Fellow readers,
Touch a hot stove once, and you don’t touch it again. Nobody has to explain thermodynamics to you. The pain shows up, the behavior changes, done. That’s a direct-acting contingency — your own behavior, your own consequence, no middleman required.
But you’ve also never touched a running table saw, and you’re probably not planning to. You didn’t have to cut yourself to learn that lesson. Someone told you “that thing will take your fingers off,” you understood the rule, and your behavior changed without a single direct consequence ever touching your skin. That’s rule-governed behavior — verbal behavior doing the work that direct experience would otherwise have to do.
Both are real. Both shape behavior. But we don’t always get to choose which one is running the show — and in ABA right now, I think we’ve let the wrong one drive.
The Problem
Here’s the uncomfortable question: when your treatment plan gets approved, what did you actually learn?
If you’re honest, probably just this — a person, on a deadline, made a decision about your submission. That’s it. That’s the whole event. And yet watch what BCBAs do with that event: they treat approval as validation. The plan got through, so the plan must be good. No pushback, no clarification request, no denial — so we must be doing this right.
That’s a direct-acting contingency doing its job a little too well. Approval feels good. Denial feels bad. Behavior follows the feeling. The problem is that “got approved” was never actually contingent on “is clinically sound.” It’s contingent on something else entirely — whether a reviewer had time to read it in its entirety, or with any real scrutiny.
Let’s do the math on that. Say a reviewer gets 15 new plans on Monday. They get through 10. The other 5 roll into Tuesday, on top of Tuesday’s fresh 15 — now they’re facing 20. Wednesday, same thing. By Friday, that reviewer isn’t reading 15 plans with real attention; they’re triaging a pile that’s tripled, against a clock that hasn’t moved. Some of those plans get read closely. Some get glanced at. Some get what our field politely calls “administratively approved” — a decision made because a decision was due, not because the plan was evaluated against medical necessity criteria line by line.
Here’s the thing: from the outside, an administratively approved plan and a rigorously reviewed plan look identical. Same portal, same status, same green checkmark. You have no way of knowing which one you got. And if you’re only responding to the direct-acting contingency — approved feels good, so I must be doing well — you have no reason to ever ask.
Why This Happens
I don’t believe this is about BCBAs being lazy or uncritical. I believe it’s about which contingency is loudest. Direct-acting contingencies are immediate, concrete, and hard to ignore — the portal updates, the dopamine hits, the to-do list gets shorter. Rule-governed behavior — “my plan should reflect behavioral cusps and social significance, document outcomes accurately, and name and adjudicate barriers honestly, regardless of what the payer says about it” — takes deliberate verbal effort to keep in view. It’s abstract. It doesn’t ping your phone.
And reviewers aren’t the villains here either. They’re operating under their own direct-acting contingency: a queue that doesn’t stop growing and a deadline that doesn’t move. Nobody wakes up planning to skim. The system just doesn’t leave room not to, once the backlog snowballs past what any one person can meaningfully read in a day.
So we end up with two groups, each responding rationally to the contingencies actually in front of them — and neither of those contingencies has much to do with whether a five-year-old’s treatment plan is any good.
The Real Cost
When approval becomes the proxy for quality, quality stops being measured at all. Nobody’s watching it. The plan that names a target behavior nobody actually cares about, the goal that isn’t developmentally appropriate, the sequencing that ignores a genuine behavioral cusp — none of that shows up in an approval status. It shows up months later, in a kid who isn’t progressing, in a family who’s losing trust, in an RBT running programs that were never sound to begin with.
And there’s a field-wide cost too. Every time we treat “it got through review” as evidence of quality, we quietly outsource our clinical judgment to a process that was never designed to make clinical judgments. We let a documentation deadline decide what our profession considers good work.
The Solution: Build Your Own Rule
You cannot control what happens on a reviewer’s desk on a Friday. You can control whether you’re relying on their glance or your own standard.
Here’s what that looks like in practice:
1. Run the “fridge-worthy” test before you submit — not after you hear back. Before a plan goes out the door, hold it against your own explicit criteria, the same way you’d decide whether a kid’s drawing earns a spot on the fridge — not because someone approved it, but because you looked at it and it was actually good: Is this goal grounded in research? Does it target a genuine behavioral cusp? Is it developmentally appropriate and socially significant for this specific learner? If you can’t answer yes to all three, approval status is irrelevant — the plan wasn’t fridge-worthy, and it wasn’t ready.
2. Separate two questions you’ve been treating as one. “Did this get approved?” and “Is this good?” are different questions with different answers. Start asking both, out loud, every time. Put them on your QA checklist as two separate line items, not one.
3. Treat approval as neutral, not as feedback. An approval tells you a decision was made under a deadline. It does not tell you the plan was read closely, and it definitely doesn’t tell you it was clinically sound. Stop reading it as a compliment.
4. Build in your own review cadence, independent of the payer’s clock. If the only eyes ever on your treatment plan belong to someone racing a Friday deadline, that’s the only quality control you have. Add a second one — a peer review, a supervisor spot-check, a structured self-audit — that isn’t tied to anyone else’s backlog.
None of this requires waiting on anyone else to fix the review process. It requires deciding that your own rule is the one you’re actually going to follow.
The Vision
Imagine a field where a BCBA’s confidence in a treatment plan has nothing to do with whether it got approved — because they already know it’s good, on their own terms, before it ever reaches a reviewer’s desk. Where “administratively approved” is a phrase we say out loud without shame, because we’ve built quality checks that don’t depend on someone else’s Friday. That’s not a fantasy. That’s just what happens when rule-governed behavior gets to do the job it was always capable of doing.
Your turn:
• When’s the last time a plan of yours got approved, and you assumed that meant it was good?
• What would your own rule — the one you’d hold a plan to regardless of payer response — actually say?
• What would it take to build a second layer of review into your process, one that isn’t racing anyone’s deadline?
Ready to dive deeper into quality improvement? Join our free community here.
With quality in mind,
Kristen
