Behavioral Experiment
A behavioural experiment turns an anxious prediction into something you can test instead of debate: state it precisely, rate your confidence, design a fair trial run, and record what actually happens. This worksheet carries you through the full cycle, including what you learned.
Test it, don't debate it
Anxious thoughts are usually predictions: if I speak up, I'll be laughed at. If I ask, they'll say no and things will be weird forever. You can argue with a prediction in your head all day and end exactly where you started. A behavioral experiment ends the debate by collecting data. You treat the prediction as a hypothesis, run a small real-world test, and let the outcome speak. There's a reason this works where arguing doesn't: avoidance. Every time you dodge the feared situation, the prediction goes untested - and untested predictions keep their power. An experiment is how a prediction finally meets the evidence.
What makes a test fair
A fair experiment has three features. The prediction is specific: not 'it will be awful,' but what you would actually see and hear if it came true - people laughing, a flat no, someone walking off. The test is doable soon: small enough to run this week. And the criteria are set in advance: decide now what result would count for the prediction and what would count against it. One more thing decides everything: no rigging. If you run the test wrapped in protections - over-rehearsing, escaping early, waiting for a perfect day - any good outcome gets credited to the protection instead of counting against the prediction. You'll plan those out below.
The prediction
Write the prediction you're going to test, in if-then form, specific enough that you'd recognize on sight whether it came true.
Confidence before the test
How likely does the prediction feel right now? Rate the felt certainty, not what you logically know - the gap between those two is exactly what this experiment measures.
Design the experiment
Plan the test. What exactly will you do, and when and where? What will you watch for while it happens - the specific signs from your prediction? And what outcome would count as the prediction being wrong? Settle that last one now, before your mind gets a chance to move the goalposts.
Rig check: safety behaviors
Safety behaviors quietly ruin experiments - whatever goes well gets credited to them. Check any you'd be tempted to use, then run the test without them, or at least without the biggest one.
Results: what actually happened
Fill this in soon after the test, while the details are fresh. Report like a camera: what did you do, and what did people actually say and do? Then set the outcome next to the prediction, line by line - which parts came true, and which didn't?
Confidence after the test
Knowing what you know now, rate the original prediction again. If the number barely moved, look for a safety behavior that snuck in and took the credit - that usually means the next run needs to be cleaner.
What the data says
What does this result tell you about the prediction - and about the bigger rule behind it, like 'people are waiting for me to slip up'? If the feared thing did happen, what did you learn about how you handled it? Name your next experiment, one notch bolder than this one.
Fewer clicks. More care.
One place for your team, your clients, and every clinical workflow, so nothing quietly slips through the cracks.
Free for practices with up to 20 clinicians.