Why your predictions are probably wrong — and how to test them

What is a behavioural experiment?

A behavioural experiment is exactly what it sounds like: an experiment you design and carry out in your own life to test whether a belief holds up against evidence. The difference between this and ordinary experience is intentionality — you formulate a specific prediction, create the conditions to test it, and observe the outcome as objectively as you can.

This matters because anxiety and low mood are maintained by beliefs that often go untested. We avoid the situations that would challenge them, and so those beliefs persist unchallenged. A behavioural experiment deliberately creates the data point that avoidance would prevent.

The structure of a behavioural experiment

Step 1 — Identify the belief to test

Begin with a specific belief, not a vague one. "If I speak up in a meeting, everyone will think I am incompetent and I will be embarrassed" is testable. "I am bad at socialising" is not.

Step 2 — Make a precise prediction

What exactly do you expect to happen? How likely do you think this is, on a scale of 0–100%? Writing this down before you carry out the experiment is essential — memory is unreliable, and without a written prediction, it is easy to shift the goalposts afterwards.

Step 3 — Design the experiment

What will you actually do? Be specific about the situation, the timing, and what you will and will not do. If you usually use safety behaviours in this situation, the experiment involves going in without them — or at least reducing them significantly.

Step 4 — Carry it out and record what happens

Observe the outcome as objectively as possible. What happened? What did other people actually do or say? What did you notice in yourself? Try to record this as close to the event as possible, before reflection and interpretation cloud your memory.

Step 5 — Compare prediction with reality

How accurate was your prediction? If the outcome was different from what you expected, what does that tell you about the belief you were testing? What would you conclude if this were a scientific experiment?

Why this is so effective

Talking about a belief in therapy can shift it intellectually. But the shift rarely feels real until the body has the direct experience of surviving — and often thriving — in a situation it was certain was catastrophic. Behavioural experiments create that direct experience.

Over time, a personal database of evidence accumulates: real situations in which feared outcomes did not materialise. This is far more convincing than any reassurance anyone else could offer, because it is your own experience, not someone else's opinion.