Behavioural diagnostic framework: meaning and application

A campaign can attract considerable attention while the intended behaviour remains unchanged. People may understand the message but lack time, skills or permission. A behavioural diagnostic framework helps investigate such explanations systematically. It starts with the action that does or does not occur in a particular situation, rather than the copy a team wants to produce.

A behavioural diagnostic framework is a structured way to investigate possible explanations for behaviour and select appropriate next steps. ‘Diagnostic’ means behavioural analysis here, not clinical diagnosis. The framework organises questions and evidence; it does not automatically establish the decisive cause. Describe behaviour, context, barriers and opportunities, then test the proposed explanation and intervention.

Michie, Van Stralen and West introduced the Behaviour Change Wheel, with COM-B at its centre: capability, opportunity and motivation in relation to behaviour. These components help organise questions about what enables an action. Michie, Van Stralen and West, 2011

The Theoretical Domains Framework offers a more detailed structure. Cane, O’Connor and Michie investigated its domain structure with experts and distinguished fourteen domains. This research does not validate every subsequent application or intervention based on the framework. Cane, O’Connor and Michie, 2012

‘Behavioural diagnostic framework’ is used here as a general description of the approach. A proprietary combination of audience segments, associations and measurement tools requires its own description and evidence. Using an established model does not automatically make the entire bespoke method scientifically validated.

This table is a practical simplification, not a complete COM-B questionnaire. The perspectives are not independent boxes that never interact. Someone can consider an action important and still not know how to perform it. Use the structure to identify missing questions, rather than to assign a single cause prematurely.

An interview shows how someone describes a situation. An association task provides clues about meanings that become accessible in response to a cue. A process observation may reveal where implementation gets stuck. None of these sources necessarily explains the entire mechanism by itself.

Choose methods according to the hypothesis. If participants say a form takes too long, investigate the steps and time required. If they expect a report to be used against them, examine its meaning, past experiences and actual follow-up. Simplifying the form alone will not automatically address the second problem.

Record contradictions. When people describe a task as easy but struggle during execution, this is not a reason to discard one source. It is a reason to investigate the task, instructions or context more closely. A useful framework makes uncertainty visible instead of hiding it behind a neat classification.

1. Define the behaviour and current situation.
2. Identify who acts and who controls relevant conditions.
3. Gather evidence about capability, opportunity and motivation.
4. Formulate several possible explanations.
5. Select a well-supported, changeable explanation for an initial test.
6. Design an intervention that addresses that mechanism.
7. Specify in advance which outcomes would demonstrate improvement.
8. Evaluate feasibility and unintended consequences as well.

Record why the intervention follows from the analysis. More explanation can address missing knowledge, but not necessarily missing authority. Greater urgency may do little when the process leaves no time to act. The reasoning should remain understandable to someone who was not involved in the research.

A fictional team wants employees to report technical faults earlier. An initial assumption is that they do not appreciate the importance. The researcher also checks whether the reporting route is accessible, the required information is available and employees know what happens after a report.

The team develops three hypotheses: the route is unclear, recording requires too many steps, or the expected follow-up is unfavourable. It gathers appropriate evidence for each. This example contains no research findings; the possible explanations remain open until investigated.

If the route proves to be the main issue, a small process change can be tested. If follow-up is perceived as unsafe and is genuinely problematic, the working process and management behaviour need attention. A poster saying ‘report it now’ is not a complete solution. Evaluation should examine useful reports and their handling, rather than campaign reach alone.

Shared associations can help distinguish groups that interpret the same situation differently. One group might connect reporting with ‘helping’, another with ‘being blamed’. That suggests a relevant research question, not evidence that the second group is less responsible.

Place these meanings alongside circumstances and behaviour. An audience profile supports analysis when it shows which considerations are shared and which conditions differ between groups. Demographics can describe a group, but they do not replace research into the mechanism explaining its behaviour.

• Starting with a campaign before specifying the behaviour.
• Attributing every problem to motivation.
• Treating a framework as a validated diagnosis of individuals.
• Using self-report alone to make a claim about actual behaviour.
• Presenting the selected solution retrospectively as the only possible explanation.
• Reporting attention or positive reactions as complete behavioural change.

Behavioural diagnosis is useful when it connects observation, explanation and action. A good framework makes that connection explicit and testable. Communication then becomes part of an appropriate solution, alongside the skills and circumstances that make action possible.