Behavioral experiments: test depressive predictions with action

Make a specific prediction ("going out will feel terrible") and test it with the actual behavior.

Why it works

Depression generates systematically negative predictions: "I won’t enjoy it," "I’ll embarrass myself," "there’s no point." These predictions guide avoidance and prevent the corrective experience that would disconfirm them. Behavioral experiments treat these predictions as hypotheses to be tested, not facts. The experiment provides direct experiential evidence against the prediction — evidence that is harder for the depressive mind to dismiss than verbal argument.

How to do it

1. Identify a specific depressive prediction: "If I go to that gathering, I will feel worse."
2. Write the prediction exactly, with a specific outcome you can check.
3. Do the activity — the one you’d otherwise avoid based on the prediction.
4. Immediately after, write what actually happened, rated on the same scale as the prediction.
5. Over time, build a personal record of prediction accuracy — most people are surprised to find depressive predictions are systematically more negative than reality.

Evidence

Behavioral experiments are used in both CBT and BA as a primary mechanism for belief change. Research on cognitive disconfirmation in depression consistently finds that experiential disconfirmation is more effective than verbal/logical disputation of negative beliefs. (clinical)

Behavioral experiments work best for mild-to-moderate avoidance and negative prediction. For very severe depressive apathy or psychomotor retardation, the capacity to execute experiments may be limited — this is not a character flaw, but a symptom requiring clinical support.

Common mistake

Confirming the prediction because the activity was imperfect ("it was okay but not great, so my prediction was basically right") — the baseline during depression is very low; "okay" disconfirms "terrible," even if it’s not "wonderful."