CBT week 4: behavior experiments

This is week 4 of our six-week, self-guided course in cognitive behavioral therapy (CBT). Last week you tested a thought on paper by weighing the evidence for and against it. This week you will test a thought in real life.

A behavior experiment is a small, planned action that checks whether a worried prediction comes true. You act a bit like a scientist. You write down what you expect, try it, and then compare your guess with what happened.¹ Research suggests that doing something can change a belief more than only thinking about it.⁶

Why this step matters

When we feel anxious or low, we tend to expect the worst. We guess that bad things are more likely than they are, that they will be worse than they turn out to be, and that we will not be able to cope.¹

Then we avoid the situation or leave early. That brings quick relief, but we never find out whether the fear was right.¹ Over time, avoided fears tend to grow.⁸

A behavior experiment breaks that loop with fresh, first-hand evidence. In a small study, CBT therapists in training tried both methods on themselves. They rated behavior experiments as more powerful for change than written thought records.⁶ Testing unhelpful beliefs and changing unhelpful actions are both core parts of CBT.^10,11

Evidence: CBT for anxiety disorders Strong evidence¹⁰

Predictions versus results

A useful prediction is specific. "It will go badly" is too vague to test. "My friend will sound annoyed and hang up within a minute" can be checked.¹ Decide ahead of time what you would see or hear if it came true.⁴

You also write an alternative prediction, a more balanced guess, and rate your belief in each from 0 to 100.¹ If belief in the old one drops after the experiment and the new one rises, your thinking has shifted.¹

Safety behaviors: how fears stay put

Safety behaviors are small things people do to feel safe or to stop a feared outcome. Examples include staying quiet in a group, always carrying water "in case," or wearing headphones so no one starts a conversation.³

These habits backfire. You never get a fair test of the fear. If nothing bad happens, you may credit the habit instead of your own coping. Some habits even bring on what you dread: staying silent can make you seem distant.³

A simple check: would you feel anxious if you could not do the habit? If yes, it is probably a safety behavior. If anxiety stays high after several tries, look for one you are still using.³ Part of planning an experiment is choosing which safety behaviors to drop.¹

How to plan and run an experiment

1. Pick a thought to test. Choose one from your week 2 or 3 notes that still feels true.
2. Write the prediction. Say exactly what you expect to happen. Rate how much you believe it, 0 to 100.¹
3. Write an alternative prediction. What might happen instead? Rate that belief too.¹
4. Design the experiment. Decide what you will do, when, where, for how long and with whom.^1,2 Start with a step that feels only a little hard, not your hardest one.²
5. Name the safety behaviors to drop. List the habits you usually lean on. Plan to do without at least one of them.^1,3
6. Plan for problems. What might get in the way? What will you do if it does?¹
7. Do it and notice. Carry out the plan. Pay attention to what really happens around you, not only to your feelings.¹ Some anxiety is normal and tends to ease if you stay.²
8. Write down the results. Record the facts. Did your prediction come true, partly come true or not come true?^1,4
9. Ask what you learned. Re-rate both predictions. Write one sentence about what this tells you.¹
10. Repeat and step up. Do the same experiment again soon, then try a slightly harder one. One good result can feel like luck. Several build real evidence.²

A worked example

This is a made-up example to show the steps.

Thought: "My old friends have moved on. If I call one of them, I will be a bother."

Prediction: "When I call Ann, she will sound rushed and get off the phone within a minute." Belief 80 out of 100.

Alternative: "Ann may be busy, but she will probably be glad to hear from me." Belief 20.

Experiment: Call Ann on Tuesday afternoon and stay on at least five minutes if she can talk.

Safety behaviors to drop: Not apologizing for calling. Not making up a "practical" reason for the call.

After: Belief in the old prediction 25. Belief in the alternative 70.

What I learned: "People can be glad to hear from me, even after a long time. I will call another friend next week." If Ann had been busy, that would also be useful. One result is one piece of evidence, not a final verdict.

Your worksheet

Copy this onto paper, or add your experiment to the free thought check tool.

Ideas for small experiments

Some experiments bring on body sensations on purpose, such as a racing heart or dizziness. If you have heart or lung problems, seizures, past fainting, neck or back injuries, or are pregnant, do these only with guidance from a doctor or therapist.⁵ If checking or washing urges are strong, see OCD.

Notes for later life

CBT works at any age. The American Psychological Association recommends group CBT for depression in older adults.⁹

Later life brings real changes, such as hearing loss or health problems. These can lead to fears like "People will get impatient if I ask them to repeat things." A small experiment, such as telling one person "I hear better if you face me," can show how others really respond. See anxiety in later life and depression in later life.

Some people become afraid of falling and start to avoid going out. Please do not test this fear on your own. A doctor or physical therapist can help you build up activity safely. See phobias.

If planning feels like a lot, keep it simple: one prediction, one small action, one sentence about what happened. A helper can write it down.

Common problems

What if the thing I feared really happened?

That is still useful. Was it as bad as you expected? How did you cope? Often the result is milder than the prediction, and you handled it better than you feared.¹ If a problem is real, you can move on to problem solving or ask for support.

I got too anxious and backed out.

That is common. It usually means the step was too big. Make a smaller "bridge" step, or repeat an easier one as a refresher, then try again.² Use your week 3 skills to check the thoughts that came up.²

Should I do this alone?

Many people can run small experiments alone. A therapist can help with harder ones.⁶ If fear or low mood keeps getting in the way of daily life, see how to find a therapist. Keep taking any medicine as prescribed, and talk with your doctor before changing anything.

Next step

In week 5: deeper beliefs, you will look at the bigger beliefs that sit under many of your automatic thoughts, and how to work with them gently. You can go back to week 3 or the CBT course overview any time.

When to get help

Education only. This page is general information written from the sources listed. It is not medical, legal or financial advice and does not replace a doctor, therapist or lawyer who knows your situation. How we write and check pages.