Behavioral Experiment

Behavioral experiments are a particularly powerful technique and are very commonly used in CBT. There are many different types of behavioral experiments which serve different purposes. A non-exhaustive list include:

  • Surveys - may be purely information-gathering
  • Experiential exercises - may allow the client to test specific beliefs (e.g. "if I carry out an interoceptive exposure exercise of hyperventilating I will pass out and die")
  • Hypothesis testing - may be designed to allow clients to gather information to test the validity of thoughts, predictions, or beliefs.

In essence, as long as a belief is clearly specified the evidence supporting it can be explored experimentally. Tests (experiments) are designed to provide clients with new information. This information may serve to confirm or disprove beliefs, or offer support in favour of one hypothesis over another.

This Behavioral Experiment CBT worksheet guides you through the essential steps required to plan and evaluate behavioral experiments. It is well suited for hypothesis testing / planned experiments ("If I do X then Y will happen") but can equally be used for data gatheing ("If I ask X then I will learn Y").

  1. Specify the client's prediction / belief / hypothesis / thought as clearly as you can "If... then..." statements often work nicely
  2. Design an experiment to test the prediction. Think about what circumstances would need to be in place for the prediction to be tested. Where would the client need to go? What would they need to do or observe? What safety behaviors might they need to minimise in order for it to be a fair test? How many times might the experiment need to be repeated in order to obtain reliable result? Think in advance about what you need to observe. What kinds of results will you need to record? Observable behaviors? Always think safety - do you need a graceful plan of retreat?
  3. Carry out the experiment
  4. Record your observations. What did you notice?
  5. Reflect upon you observations. Was the original prediction supported by the results of the experiment? Is there another belief which might account better for the results that you observed?
  • Bennett-Levy, J., Butler, G., Fennell, M. J. V., Hackmann, A., Mueller, M., & Westbrook, D. (Eds.) (2004). The Oxford handbook of behavioural experiments. Oxford: Oxford University Press.

Other resources


free & pro