How Poor Risk Assessment Terminology Is Eroding Safety Culture

Safety culture is understood as being made up of three interrelated aspects:

Psychological (individual and group values, attitudes, and perceptions),

Behaviour (what people actually do), and Organisational (systems, environment, policy, and procedure).

Of the three, Organisational factors are the most overlooked as a key influencing factor of culture — yet they are frequently the origin point for the other two. This case study examines one specific organisational failure — confused terminology in risk assessment — and traces how it cascades through Behaviour into Belief, quietly normalising risk across an organisation.

the risk assessment challenge

Effective risk assessment depends on four distinct concepts being clearly understood:

• Task — a defined activity with clear boundaries, a start and end, and steps.

• Hazard — something with the potential to cause harm. That potential comes from energy.

• Risk — the combination of likelihood and severity of a harmful consequence.

• Control — a measure that reduces the likelihood or severity of harm.

  
  

In practice, when workers and supervisors are asked to identify hazards, the most common answers are not hazards at all. They are tasks (e.g. confined space entry), risk-increasing factors (e.g. complacency, time pressure, poor maintenance), controls (e.g. isolations, COSHH assessments) or the lack of controls. Genuine hazards, the energy sources capable of causing harm, are often conspicuously absent from the list.

This is not simply a training gap. It reflects a widely held but incomplete mental model of what a hazard actually is.

the organisational root cause

The confusion is not incidental — it is frequently built into the systems organisations use to manage risk. Many risk assessment software platforms populate their hazard libraries with exactly the same category errors: Human Factors, Poor Lighting, Insufficient Egress, Conflicting Work Activities, Isolation Verification. None of these are hazards. They are conditions or factors that influence the likelihood of an incident occurring or the severity should it occur — but they are not the thing that transfers harm to a person's body.

This matters because these systems are not neutral. They are the Organisational layer of culture in its most literal form: a policy, a procedure, a piece of infrastructure that shapes what people are prompted to think and say. When a dropdown menu offers "Fatigue" as a hazard example, it teaches the workforce — at scale, repeatedly, every time a risk assessment is completed — that this is what a hazard looks like. The system is not a passive record-keeping tool; it is an active training mechanism, and it is training the wrong lesson.

why energy is the missing link

Safety science has understood hazard as an energy phenomenon since the 1960s. Gibson (1961) first modelled accidents as an energy source making contact with a vulnerable target through a failed or absent barrier. Haddon (1968, 1980) extended this, arguing that injury occurs when energy is transferred to the body in a form or quantity exceeding what it can tolerate, and that prevention should focus on removing, reducing, isolating, or controlling that energy. Later work (Fleming; Ball; Albert, Hallowell & Kleiner) translated the theory into ten practical energy types: gravity, motion, mechanical, electrical, pressure, sound, radiation, biological, chemical, and temperature.

Field research gives this real weight. Hallowell's studies, based on thousands of hours of direct worker observation, found that only around 45% of hazards present on a task were recognised by workers without structured training. When workers were trained to think in terms of energy sources rather than tasks or conditions, hazard recognition improved by roughly 30%. As Hallowell puts it, every injury is the result of undesirable contact between a person and an energy source.

This gives us a precise definition chain that resolves the industry's confusion:

1. Definition — A hazard is anything with the potential to cause harm.

2. Question — What is it that actually causes the harm? Energy.

3. Mechanism — Harm occurs when that energy transfers to the body in excess of what it can absorb or resist.

Applying this test immediately disqualifies "Fatigue" and "Time Pressure" as hazards — they influence the likelihood of contact with an energy source, but they transfer nothing to the body themselves. They are risk-increasing factors, not hazards.

The Cascade Through Interrelated Aspects of Culture

Organisational 

When software systems and procedures present tasks, risk factors, and controls as if they were hazards, they embed the error into the daily workflow. 

Behaviour 

Workers complete risk assessments by listing what the system prompts them to list. Genuine energy sources go unrecorded and, by extension, uncontrolled.

Psychological 

Over time, repeated reinforcement shapes what people genuinely believe a hazard is. The workforce comes to perceive risk assessment as a paperwork exercise about conditions and procedure, rather than a discipline for identifying the specific energy that could kill or injure them. This is risk normalisation: the energy source itself fades into the background, because the language used to describe it never named it in the first place.

Conclusion

Poor risk assessment terminology is rarely a knowledge problem alone — it is an organisational design problem. Systems that fail to distinguish tasks, hazards, risks, and controls do not just produce imperfect paperwork; they actively shape what people believe a hazard is and, ultimately, how seriously they take the energy sources that can genuinely harm them. Correcting the terminology at the system level — grounding "hazard" explicitly in energy, using the definition–question–mechanism chain — is a direct, evidence-based intervention at the Organisational level, with a demonstrated route to improving both Behaviour and Belief.

References: Gibson, J.J. (1961); Haddon, W. (1968, 1980); Fleming, M. (2009); Ball, L. (NASA); Albert, A., Hallowell, M.R. & Kleiner, B.M. (2014); Hallowell, M.R. (Construction Safety Research Alliance, CU Boulder).