Glossary
Behavioral experiments
What are behavioral experiments?
Behavioral experiments are structured activities that test a person's beliefs or predictions about what will happen if they change their behavior. They are used in cognitive therapy to collect evidence relevant to assumptions; the result can challenge, refine, or support a prediction.
How it works
Start with a prediction: ‘If I speak up in the meeting, people will think I am incompetent.’ Then specify what you could observe, try the action, and compare the observations with the prediction. You cannot directly observe what everyone thinks, so the test needs a clearer sign, such as the responses to your contribution. Consider alternative explanations too.
Behavioral experiment example
A person with social anxiety who believes 'Everyone will stare if I eat alone in a restaurant' designs an experiment: eat alone, count how many people actually look at them for more than two seconds. The exercise illustrates a testable prediction. Record what happens rather than deciding beforehand that nobody will look. A glance also does not, by itself, tell you what the other person thinks.
Why it matters
Behavioral experiments make a belief testable. In one study of cognitive-therapy trainees, participants rated experiments as more compelling than thought records. That self-practice study does not establish that experiments always produce more belief change than discussion in clinical care.
Sources and further reading
BCT Taxonomy v1 defines technique 4.4 as testing behavioral hypotheses through data collection and interpretation.