Treat your study time as decision practice. For every topic you revise, ask three questions: what is the hazard, what is the consequence if it goes wrong, and which control best reduces it at source? Work through short written scenarios, rate them on a matrix, and compare your answers with a study partner. Administrative details about the credential belong with IOSH, the issuing body at https://www.iosh.com/; your study time is better spent on judgement than on logistics.
Hazard versus risk: the distinction that everything else builds on
A hazard is something with the potential to cause harm; a risk combines how likely that harm is with how severe it would be. Confusing the two words derails every later judgement, so anchor the distinction with concrete pairs before studying anything else.
Test yourself with pairs rather than definitions. A trailing cable is a hazard; the risk is that someone trips, and that risk depends on where the cable runs, how busy the walkway is, and how badly a fall would injure someone. Electricity is a hazard; the risk differs enormously between a faulty kettle in an office and live switchgear in a substation. Same hazard, very different risks.
This distinction matters because controls act on different parts of it. Some controls reduce likelihood (training, segregation, maintenance schedules); only a few reduce severity at source (eliminating the hazard entirely, or reducing the energy involved). If you write 'risk of electricity' in a risk assessment, you cannot decide which control fits. If you write 'risk of electric shock from damaged flex during use', a control choice becomes obvious. Practise rewording vague phrases into hazard-plus-consequence sentences.
Rating the matrix: separating likelihood from severity under scenario pressure
A risk rating matrix scores likelihood and severity separately and multiplies or combines them into a rating that drives priority and control choices. The discipline is to justify each score in writing instead of producing a single gut-feeling number.
The common shortcut is to average the two dimensions: 'the driver is trained and the site is small, so the risk is medium'. That collapses two different questions into one. Severity asks: if this goes wrong, how bad is the worst credible outcome? Likelihood asks: how often is the task done, how exposed are people, and what already exists to prevent the event? Training affects likelihood; it rarely reduces the severity of the event if prevention fails.
Worked scenario: a warehouse uses a counterbalance forklift in an aisle shared with pickers. A learner rates the risk 'medium — drivers are trained and speed is limited'. The better decision: severity of a pedestrian-forklift collision is catastrophic regardless of training, so score severity first and defend it; then assess likelihood in the light of traffic volumes, visibility, and any segregation already present. The rating now points directly at the control priority — physical segregation — rather than at more driver reminders.
The hierarchy of control: why order of preference is the whole point
The hierarchy of control ranks risk controls from most to least effective: elimination, substitution, engineering controls, administrative controls, and finally personal protective equipment. The order is the teaching point, because it says reduce risk at source before relying on human behaviour or equipment worn by individuals.
Each tier works differently. Elimination and substitution change the task or the material itself, so the protection cannot be forgotten or skipped. Engineering controls separate people from the hazard through guarding, barriers, ventilation, or interlocks — they protect everyone in the area without individual cooperation. Administrative controls (procedures, training, signage, supervision) depend on people following rules, and PPE protects only the wearer and only when worn correctly.
The table below is worth reconstructing from memory during revision. When you can place any control a scenario mentions into its tier — and spot when a plan leans on the lower tiers while ignoring the top ones — you have internalised the concept rather than memorised a list. Residual risk is what remains after controls; a good answer always states what is left over and whether it is tolerable.
| Control level | How it works | Workplace example | Typical weak answer |
|---|---|---|---|
| Elimination | Removes the hazard entirely | Design out a work-at-height task with self-cleaning materials | Skipping straight to harnesses |
| Substitution | Replaces hazard with something less dangerous | Water-based coating instead of a solvent-based one | Same chemical, more gloves |
| Engineering controls | Physically separate people from the hazard | Fixed guards, pedestrian barriers, local exhaust ventilation | Relying on a warning sign alone |
| Administrative controls | Change how people work | Permit-to-work, training, supervision, job rotation | Assuming training alone is sufficient |
| PPE | Protects the individual wearer | Eye protection, hearing protection, gloves | Presenting PPE as the main control |
Choosing a control in a scenario: a worked height task from first principles
Scenario questions reward a fixed reasoning route: state the hazard and consequence, propose an option from the top of the hierarchy first, then justify why lower-tier controls are supplementary rather than primary.
Worked scenario: two workers are to clear moss from gutters on a single-storey depot. A plausible first draft says: 'ladders with harnesses, workers briefed to take care, hi-vis worn'. The mistake is structural — it opens at the bottom of the hierarchy and treats care as a control. Falls from height have high severity whatever the worker's competence, so a plan built on individual care protects least exactly where consequences are worst.
The better decision: ask first whether the task can be eliminated or redesigned — for example, a telescopic gutter-clearing tool operated from the ground, or gutter guards that reduce the cleaning frequency. If work at height remains, an elevated work platform is an engineering control that removes the fall hazard from the working method rather than arresting the fall. A rescue plan and harnesses then support the chosen method instead of substituting for one. This ordering matters because it produces controls that survive a forgetful or rushed day.
Incident investigation: tracing immediate causes back to management failures
Investigation links three levels: the immediate cause (the unsafe act or condition), the underlying causes (why it was possible), and root causes in the management system (planning, training, maintenance, supervision). Stopping at the immediate cause produces blame and no improvement.
Practise the tracing chain on paper. A worker's hand contacts a moving conveyor (injury). Immediate cause: guard removed. Underlying cause: the guard jammed the product, so production pressured staff to work around it; no reported fault process. Root cause: the system buys conveyors whose guards interfere with the job and has no route for workers to raise design faults. Each level names a different owner and a different fix.
Distinguish near misses from injuries when you practise: a near miss is an event that could have caused harm but did not, and it is valuable precisely because it costs nothing to learn from. A good investigation answer identifies who should gather evidence, notes that findings must feed back into risk assessments and procedures, and avoids conclusions based on blame — blame ends the conversation before the system failure is found.
Roles, competence, and documentation: who is answerable for what
Management-level safety training centres on accountability: organisations owe duties as employers, and managers translate those duties into assessments, controls, supervision, and records. Competence means the right mix of training, knowledge, experience, and awareness of one's own limits.
Practise sorting statements by role. Setting policy and resourcing controls sits with the organisation and its senior leadership; day-to-day enforcement of a segregation rule sits with a supervisor; reporting a near miss and following safe procedures sits with every worker. A competent person appointed to advise is not a substitute for line management — competence means knowing when to escalate beyond your own limits, not having an answer to everything.
Documentation is the visible proof of the process, which is why risk assessment records, inspection findings, training records, and investigation reports are repeatedly assessed. When you write a practice risk assessment, check that it records who assessed, what the hazards and controls are, who is responsible for actions, and when review will occur. A risk assessment that is never reviewed after change — a new machine, a new substance, an incident — has failed its purpose, and saying so in an answer demonstrates understanding rather than recall.
A preparation sequence with a self-check rubric and readiness checks
Structure revision in three passes: concepts first, scenario decisions second, and timed mixed practice last. Use the rubric below to mark your own written answers, and finish when the readiness checks feel routine rather than strained.
Adaptable sequence: pass one, build one-page summaries of hazard/risk, the matrix, the hierarchy of control, incident causation, and roles, each with two of your own examples. Pass two, write ten short scenario answers using the fixed reasoning route from section four, and swap with a partner to compare control choices. Pass three, mix timed questions with flashcard recall, then rebuild the hierarchy table and the causation chain from a blank page.
Practical exercise: pick five everyday tasks — say, carrying boxes upstairs, deep-frying at home, cleaning windows from a ladder, changing toner in an office printer, and cycling to work. For each, write the hazard and consequence, rate likelihood and severity separately on a five-by-five matrix, and list one control from each hierarchy tier. Expected observations: your severity scores wobble between sittings, your first drafts put PPE second, and your controls target people more often than the task. Those wobbles are exactly what the rubric exists to catch.
Self-check rubric: did you state the consequence before the rating (2 points); did you justify severity independently of likelihood (2); did you propose an elimination or substitution option before PPE (2); did you name residual risk and a review trigger (2); could a peer reconstruct your reasoning from your written answer alone (2)? Ten or more suggests sound decision habits; below eight, repeat the exercise with different tasks. These scores are learning milestones only, not predictions of any assessment outcome.
Readiness checks before you finish: you can define hazard and risk in one sentence each without notes; you can place any named control in its correct hierarchy tier instantly; you can rate two fresh scenarios and agree with a partner within one matrix cell, with reasons; you can trace a described incident from immediate cause to root cause unaided; and you can state what should trigger a risk assessment review. When all five hold, you have converted the syllabus into judgement.
References and further reading
Use these references to explore the concepts and check the latest information from the relevant organizations.