How to Turn a Worry Prediction Into a Small Behavioral Experiment

A behavioral experiment is not an adult proving that a child’s worry is silly. It is a planned way to gather new information about a specific prediction. It should be collaborative, proportionate, and safe.

In brief: Write the prediction precisely, rate expected likelihood, choose the smallest informative test, observe what actually happens, and update the prediction without demanding that anxiety disappear.

Make the Prediction Testable

“School will be awful” is too broad. Ask what the child expects to happen: “If I ask the teacher a question, everyone will laugh for the rest of class.” Record the predicted outcome and, when appropriate, a 0–100 likelihood estimate.

Do not argue with the estimate. Curiosity is more useful: “What could we observe that would give us more information?”

Design the Smallest Useful Test

The test might be asking a prepared question after class, then during a quiet work period, before trying it in a larger discussion. Agree on what counts as evidence. Notice not only whether someone laughs, but how many people respond, how long the reaction lasts, and whether the child can recover.

The experiment must not involve genuine danger, humiliation, deception, or withholding necessary support. High-risk fears require qualified assessment—not a DIY test.

Review All the Data

Afterward, compare the prediction with observations:

What happened as predicted?
What happened differently?
What helped the child cope?
What remains uncertain?
What would be a fairer prediction next time?

An outcome can be mixed. Someone may glance over, the child may feel embarrassed, and the feared catastrophe may still not occur. The lesson is not “nothing bad ever happens”; it is “we can test predictions and respond to real outcomes.”