Behaviour Change Technique

How a behaviour change technique contributes to an intervention

A behaviour change technique is a defined, observable and replicable component intended to alter behaviour or a process that influences behaviour. This page explains how techniques are identified and selected, how they differ from intervention functions and delivery methods, and how their implementation, effectiveness and costs can be evaluated.

A technique is one active component rather than the entire intervention. An intervention may contain several techniques delivered through different people, technologies and settings, and the same technique may produce different effects depending on its target, timing and context.

Defining the target behaviour

A technique should be selected for a clearly specified behaviour. Broad aims such as improving health or increasing engagement are not sufficiently precise to support replication or evaluation.

The target behaviour should identify:

• Who needs to perform the behaviour.
• What the person or group must do.
• When and how often it should occur.
• Where it should occur.
• The people or systems involved.
• The current and intended level of performance.

For example, increasing physical activity is broad, while walking for 30 minutes after work on at least five days each week defines a more observable target. A clear target also allows the intervention to distinguish initiation, maintenance and relapse.

Techniques, mechanisms and outcomes

A behaviour change technique is what the intervention delivers. A mechanism of action is the process through which the technique is expected to influence behaviour, and an outcome is the change that is measured.

For example:

• Technique: Goal setting for behaviour.
• Proposed mechanism: Increased intention, commitment or self-regulation.
• Behavioural outcome: Completion of the intended activity.
• Health outcome: A later change in symptoms, risk or wellbeing.

These levels should not be collapsed. Observing a behavioural change does not prove that the proposed psychological mechanism caused it, and a change in a mechanism does not guarantee a clinically meaningful health outcome.

Common behaviour change techniques

Standardised taxonomies give techniques shared labels and definitions so interventions can be described and compared. The relevant technique should be identified from what is actually delivered rather than from the name of the programme.

Examples include:

• Goal setting for behaviour or outcomes.
• Action planning.
• Self-monitoring of behaviour or outcomes.
• Feedback on behaviour.
• Review of behavioural goals.
• Problem solving.
• Instruction on how to perform the behaviour.
• Demonstration of the behaviour.
• Social support.
• Prompts or cues.
• Information about health consequences.
• Restructuring the physical or social environment.
• Graded tasks.
• Behavioural practice or rehearsal.
• Rewards linked to behaviour or outcomes.

Labels should be used with their controlled definitions. Describing an intervention as educational, motivational or digital does not identify the specific technique it contains.

Distinguishing techniques from intervention functions

An intervention function describes the broader way an intervention is intended to change behaviour, such as education, persuasion, training, incentivisation, enablement, restriction or environmental restructuring. A behaviour change technique describes the specific component used to deliver that function.

For example, a training function may use demonstration, rehearsal and feedback. An enablement function may use problem solving, action planning and social support.

One technique may support several functions, and one function may require several techniques. Keeping the levels separate improves intervention design and reporting.

Distinguishing techniques from delivery methods

The delivery method describes how a technique reaches the intended person. Face-to-face counselling, text messaging, an app, a group session and printed material are delivery modes rather than techniques.

An app may deliver self-monitoring, feedback and prompts, while an in-person session may deliver the same techniques through conversation and demonstration. Effects can still differ because the mode changes intensity, reach, personalisation, cost and engagement.

A complete intervention description should therefore state both the technique and how it is delivered.

Selecting techniques using behavioural diagnosis

Technique selection should respond to the factors preventing or supporting the target behaviour. Choosing familiar techniques without identifying the behavioural problem can add burden without addressing the main barrier.

Potential determinants include:

• Knowledge and understanding.
• Practical skills.
• Confidence and perceived capability.
• Beliefs about consequences.
• Goals and priorities.
• Memory and attention.
• Habit and automatic responses.
• Social norms and support.
• Physical opportunity and environmental constraints.
• Access to time, money, equipment or services.

Information about consequences may help when knowledge is the barrier but may have little effect when a person lacks transport, time or physical access. Environmental or service changes may be necessary before an individual-level technique can work.

Specifying a technique for replication

Naming a technique is not enough to make the intervention reproducible. The description should explain its content, target, dose, timing, provider and delivery process.

Specification should include:

1. Identify the technique using a recognised definition where applicable.
2. State the target behaviour and intended recipient.
3. Describe the content delivered to the participant.
4. Specify the provider and required training.
5. Define the mode and setting of delivery.
6. State the schedule, intensity and duration.
7. Explain tailoring rules and permitted adaptations.
8. Describe materials and technology.
9. Define fidelity measures for delivery and receipt.
10. Link the technique to its proposed mechanism and outcome.

The description should allow another team to distinguish the essential component from features that can be adapted locally.

Dose, sequence and combination

Behaviour change techniques can vary in frequency, duration and intensity. More exposure is not always better because repeated prompts, monitoring or counselling can create burden, disengagement or alert fatigue.

Techniques may also depend on sequence. Instruction may precede rehearsal, while feedback requires behaviour or performance data to exist first. Maintenance techniques may be introduced after initiation rather than delivered at the same intensity throughout.

When interventions contain several techniques, their effects may be additive, synergistic, redundant or conflicting. A package that changes behaviour does not reveal which component was necessary without an appropriate study design or process analysis.

Tailoring and personalisation

Tailoring adapts technique content or delivery according to participant characteristics, behaviour or response. It can improve relevance but can also reduce consistency and make replication difficult.

Tailoring rules should state:

• Which information triggers adaptation.
• Which components can change.
• Who or what makes the tailoring decision.
• Which components must remain consistent.
• How adaptations are recorded.

Unrecorded clinician judgement may be appropriate in practice but creates uncertainty about which intervention was evaluated. A flexible protocol should preserve both individual responsiveness and analytical traceability.

Fidelity and engagement

Fidelity concerns whether the intended technique was delivered as planned. Engagement concerns whether participants encountered, understood or used it. Neither should be assumed from programme availability.

Evaluation may distinguish:

• Design fidelity: The intervention materials contain the intended technique.
• Delivery fidelity: Providers or systems deliver the component as specified.
• Receipt: Participants understand or attend to the component.
• Enactment: Participants use the relevant skill or strategy in daily life.

Low effectiveness may result from an ineffective technique, weak delivery, low receipt or environmental barriers. These explanations have different implications for redesign.

Measuring behavioural outcomes

Outcome measurement should correspond to the defined target behaviour. Self-report can capture behaviours that are difficult to observe, while devices, records or direct observation may reduce some recall and reporting problems.

Measures may include:

• Frequency or duration of behaviour.
• Proportion of opportunities in which the behaviour occurs.
• Adherence or persistence.
• Time to initiation or discontinuation.
• Objective sensor or device data.
• Service or prescribing records.
• Validated self-report scales.

Measurement itself can change behaviour, particularly when self-monitoring is also part of the intervention. The evaluation should recognise when data collection functions as an active technique.

Evaluating effectiveness

An effectiveness study should compare the intervention containing the technique with a relevant alternative and measure both behavioural and health outcomes when appropriate. The comparator may contain similar techniques, so usual care should be described rather than treated as an empty condition.

Important questions include:

• Did the target behaviour change?
• Was the change maintained?
• Did the proposed mechanism change?
• Did the behavioural change improve health or service outcomes?
• Which groups benefited?
• Was the technique delivered and received as intended?
• Were there unintended consequences?

Mediation analysis may examine whether change in a proposed mechanism or behaviour explains later outcomes. It requires appropriate timing, measurement and assumptions and should not be treated as automatic proof of mechanism.

Factorial and optimisation designs

Multi-component interventions can be studied using factorial, fractional factorial, sequential or adaptive designs. These approaches may help estimate the contribution of individual components and identify an efficient intervention package.

Component evaluation should consider interactions. A technique that has little effect alone may be important when paired with another technique, while two effective components may duplicate one another.

Optimisation should consider cost, burden, feasibility and equity as well as average effect. The smallest package is not necessarily the most valuable if removing a component disproportionately reduces benefit for a high-need group.

Economic evaluation of behaviour change techniques

Economic evaluation can compare the additional costs and consequences of interventions containing different techniques. Costs should include development, training, delivery, monitoring, technology and participant time when relevant to the perspective.

For technique package (j), total programme cost can be represented as:

$$ C_j = C_{development,j} + C_{training,j} + C_{delivery,j} + C_{technology,j} + C_{monitoring,j} $$

Average cost per participant is:

$$ Average\ cost_j = \frac{C_j}{N_j} $$

where (N_j) is the number of participants reached. Average cost may decline with scale when development costs are fixed, but staffing, licensing and support costs may increase with participation.

The analysis should also measure downstream healthcare use and health outcomes. A low-cost technique is not good value if it produces little sustained change, while a more intensive technique may be cost effective when it generates meaningful long-term benefit.

Modelling long-term health outcomes

Behavioural studies may observe change for a shorter period than is needed for economic evaluation. Models can link behaviour to later health outcomes, but the relationship and duration of effect should be supported by evidence.

Key uncertainties include:

• How long behavioural change persists.
• Whether effects decay after support ends.
• How behaviour affects disease risk.
• Whether repeated intervention is required.
• Whether participants compensate through other behaviours.
• Whether trial engagement can be reproduced at scale.

Assuming that a short-term behavioural change continues indefinitely can substantially overstate health gain and cost effectiveness.

Equity and accessibility

A technique may be more feasible or effective for people with particular resources, literacy, language, disability access or digital skills. Average effectiveness can therefore conceal unequal reach or benefit.

Equity assessment should examine:

• Who is offered the intervention.
• Who enrols and remains engaged.
• Whether materials are accessible and culturally appropriate.
• Whether participation requires time, devices, transport or money.
• Whether tailoring addresses different barriers.
• Whether the technique increases burden on people with fewer resources.
• How effects differ across relevant population groups.

Individual-level techniques should not be used to place responsibility on patients for behaviours constrained by poverty, service access or environmental conditions.

A simplified example

Suppose an intervention aims to improve adherence to daily medicine. It uses goal setting, self-monitoring through an app and weekly feedback from a pharmacist.

The evaluation finds improved adherence during the 12-week support period but a decline after feedback ends. Process data show that participants continued self-monitoring less frequently once pharmacist contact stopped.

The finding suggests that the intervention package changed behaviour but does not show whether feedback, self-monitoring or their combination produced the effect. A future study could test a lower-cost feedback schedule, maintenance prompts or different component combinations while continuing to measure adherence after active support ends.

Common misunderstandings

A behaviour change technique is a component with defined content, not a general description of an intervention's purpose or delivery channel.

Common misunderstandings include:

• Education is an intervention function, not a sufficiently specific technique.
• An app is a delivery mode, not a behaviour change technique.
• Naming a technique does not establish that it was delivered with fidelity.
• More techniques do not automatically produce a stronger intervention.
• A change in knowledge does not prove that behaviour changed.
• A short-term behavioural effect does not establish long-term health benefit.
• Self-reported behaviour is not automatically invalid, but its limitations should be assessed.
• A multi-component intervention does not reveal the effect of each component.
• Individual behaviour should not be separated from environmental and structural constraints.

Interpreting evidence about a technique

The effect of a behaviour change technique depends on the target behaviour, population, delivery, dose, accompanying components and context. Evidence about a technique used in one intervention should not be transferred automatically to another setting.

A useful evaluation identifies the active component clearly, tests whether it was delivered and received, and traces the pathway from the proposed mechanism to behaviour and health. This allows decision makers to judge not only whether an intervention worked, but also what may be necessary to reproduce its effects at an acceptable cost.