Run uncertainty experiments

Deliberately leave small things uncertain and observe what actually happens.

Why it works

Reassurance-seeking and checking are maintained by negative reinforcement: they produce short-term relief that prevents the disconfirmation of catastrophic predictions about uncertainty. Exposing yourself to real uncertainty — and not checking — tests whether the feared outcome (things go wrong, you cannot cope) actually occurs. Repeated disconfirmation erodes the belief that uncertainty is dangerous.

How to do it

1. Identify a low-stakes uncertain situation you would normally check or seek reassurance about.
2. State your prediction: "If I don’t check, _____ will happen."
3. Refrain from checking for an agreed period.
4. Record what actually happened and compare to your prediction.

Evidence

CBT protocols for GAD that include behavioral experiments targeting IU show significant reductions in worry and GAD symptoms in controlled trials. (rct)

Effects are strong for GAD; generalization to normal-range worry is plausible but less directly studied.

Sources

• Dugas & Robichaud (2007), Cognitive-Behavioral Treatment for Generalized Anxiety Disorder — protocol evidence summary
• van der Heiden et al. (2012), RCT of uncertainty-based CBT for GAD, Journal of Consulting and Clinical Psychology

Common mistake

Selecting experiments that are too high-stakes initially — the brain’s threat system overwhelms the learning before the result can be registered. Start with genuinely trivial uncertainties.