Study IOSH Managing Safely as a sequence of decisions: name the hazard specifically, assess the risk using likelihood and severity separately, rank it on a matrix, control it using the hierarchy of risk control, investigate any incident down to underlying causes, and measure performance with both proactive and reactive indicators. Practise each step on real work areas you know, and check yourself with a rubric rather than rereading notes. Course delivery and assessment format are set by licensed training providers, so confirm administrative details with your provider or the issuer at https://www.iosh.com/.
Hazard versus risk: the distinction the whole course rests on
A hazard is something with the potential to cause harm; a risk is the combination of how likely that harm is and how severe it would be. Every later topic — assessment, control, measurement — assumes you can keep these two ideas separate.
Compare the two definitions directly. 'Trailing cable across a walkway' is a hazard: it has potential to cause harm. 'Someone tripping over that cable and fracturing a wrist during shift changeover' describes a risk, because it specifies the likelihood of the event and the severity of the outcome. When you revise, force yourself to restate every hazard in risk form: who could be harmed, how, and how badly. If your sentence contains no people and no consequence, you are still describing a hazard.
This distinction matters because the tools apply to different objects. You identify hazards; you assess, rate, and control risks. Exam-style options often blur the terms deliberately — one option names a hazard when the question asks for a risk description, or vice versa. Build the habit of asking two questions about any situation: what has the potential to cause harm, and what is the realistic likelihood and severity of that harm happening in this specific place, to these people, doing this task?
- Hazard test: does the statement name a source of potential harm (substance, machine, activity, environment)?
- Risk test: does the statement identify who could be harmed and describe a plausible consequence with a sense of likelihood?
- Rewrite drill: take five hazards from your own workplace and convert each into a one-sentence risk description.
Running the risk assessment steps on a real work area
Work through the assessment stages in order: identify the hazards, decide who might be harmed and how, evaluate the risks and decide on controls, record the findings, and review when things change. Practise the sequence until it runs as one habit.
The steps are sequential for a reason: skipping the 'who might be harmed and how' stage produces vague controls. Cleaners, new starters, contractors, and visitors experience the same area differently. A loading bay risk looks different for a driver who spends five minutes there than for a yard worker who spends the whole shift. When you practise, name the groups explicitly and describe the harm pathway — struck by, slipped on, exposed to — rather than writing 'staff' and 'injury'.
Scenario 1 — the matrix mistake. A warehouse supervisor rates a trailing cable across the main picking route as 'low risk' because no one has tripped yet. The plausible mistake is using incident history as a proxy for likelihood. The better decision: separate the two dimensions. People cross that route constantly, so likelihood of contact is high; a trip on a hard floor can cause a fracture or head injury, so severity is also significant. The correct rating is high on both counts regardless of the clean record. Why it matters: absence of past incidents says nothing about future likelihood, and a low rating would justify no action. Treat 'it hasn't happened yet' as a warning sign in your reasoning, never as evidence.
Choosing controls by hierarchy instead of reaching for PPE
The hierarchy of risk control ranks control measures from most to least effective: eliminate the hazard, reduce or substitute it, use engineering controls to isolate people from it, then administrative controls, with personal protective equipment as a last layer — not a first choice.
The ranking reflects reliability. Elimination removes the hazard entirely, so it cannot fail at the moment of exposure. Engineering controls work whether or not a person remembers anything. Administrative controls depend on people following rules, and PPE depends on people wearing it correctly, at the right time, in good condition. When an exam scenario offers several controls, classify each option into its hierarchy level first; the better answer is usually the highest-ranked one that is reasonably practicable in that situation.
Scenario 2 — the gloves-first error. A workshop uses a solvent that irritates skin, and a supervisor's first proposal is chemical-resistant gloves and splash goggles. The plausible mistake is treating PPE as the solution. The better decision: work down the hierarchy. Could the solvent be eliminated or substituted for a lower-hazard product? Could local exhaust ventilation or closed containers reduce exposure? Only then do gloves become a supplementary layer. Why it matters: PPE leaves the hazard fully present, protects only the wearer, and fails silently when a glove is torn or forgotten. In scenario questions, always ask what could be changed about the task, the substance, or the equipment before anything worn on the body.
| Hierarchy level | What it does | Workshop solvent example |
|---|---|---|
| Eliminate | Removes the hazard completely | Stop using the solvent-based process |
| Reduce / substitute | Lowers the hazard or exposure potential | Switch to a water-based cleaner |
| Engineering control | Isolates people from the hazard without relying on behaviour | Local exhaust ventilation, closed containers |
| Administrative control | Changes how people work | Job rotation, restricted access, training, signage |
| PPE | Protects the individual at the point of exposure | Gloves and goggles as a final layer |
Investigating incidents beyond 'carelessness'
Investigate incidents by distinguishing immediate causes — the unsafe act or condition present at the time — from underlying and root causes such as organisation, supervision, training, and design. 'Human error' is a description, not a conclusion.
Practice cause-mapping in layers. For a pallet truck striking a racking upright: the immediate cause is the collision; asking why it happened may reveal a blind corner, an unclear traffic route, pressure to work quickly, or no refresher training. Each layer must be followed until you reach something the organisation can actually fix. A good test: for every cause you write, ask 'why does that happen here?' If you cannot answer, you have not reached the underlying cause.
This layering matters because the corrective actions differ at each level. 'Tell the driver to be careful' addresses only the immediate cause and leaves every contributing condition in place. 'Re-route pedestrian traffic and add a one-way system at the corner' addresses the conditions that produced the error. In exam-style scenarios, options that blame the individual without changing the conditions around them are the weaker choices; options that change the workplace, the system, or the supervision of the task are stronger, because they prevent recurrence rather than hoping for better behaviour.
- Immediate cause: the act or condition directly involved in the event.
- Underlying cause: the workplace or system condition that allowed it — design, maintenance, supervision, workload, training.
- Root cause check: keep asking why each cause occurred until the answer is something management can change.
Manager responsibilities: organising for safety, not delegating it
The course frames safety as a management responsibility: plan, organise, control, monitor, and review health and safety as part of running the work. Every worker has a role, but the ability to direct resources and change systems sits with managers.
Compare two approaches. A reactive manager responds after harm occurs: incident reports, blame, added rules. A managing approach plans safe systems before work starts, organises competent people and resources, controls the work so the planned way is the done way, monitors through observation and measurement, and reviews after incidents or change. Revision should focus on what each element means in practice — for example, 'control' means making sure the risk assessment you wrote is actually reflected in how the task is done on the floor, not filed in a folder.
Competence is a recurring thread. Where the course discusses roles, the point is that the person managing safety must have the training, knowledge, and experience appropriate to the situation, and must know the limits of their own competence — knowing when to seek specialist advice. In scenario questions, watch for situations where a manager exceeds their competence: authorising a confined-space entry without assessment, or signing off a risk assessment for a process they do not understand. The stronger answer recognises the limit and escalates or brings in the right expertise rather than proceeding.
A practice exercise with a self-check rubric
Build one complete risk assessment for an area you know well, then score it against a fixed rubric. Repeat on three different areas — one office-like, one plant or vehicle-related, one involving substances or manual handling.
The exercise: choose your own workplace area or a familiar one, and write a full risk assessment in about twenty minutes — five hazards, affected groups, likelihood and severity ratings with reasons, controls mapped to hierarchy levels, and a review trigger. Do this closed-book. Then score it against the rubric below and note which criterion failed first. Across three different area types you will see which hazard categories you consistently overlook; that pattern, not the score itself, tells you what to study next. A rubric score is a learning milestone, not a prediction of any exam outcome.
Expected observations when you do this properly: your first attempts name hazards too generally ('unsafe stairs') and your controls cluster at the bottom of the hierarchy. On the second or third attempt, hazards become specific ('unsecured ladder positioned on a sloped surface'), each risk names a person and a harm, and at least one control sits above the administrative level. Adaptable preparation sequence: first block, definitions and matrix drills; second, control-classification sorting from mixed lists; third, three timed risk assessments with the rubric; fourth, incident cause-mapping from scenarios; final block, mixed scenario practice and rubric re-check.
- Hazard named specifically and by source (2 points) — 'trailing cable across picking route', not 'tripping hazard'.
- Affected people identified by group and task (2 points).
- Likelihood and severity rated separately, each with a stated reason (2 points).
- Controls mapped to correct hierarchy levels, with at least one above administrative where practicable (2 points).
- A review trigger stated — change of process, equipment, personnel, or after an incident (2 points).
- Self-check guide: 9–10 shows the sequence is fluent; 6–8 means revise one weak step; below 6, repeat the exercise on an easier area first.
Measuring performance: proactive and reactive indicators
Reactive indicators measure harm that has already occurred, such as incidents and near-miss reports; proactive indicators measure activity before harm, such as inspections completed, training delivered, and observations of safe behaviour. A balanced system uses both.
Compare the two types when you revise. Reactive data tells you where the system has already failed, and it arrives too late to prevent the harm it reports. Proactive data tells you whether the controls you designed are actually functioning — are inspections happening on schedule, are actions from previous assessments closed out, are people observed working the way the assessment assumed? When a scenario mentions a workplace that only looks at incident numbers, the stronger answer introduces proactive measures, because a fall in incident numbers can equally mean hazards are falling or that reporting has stopped.
Near-miss reporting belongs in this topic and is easy to misplace. A near miss is an event that could have caused harm but did not on this occasion — and it is a leading source of learning precisely because no one was injured, so the underlying causes can be fixed before severity materialises. Link it back to earlier sections: a near-miss report should trigger investigation into immediate and underlying causes, and its findings should feed the review of the relevant risk assessment. Practise drawing that loop: measure, investigate, revise controls, review, measure again.
References and further reading
Use these references to explore the concepts and check the latest information from the relevant organizations.