HBM 22: Theoretical Domains Framework

A construction safety manager conducts a COM-B analysis to understand why workers on a bridge rehabilitation project are not consistently wearing fall protection harnesses. The analysis identifies “psychological capability” and “reflective motivation” as deficient. These are useful starting points, but they raise more questions than they answer. Is the psychological capability problem about knowledge (workers do not know when harnesses are required), about attention (workers forget in the rush of work), or about decision-making (workers misjudge the risk on familiar tasks)? Is the motivation problem about beliefs (workers think falls are unlikely), about goals (safety competes with productivity), or about identity (workers see harness use as a sign of inexperience)?

COM-B tells the team which broad component is the barrier. But for designing a specific, targeted intervention, the team needs more detail. The Theoretical Domains Framework (TDF) provides exactly this — a magnifying glass that breaks COM-B’s broad categories into 14 specific, assessable behavioral domains.

Why Detailed Barrier Analysis Matters

In transportation, construction, and planning practice, broad diagnoses often lead to broad interventions. If the diagnosis is “motivation is low,” the response might be a generic motivational campaign. If the diagnosis is “capability is lacking,” the response might be a general training session. These broad responses frequently fail because they do not target the specific mechanism that is actually blocking the behavior.

Consider the difference:

• “Workers lack psychological capability” → generic safety training
• “Workers lack knowledge of when harnesses are required for specific tasks” → task-specific decision aids
• “Workers have adequate knowledge but forget to don harnesses when transitioning between ground-level and elevated work” → context-triggered reminders and environmental cues

The second and third diagnoses point to entirely different interventions. The TDF’s value lies in making these distinctions visible and systematic.

The TDF Explained

The Theoretical Domains Framework was developed by a collaboration between behavioral scientists and implementation researchers. The original version (Michie et al., 2005) identified 12 domains; the validated refinement (Cane et al., 2012) established 14 domains. These 14 domains represent a synthesis of 33 theories of behavior and behavior change, with 128 constructs organized into distinct, assessable categories.

The Relationship Between COM-B and TDF

TDF and COM-B are not competing frameworks. They are nested:

Core Constructs: The 14 Domains

Causal Logic

The TDF does not propose a single causal pathway. Instead, it provides a comprehensive checklist of determinants, any of which can serve as a barrier or facilitator for any given behavior:

The causal logic is pragmatic: the TDF assumes that behavior is multiply determined, and that different behaviors in different contexts will have different dominant barriers. The framework does not prescribe which domains are most important in general — that is an empirical question answered by the assessment.

Data Needed

TDF-based studies typically use qualitative or mixed-methods data:

SAMPLE TDF INTERVIEW QUESTIONS (Construction Safety — Harness Use)
KNOWLEDGE
“Can you describe when fall protection is required on this project?”
“What are the site-specific rules for harness use?”
SKILLS
“Walk me through how you inspect and put on your harness.”
“Have you received training on the specific harness system used here?”
SOCIAL/PROFESSIONAL ROLE AND IDENTITY
“Is wearing a harness something that experienced workers do, or is
it more associated with newer workers?”
“How does harness use fit with being a skilled [trade] worker?”
BELIEFS ABOUT CAPABILITIES
“How confident are you that you can work efficiently while wearing
a harness?”
“Are there tasks where the harness makes your work significantly
harder?”
BELIEFS ABOUT CONSEQUENCES
“What do you think would happen if you didn’t wear a harness for
a quick task at height?”
“Have you seen or heard about fall incidents on projects like this?”
ENVIRONMENTAL CONTEXT AND RESOURCES
“Are harnesses easily accessible when you need them?”
“Is there time in the schedule to properly don and adjust harnesses?”
“Does the weather or work environment affect harness use?”
SOCIAL INFLUENCES
“Do your supervisors consistently wear harnesses when at height?”
“What would happen if you reminded a coworker to put on their harness?”
GOALS
“When you’re working at height, what is your main priority?”
“Does harness use ever conflict with getting the job done on time?”
MEMORY, ATTENTION, AND DECISION PROCESSES
“Have you ever forgotten to put on a harness when moving to elevated
work?”
“Are there transitions during the day when harness use gets overlooked?”
EMOTION
“How do you feel about working at height with and without a harness?”
“Is there frustration associated with harness use?”
BEHAVIOURAL REGULATION
“Do you have a personal routine for checking your harness?”
“What would help you remember to use the harness consistently?”

Methods

• Qualitative framework analysis: Transcripts from TDF-based interviews are coded deductively using the 14 domains as the coding framework. Coded data are analyzed to identify which domains contain the most barriers.
• TDF-based coding: A structured coding process where two or more researchers independently code interview data to TDF domains, then reconcile disagreements. This produces a domain-level summary of barriers and facilitators.
• Domain relevance ranking: After coding, domains are ranked by frequency, specificity, and importance to identify the key barriers. Not all 14 domains will be relevant for every behavior.
• Mixed methods: Combining qualitative TDF interviews (for depth and mechanism) with quantitative TDF questionnaires (for prevalence and generalizability across the target population).
• Linkage to BCW: Once key barrier domains are identified, they are mapped to COM-B components and then to BCW intervention functions using published linkage matrices.

Construction Safety Example: Why Harness Compliance Fails

A construction company managing a highway bridge rehabilitation project observes that fall protection harness use is inconsistent — around 65% compliance in weekly audits, despite training, available equipment, and clear policies. A TDF-based study involving 24 semi-structured interviews with ironworkers, laborers, and supervisors reveals the following barrier profile:

Detailed Findings by Domain

What TDF Reveals That COM-B Alone Could Not

A COM-B analysis would have identified “reflective motivation” and “social opportunity” as barriers. TDF reveals the specific mechanisms within those broad categories:

Strengths

Limitations

Best Use Case

The TDF is the right framework when a team needs detailed, specific barrier and facilitator identification beyond what COM-B provides. It is particularly valuable when:

• A COM-B analysis has identified broad barriers but more specificity is needed for intervention design
• The target behavior is complex and potentially influenced by many different factors
• Previous interventions have failed and the team suspects the real barrier was not identified
• The team wants a theoretically grounded, systematic approach to qualitative research on behavioral determinants
• The study will inform the design of a multi-component intervention where each component must target a specific mechanism

TDF is less useful when the barriers are obvious, when the behavior is simple and well-understood, or when the team lacks the resources for rigorous qualitative research.

Key Takeaway

Key References

Exercises and Discussion Questions

The next post turns from diagnostic frameworks to a different approach: rather than identifying individual psychological barriers, nudge theory and choice architecture redesign the decision environment itself to make desired behaviors easier, more salient, or the default option.