The REFINE™ framework
Same behavior.
Different possible drivers.
Two children can do exactly the same thing for entirely different reasons. We cannot know why something is happening by looking at the behavior alone — and the support we choose is only as good as our understanding of what is driving it.
REFINE is a reasoning tool — not a diagnosis.
The observation trap
The sentences we reach for first close the question instead of opening it.
- 01
Label
A word stands in for a description.
- 02
Assumption
The label implies why it is happening.
- 03
Mismatched response
We design a support for the wrong driver.
The goal is not to remove judgment. It is to delay conclusions long enough to improve them.
Try it
Pick a behavior you recognise.
Each of these has several genuinely different possible drivers — and each driver needs a different kind of support.
Choose one above to see what could be underneath it.
The six stages
Six stages, in order, every time.
Each stage has one job and one guiding question. Open any stage for a little more, including an example.
- R
RECOGNIZE
Describe what you actually see and hear.
What is happening?
- E
EXPLORE
Gather context, patterns, demands, and possible contributors.
What else might be influencing it?
- F
FOCUS
Narrow to the strongest one or two working hypotheses.
What seems most likely?
- I
IMPLEMENT
Choose and try a support matched to the hypothesis.
What will we try?
- N
NOTICE
Watch for meaningful changes in behavior, effort, regulation, and participation.
What do we see?
- E
EVALUATE
Keep, adjust, combine, refer, or rethink.
What's next?
EVALUATE returns to RECOGNIZE with a better description. The cycle is the point.
REFINE in action
You will not find a screen called “stage three.”
REFINE is reasoning that runs through the work, not a menu you click through. You meet the stages naturally: while describing what you are seeing, while answering questions that widen the picture, while reading a profile that narrows it to the areas most worth exploring.
Then again while choosing a support and trying it at home, while watching what actually changes, and while deciding whether to keep going, adjust, or rethink the whole idea. The same six letters appear quietly on those screens so you can see where you are — nothing more.
Want to go deeper?
There is more to each stage than one page can hold.
Observation versus interpretation, the wider map of possible contributors, matching supports to a hypothesis, the measures of meaningful change, and what to do when the first design does not work. We are building that teaching out stage by stage with the educators piloting this framework.
Educational and developmental exploration tool. Not a medical or clinical diagnosis. This describes patterns in what you observed — it does not identify a condition.