The behavioural components of Cognitive Behavioural Therapy

As the name would suggest, Cognitive Behavioural Therapy (CBT) consists of techniques to help us to develop more adaptive and healthy cognitions and behaviours with a recognition that what we think and what we do are deeply connected.  An important part of the therapeutic process is the performance of ‘behavioural experiments’.

Behavioural experiments are planned activities based on experience, experimentation and observation, undertaken in and between CBT sessions. Their design is derived from a cognitive formulation of the problem and their aims are to test the validity of our existing beliefs (about ourselves, others and the world), construct and test new, more adaptive beliefs, and contribute to the development and verification of the shared cognitive conception of the problem.

Behavioural experiments are amongst the most powerful tools available in CBT. They enable us to actively engage in the process in a way that purely verbal methods cannot: they allow us to see, hear, feel, taste, smell and touch, creating a much more effective method of learning and a much higher probability of positive change, whether for depression, anxiety, eating issues, sexual concerns or addictions.

By becoming empowered to take breakthroughs in the therapeutic setting home, we become equipped with the tools to tackle our problems not just in the context of therapy, but also in ‘the real world’: this is an important reason why CBT is consistently shown to yield lower relapse rates than medication or other therapies: the short, solution-focused CBT sessions essentially train us to become our own therapist.

The precise nature of the behavioural experiments that a therapist might use are entirely dictated by the specific needs and requirements of the client: experiments are shaped around the thoughts and feelings particular to that person, though there are common techniques used for each type of presenting complaint.

Those who are feeling low or depressed for example often have trouble engaging with their daily activities and duties: they may cancel engagements with friends or skip work because they don’t anticipate that anything positive can come from their day. A task useful here is activity scheduling (monitoring and modifying activity levels by keeping  diary of activity and planning future activities).

Activity scheduling can be a rich opportunity for planning behavioural experiments, but also can be an experiment in itself: it can be particularly useful for disproving a person’s belief that they are ‘useless’ or ‘not achieving anything’ because. by keeping an hour by hour record of what we have been getting up to, we can often realise that we have achieved more than we thought, even if only small tasks.

Though the specific types of behavioural experiment available are as varied and diverse as the people who attend therapy, behavioural experiments can be thought to fall into two broad types: active and observational.

Active experiments are the most common and involve the person themselves taking the lead role. Once they have identified with the therapist an unhelpful or problematic cognition or pattern of behaviour, they are tasked with deliberately thinking or acting in a different way in a real life situation. Afterwards, often during the therapeutic session with the help of the therapist, they are encouraged to note what happened and reflect on its implications for their future thoughts and behaviour.

Sometimes the real-life situation to test in may be rare, or the person may need practice in a more controlled and comfortable environment: in these cases role play is often used with the therapist, so that the person is equipped to handle future relevant situations as they arise.

Observational experiments involve the person as a data gatherer rather than an actor i.e. they are not directly involved. They, like role playing active experiments, are particularly useful when the thought of direct action is too anxiety-provoking (as with phobias), or when more information is required before progressing to an active experiment.


The House Partnership, 31st March 2012