing it.

What a Connection-First Session Actually Looks Like#

A connection-first session looks boring from the outside. A clinician sits on the floor near the client. There is no token board. There is no SD. There is no "first this, then that." The clinician is watching. What does the client pick up? What do they put down? What do they laugh at? What do they walk away from? That is the data for the day.

Valerie ran sessions like this every day with a student while she was still waiting for consent on a practical functional assessment. She did not call it intervention. She called it daily rapport building. She sat with him, watched what he picked up, watched how he liked to interact, and let the work of the day be that.

It took a very long time for us to get consent to conduct that practical functional assessment with the student. But while we were waiting, I was like, I'm not waiting. Like I started taking the student every day for daily rapport building sessions. I just wanted to see what he was interested in, how he liked to interact. From the talk. Valerie via Dr. Megan DeLeon

A session like this is not a vibe. It is a probe. The clinician is screening for a few specific things. Does the client seek the adult out at all? Does the client tolerate the adult moving closer? Does the client share a glance, a sound, an item? If the answer to most of those is no, the session stays at this level. The session is the assessment.

The Three Phases: Connection, Skill, Independence#

Dr. Megan DeLeon's model has a spine. There are three phases, and they tell a BCBA where in the arc a client sits and what the next clinical step is.

Phase one is developing that connected relationship. So you're increasing engagement with adults and other items. In phase two, we can start working more on skill development around increasing duration and join a wider array of items. Phase three is independent. So increasing independence with enjoyment and engagement without the adult present. From the talk — Dr. Megan DeLeon

Phase one is the part most plans skip. The goal is engagement with people and with items. The clinician is not adding demands. The clinician is trying to be a person the client wants in the room. The data is interaction time, items shared, and approach behavior. When those numbers move, the client is ready for phase two.

Phase two is where skill goals start to show up. Duration of joint play goes up. The set of items the client will share gets wider. Targets enter the plan, but they are stitched into the play, not stacked on top of it. The clinician is still following the client's lead more than they are leading.

Phase three is the goal most parents and adult clients name when you ask them. Engagement and enjoyment without the adult having to drive it. A kid who plays with a sibling. An adult who picks a leisure activity and stays with it. The clinician fades out on purpose.

Naming the phases on paper matters. It tells a team, a family, and a funder where the client is and what the next step looks like.

Joint Activity Routines (Borrowed From ESDM)#

A joint activity routine is the specific clinical move that powers a phase-one session. The name comes from the Early Start Denver Model (ESDM), and it is worth saying out loud because it gives the BCBA a procedure with a citation, not just a feeling.

Joint activity routines are described in Early Start Denver model. If you're not familiar with it, I highly recommend checking that out. Joint activity routines are a wonderful way to follow someone's lead, let them be on their agenda, and then connect with them and pull in some of other ways to share an agenda. From the talk — Dr. Megan DeLeon

The routine has a shape. The client picks an activity. The clinician joins it. The clinician adds a small variation that the client can accept or reject. The clinician follows the client's response. Over time, the activity becomes a shared thing, with a beginning, a middle, and an end that both people know.

What this gives a BCBA is a teachable unit. A supervisee can practice it. A parent can practice it. The data point is whether the variation got accepted and how long the joint piece lasted. It is not vibes. It is a routine.

Imitation With Your Own Set of Materials#

There is a small move inside joint activity routines that deserves its own paragraph because it is the one thing supervisees get wrong most often. The clinician imitates the client. But the clinician does it with their own materials.

A clinician sat with a kid who liked wooden, neutral colored blocks. She did not reach for his blocks. She brought her own set, colored ones, and started building near him. When the kid noticed, she imitated what he was doing. Knock one over, she knocks one over. Build a tower, she builds a tower. Same activity, separate piles.

Two reasons this matters. First, the client does not have to share to be in the activity. The pressure point that ends most sessions early is "give me one." If the clinician has their own set, that pressure point never shows up. Second, imitation builds approach. Humans, of any age, tend to like being around people who do what they do. The clinician is teeing up the very thing they want to measure.

A working clinical note from a session like this reads: "Brought a parallel set of blocks. Built near client for 8 minutes. Client glanced 3 times, smiled twice, no protest. Variation: I knocked mine over and laughed. Client knocked his over and laughed."

When to Add the First Demand (and What Counts as a Demand)#

The most common question a supervisee asks after watching a phase-one session is when to add the first demand. The honest answer is later than feels comfortable. The clinical answer is when the phase-one data tells you so.

A demand is anything the clinician puts in front of the client that the client did not choose. A worksheet is a demand. A request to point is a demand. A request to wait is a demand. A request to clean up is a demand. If a BCBA does not call those demands, the plan will drift into phase two before the client is ready.

The signal that phase one has done its work is approach behavior. The client comes toward the clinician on their own. The client offers an item. The client tolerates the clinician moving closer without protest. When those data points are stable across two or three sessions, the first demand can go in. The first demand should be small, embedded in an activity the client already likes, and easy to back off if the client protests.

If the client protests, the response is not to hold the line. The response is to pull the demand, sit back down in the joint activity routine, and try again the next day with something smaller. That is the part the old ABA training would call a mistake. In this model, it is the protocol.

Two Case Sketches: A Preschooler and a 30-Year-Old#

The case for this approach is that it works at any age. Two sketches from the talk make that point.

The first is a preschooler whose team was still waiting on consent for a practical functional assessment. Valerie did not wait. She ran daily rapport sessions, watched what the student picked up, and let interaction be the day. By the time consent came through, she already had a list of preferred items and a baseline of approach behavior. The assessment was easier to run because the student already knew her.

The second is David, a 30-year-old. Alana came in ready to work on skill development. Dr. Megan DeLeon pulled her back.

Megan was like, okay, let's back up a little bit. Let's just work on this connected relationship first. So we really looked at choice assessment and just observing what David would do throughout the day. From the talk. Alana via Dr. Megan DeLeon

Same move. Watch what the person picks. Use that as the floor of the relationship. Build joint activity routines from there. The kid had blocks. The adult had something else. The clinical move was the same.

This is the section to keep in mind when a team or a funder pushes back with "he's too old for that." Connection is not a developmental stage. It is a service-delivery sequence.

How to Write This Into Your Treatment Plan So Insurance Reads It#

A treatment plan that names phase one as "rapport building" tends to get pushed back. A treatment plan that names phase one as a clinical phase with measurable objectives tends to get approved.

Three small writing moves help.

Name the model. Cite Dr. Megan DeLeon's three-phase progression and joint activity routines from the Early Start Denver Model (ESDM). A funder who sees a named model with a citation reads the rest of the plan as clinical, not as missed sessions.

Write phase-one objectives in data terms. "Client will independently approach the clinician within 30 seconds, across 3 of 5 opportunities, for 3 consecutive sessions." That is a target. It is also exactly what the clinician is doing on the floor.

Tie phase one to a phase-two trigger. Say in writing what data will move the client to phase two and what the first phase-two target will be. That tells the funder this is a sequence, not a stall.

A plan written this way reads as ABA, because it is. It just sequences connection before demand.

Frequently asked questions#

Can I run an ABA session with no demands at all?

Yes, for a phase-one session, and the plan should say so. The session has a clinical purpose: screen for approach behavior, build joint activity routines, and watch what the client picks when nothing is being asked. The data are interaction time, items shared, and protest behavior. A session like this is not a missed session. It is the assessment for phase two.

When do I add demands back in after connection-only sessions?

When the phase-one data show stable approach behavior across two or three sessions. The client comes toward the clinician on their own. The client offers items. The client tolerates the clinician moving closer without protest. The first demand should be small, embedded in an activity the client already likes, and easy to pull back if the client protests. Pulling back is the protocol, not a mistake.

Is connection-based behavior intervention the same as naturalistic teaching?

They overlap, but they are not the same. Naturalistic teaching is a teaching style that embeds targets in everyday activities. Connection-based behavior intervention is a service-delivery sequence with three phases, and phase one has no targets at all. Naturalistic teaching shows up most in phase two, once the connected relationship is already there.

Watch the full talk#

Dr. Megan DeLeon walks through the three-phase progression, the assessment skills that signal a client needs phase one, the choice assessment and SORF as screening tools, and case sketches across ages from preschool to adult. If you write treatment plans that fund relationship work, this is the talk to send your team and your supervisor before the next plan goes out.