SHARE Knowledge Hub

Risk Perception and Decision-Making

People perceive and act on risk through attention, experience, familiarity, workload, social cues and the way work is designed. A worker may miss a weak signal or accept a normalised shortcut even when a written assessment exists. Safer decisions come from visible hazards, usable controls, time to stop and ask, supervisor support and learning why the choice made sense in context.

Updated 2026-10-19GB and NIVersion 2.2

Understanding Risk Perception and Decision-Making

People perceive and act on risk through attention, experience, familiarity, workload, social cues and the way work is designed. A worker may miss a weak signal or accept a normalised shortcut even when a written assessment exists. Safer decisions come from visible hazards, usable controls, time to stop and ask, supervisor support and learning why the choice made sense in context.

Why this matters at work

Repeated successful exposure can make a hazard feel normal even when its physical consequence has not changed. Experience is valuable, but uneventful work can coexist with weak barriers, luck or unreported near misses. Compare actual work with objective control standards and ask what would prevent harm if the next error or defect occurs. Do not treat confidence, willingness or years of service as a substitute for evidence that the present method is safe. Use direct observation, worker discussion and independent challenge to identify workarounds and missed cues. Explain the harm pathway and required controls in a way relevant to the task, and provide usable equipment, time and supervision. Review decisions separately from outcomes: a risky shortcut that succeeds is still a weak decision, while an adverse event does not prove every earlier judgement was unreasonable. Record changes and feedback rather than labelling people complacent. HSE's human-factors approach considers job, individual and organisational conditions together, including how management choices shape behaviour.

Record material assumptions, contrary evidence, unresolved disagreement and the reason for the selected option. Assign responsibility for obtaining missing evidence and define triggers that reverse or reopen the decision. Make it possible for junior, contractor or remote staff to challenge without detriment. The leader should show that challenge can affect outcomes, not simply solicit comments after approval. Match the depth to the consequence: a routine low-risk choice needs less structure than a decision depending on uncertain critical protection, but both benefit from explicit conditions and clear authority.

How organisations manage this in practice

Unsafe shortcut: treating bias awareness as an individual training topic detached from hazard control and management systems. Escalate when the decision, affected population, owner, legal route or safety consequence is not defined. Pause, restrict or add proportionate interim controls where delay could expose people to serious harm. Another competent reviewer should be able to reconstruct the question, evidence, uncertainty, options, challenge, decision, owner, implementation and review trigger.

For practical application, rank sources by legal status, relevance, quality, date, independence and applicability to the actual undertaking. Keep the assessment connected to design, resources, supervision, procedures, competence, maintenance, consultation, emergency arrangements and management of change.

For practical application, define terms with workplace examples and distinguish measured frequency, estimated probability, subjective confidence, concern, acceptability and risk appetite. Check normal work, startup, shutdown, cleaning, maintenance, fault, emergency, change, contractor and foreseeable misuse conditions before concluding.

Conflicting cues require a protected pause

A recorded isolation and an unexpected field indication are conflicting evidence. Familiarity, a deadline or agreement between two indicators does not resolve the actual energy state.

Safety team reviewing photographed evidence and an isolated work area. Photorealistic illustration.
Illustrative UK workplace scene. Read the practical checks below for responsibilities, evidence and limits. Download image · Download smaller image
  1. Name the discrepancy

    Record what the document says and what the actual indication shows, without assigning blame or making a diagnostic judgement about perception.

  2. Pause affected work

    Keep people protected and use the site escalation route rather than interpreting the unexpected state as permission.

  3. Ask the competent verifier

    Resolve the equipment-specific cause and evidence through the authorised safe system; no generic valve procedure is supplied.

  4. Improve the information design

    Review labels, workload, authority and handover after the discrepancy is resolved. Record the decision; there is no automatic restart tick.

Read the complete task-specific control examples

ERICPD is a planning cross-map: Discipline verifies and sustains every control. ESEAP compares effectiveness. A label does not authorise the task; retain the owner, evidence and stop or change decision.

Risk Perception and Decision-Making · ERICPD

Eliminate
Move the pedestrian task and delivery point so the two routes never meet, after checking that the new route does not create another blind crossing.
Reduce
Replace reversing through the busy area with a forward-only vehicle route where vehicle geometry allows.
Isolate
Install fixed barrier and self-closing gated crossing outside the swept path; do not count a painted line, poster or briefing as a barrier.
Control
Test speed limiter and visual/audible warnings; provide a clear stop cue and authorised delivery hand-off for the remaining crossing.
PPE
Keep suitable high-visibility clothing as a residual visibility aid; it does not prevent crush injury and must not become the basis for accepting the lane.
Discipline
Use near-miss reports and worker observation to test whether the design survives time pressure; refresh instructions and fix the layout when shortcuts recur.

Risk Perception and Decision-Making · ESEAP

Elimination
Move the pedestrian task and delivery point so the two routes never meet, after checking that the new route does not create another blind crossing. HSE workplace-transport guidance
Substitution
Replace reversing through the busy area with a forward-only vehicle route where vehicle geometry allows. Staggered breaks reduce exposure but do not substitute the hazardous movement. HSE workplace-transport guidance
Engineering controls
Install fixed barrier and self-closing gated crossing outside the swept path; do not count a painted line, poster or briefing as a barrier. HSE workplace-transport guidance Install and prove speed-limiting equipment, warning/sight-line aids and physical visibility improvements after segregation design; instructions do not replace their function. HSE workplace-transport guidance
Administrative controls
Maintain the traffic plan, authorised human roles, inspection, functional-test arrangements, briefings and response to defects. Suspend conflicting movement when the designed controls fail. HSE workplace-transport guidance Use near-miss reports and worker observation to test whether the design survives time pressure; refresh instructions and fix the layout when shortcuts recur. HSE workplace-transport guidance
PPE
Keep suitable high-visibility clothing as a residual visibility aid; it does not prevent crush injury and must not become the basis for accepting the lane. HSE workplace-transport guidance

Perception starts with a signal reaching the senses. A visual indicator, audible alarm, smell or vibration is a stimulus; poor lighting, noise, display contrast or a vision/hearing limitation can prevent detection. Screen for task-relevant capability where justified, adapt the interface and provide a second cue or independent check rather than assuming one person will always notice. Perceptual set means prior expectation shapes interpretation: a familiar amber lamp may be read as a nuisance instead of a new fault. Distortion can also arise from ambiguous displays, group norms or misleading context. Stress and time pressure narrow attention; fatigue slows information processing and increases missed cues. Selective attention is necessary in a busy task but may filter out a rare critical signal. HSE human-factors guidance sets out sensory and perceptual limits; HSE fatigue guidance explains the attention effects. These mechanisms are prompts for work design, not diagnoses of a worker.

What is risk perception?

Risk perception is how people notice and interpret hazards. A vision test alone cannot prove every warning will be detected: display design, lighting, noise, workload and a second cue matter. Familiarity, time pressure and peer norms can make a weak signal seem normal, so the employer should design usable barriers and a supported stop route rather than rely on warnings alone.

From stimulus to decision: where a critical cue can be lost

From stimulus to decision: where a critical cue can be lost. SIGNAL REACHES THE SENSES: A visual display, audible alarm, smell or vibration provides a cue. Lighting, contrast, noise or task-relevant sensory limitations may prevent detection. ATTENTION SELECTS: Busy work requires selective attention. Stress, workload and fatigue can filter out a rare critical cue or slow information processing. EXPECTATION INTERPRETS: Prior experience creates a perceptual set. A familiar amber lamp may be treated as a nuisance; an ambiguous display or team norm can distort the meaning. CONTEXT SHAPES CHOICE: Familiarity, peer behaviour and schedule pressure can make the shortcut feel normal even when the person knows the hazard. An injury-free history does not prove that a control works. DESIGN SUPPORTS SAFE ACTION: The employer improves cues, workload and equipment, provides a second cue or independent check where required, and supports the authority to stop and escalate uncertainty.
Learning point: This is a work-design mechanism, not a test of a person’s character or a sensory diagnosis. A vision test alone cannot guarantee detection of every warning.
Read the complete visual as text
  1. 1. SIGNAL REACHES THE SENSES: A visual display, audible alarm, smell or vibration provides a cue. Lighting, contrast, noise or task-relevant sensory limitations may prevent detection.
  2. 2. ATTENTION SELECTS: Busy work requires selective attention. Stress, workload and fatigue can filter out a rare critical cue or slow information processing.
  3. 3. EXPECTATION INTERPRETS: Prior experience creates a perceptual set. A familiar amber lamp may be treated as a nuisance; an ambiguous display or team norm can distort the meaning.
  4. 4. CONTEXT SHAPES CHOICE: Familiarity, peer behaviour and schedule pressure can make the shortcut feel normal even when the person knows the hazard. An injury-free history does not prove that a control works.
  5. 5. DESIGN SUPPORTS SAFE ACTION: The employer improves cues, workload and equipment, provides a second cue or independent check where required, and supports the authority to stop and escalate uncertainty.

Risk perception is not simply knowledge of the hazard. A worker may understand that work at height is dangerous but still overreach if they are rushed, watched by peers, badly equipped or used to doing the task that way. A leader may understand process safety theory but still focus on injury rates because those numbers are familiar and easy to report.

HSE's human factors framing is useful because perception and decision-making are shaped by the job, the individual and the organisation. Good decisions come from well-designed tasks, usable procedures, realistic workload, clear information, supportive supervision and a culture where challenge is welcomed.

InfluenceHow it changes perceptionExample
FamiliarityRepeated exposure can make risk feel normal.A team stops noticing an unguarded edge because it has been present all week.
ControlPeople tolerate risks they feel they can control.A driver underestimates fatigue because they think experience compensates.
Peer normsGroup behaviour defines what feels acceptable.New starter copies unsafe shortcuts to fit in.
Leadership cuesPeople infer priorities from what leaders ask about.Programme pressure makes a permit check feel optional.
AvailabilityRecent vivid events feel more likely than hidden chronic risks.Slips get attention while silica exposure is ignored.
Optimism biasPeople assume serious harm is unlikely to happen to them.Experienced technician works live because it has always been fine.
UncertaintyAmbiguous information may be ignored or overreacted to.Gas detector intermittent alarm is treated as instrument nuisance.

How can the hierarchy of controls improve risk decisions before relying on individual judgement?

A supervisor should map the real decision, time pressure, information and equipment, then make the safe action practical: clear cues, independent checks for critical steps, authority to stop and a route to escalate uncertainty. Test the safeguard during real work and review near misses without reflex blame.

Design the decision before time pressure arrives

Design the decision before time pressure arrives. INFORMATION: Weak: Risk hidden in long RAMS or technical jargon. Design response: Simple critical-control briefing and visible cues. TIME PRESSURE: Weak: Decision made while work is waiting. Design response: Pre-planned hold point before irreversible step. AUTHORITY: Weak: Worker unsure if they can stop work. Design response: Explicit stop-work authority and no-blame escalation. PEER INFLUENCE: Weak: Shortcut is the team norm. Design response: Supervisor reinforces standard and invites challenge. UNCERTAINTY: Weak: Ambiguous sign treated as nuisance. Design response: Escalation trigger for abnormal condition. RECOVERY: Weak: No chance to catch error once made. Design response: Independent check, interlock or second verification.
Learning point: The supervisor maps the real choice, information, time and authority. Test usable cues, supported stops and independent checks during actual work; investigate near misses without reflex blame.
Read the complete visual as text
  1. INFORMATION: Weak: Risk hidden in long RAMS or technical jargon. Design response: Simple critical-control briefing and visible cues.
  2. TIME PRESSURE: Weak: Decision made while work is waiting. Design response: Pre-planned hold point before irreversible step.
  3. AUTHORITY: Weak: Worker unsure if they can stop work. Design response: Explicit stop-work authority and no-blame escalation.
  4. PEER INFLUENCE: Weak: Shortcut is the team norm. Design response: Supervisor reinforces standard and invites challenge.
  5. UNCERTAINTY: Weak: Ambiguous sign treated as nuisance. Design response: Escalation trigger for abnormal condition.
  6. RECOVERY: Weak: No chance to catch error once made. Design response: Independent check, interlock or second verification.

This decision design follows HSE human-factors risk-assessment guidance: examine the job, cues, environment and organisation before blaming an operator. For a Northern Ireland site, select the separate HSENI risk-assessment route when recording duties and ownership.

Do not rely on a worker making a perfect judgement in a bad decision environment. If the job is rushed, the procedure is unclear, the alarm is confusing, the supervisor rewards speed and the right equipment is missing, the decision is already being shaped in the wrong direction.

Decision design means building cues and controls into the system: hold points, permits, checklists, two-person verification, clear stop triggers, independent review, peer challenge, fatigue controls and escalation rules.

Decision factorWeak designStronger design
InformationRisk hidden in long RAMS or technical jargon.Simple critical-control briefing and visible cues.
Time pressureDecision made while work is waiting.Pre-planned hold point before irreversible step.
AuthorityWorker unsure if they can stop work.Explicit stop-work authority and no-blame escalation.
Peer influenceShortcut is the team norm.Supervisor reinforces standard and invites challenge.
UncertaintyAmbiguous sign treated as nuisance.Escalation trigger for abnormal condition.
RecoveryNo chance to catch error once made.Independent check, interlock or second verification.

Illustrative decision trap: workers have become used to stepping across a live vehicle lane because no one has been struck. Normalised success is not evidence that the crush risk is controlled. The owner should remove the encounter first, then test safer substitutions and engineered separation before relying on warning, attention or PPE. ERICPD (Eliminate, Reduce including substitution, Isolate, Control, Personal Protective Equipment, Discipline) prompts the decision; the HSE five-level hierarchy identifies the control type, and the two are related rather than identical. See HSE workplace-transport guidance for the vehicle route and HSE hierarchy of controls for control order.

On a narrow screen, scroll the control table sideways for the owner and stop decision.

How can the same risk be judged differently by different people?

A worked example shows how role, pressure and experience change risk perception.

Pump restart: make every perspective explicit

Pump restart: make every perspective explicit. OPERATOR: Possible view: Familiar task feels routine. Challenge: Prompt: what is different from normal? SUPERVISOR: Possible view: Delay feels like the main risk. Challenge: Prompt: what control must not be traded for schedule? ENGINEER: Possible view: Temporary bypass feels technically manageable. Challenge: Prompt: what is the safety function and authorisation route? H&S ADVISER: Possible view: Sees procedural and handback weakness. Challenge: Prompt: what evidence shows the job is controlled? SENIOR MANAGER: Possible view: May see production and customer impact. Challenge: Prompt: what is the worst credible consequence and who owns residual risk? TEAM: Possible view: May follow the dominant voice. Challenge: Prompt: invite challenge before final decision.
Learning point: A maintenance team is under pressure to restart a pump before a batch is lost. These are competing perspectives on one decision, not six independent permissions to bypass protection.
Read the complete visual as text
  1. OPERATOR: Possible view: Familiar task feels routine. Challenge: Prompt: what is different from normal?
  2. SUPERVISOR: Possible view: Delay feels like the main risk. Challenge: Prompt: what control must not be traded for schedule?
  3. ENGINEER: Possible view: Temporary bypass feels technically manageable. Challenge: Prompt: what is the safety function and authorisation route?
  4. H&S ADVISER: Possible view: Sees procedural and handback weakness. Challenge: Prompt: what evidence shows the job is controlled?
  5. SENIOR MANAGER: Possible view: May see production and customer impact. Challenge: Prompt: what is the worst credible consequence and who owns residual risk?
  6. TEAM: Possible view: May follow the dominant voice. Challenge: Prompt: invite challenge before final decision.

A maintenance team needs to restart a pump before production loses a batch. The operator sees low risk because the task is familiar. The supervisor sees schedule pressure. The engineer sees a temporary bypass as acceptable for a short period. The H&S adviser sees a weak isolation and unclear handback.

The mature response is not to accuse one person of poor judgement. It is to make the decision explicit: what has changed, what controls are bypassed, what could happen, who can authorise the decision, what temporary controls exist and how long the condition may last.

PersonLikely perceptionBetter decision support
OperatorFamiliar task feels routine.Prompt: what is different from normal?
SupervisorDelay feels like the main risk.Prompt: what control must not be traded for schedule?
EngineerTemporary bypass feels technically manageable.Prompt: what is the safety function and authorisation route?
H&S adviserSees procedural and handback weakness.Prompt: what evidence shows the job is controlled?
Senior managerMay see production and customer impact.Prompt: what is the worst credible consequence and who owns residual risk?
TeamMay follow the dominant voice.Prompt: invite challenge before final decision.

A permit holder sees conflicting cues

Task, material and setting: A maintenance team is behind schedule; a permit says isolation is complete, but a local indicator suggests pressure remains.

People and prerequisites: The team has authority to stop and asks an independent competent person to verify the energy state.

Before: Redesign permit hand-off so the field indicator and physical isolation check must agree before work.

During: Supervisor observes use on a live job and checks whether time pressure still discourages challenge.

Hold or stop: Hold the task when isolation evidence conflicts or no competent verifier is available.

Completion and hand back: Investigate the decision context, correct the interface and retest at the next shutdown.

Evidence: Keep permit, discrepancy report, worker account, design change and field verification.

A familiar route masks vehicle risk

Task, material and setting: Warehouse staff routinely cross a forklift route because deliveries are usually quiet.

People and prerequisites: Supervisor observes peak and quiet periods, sight lines, separation and why the shortcut is convenient.

Before: Provide a protected crossing or separate route, change layout and coordinate delivery windows.

During: Check the route during real peak work and ask workers whether the safer path is usable.

Hold or stop: Stop crossing where vehicles and people are not safely separated.

Completion and hand back: Review near misses and route use after redesign, then change layout again if shortcuts persist.

Evidence: Keep observations, layout decision, worker feedback, controls and effectiveness review.

Which decision traps matter most in safety?

Common traps include familiarity, optimism, confirmation, authority gradient, groupthink and outcome bias.

Bias does not mean people are careless. It means the human mind uses shortcuts. Those shortcuts are often useful, but they can fail in high-risk work, low-frequency events, complex systems or situations with pressure and uncertainty.

A good safety system anticipates bias and adds challenge, evidence and recovery controls. It asks: what would change our mind, what are we assuming, and who has the authority to stop?

During a pump restart, a familiar minor alarm may be dismissed because previous restarts succeeded. The supervisor should compare the alarm with the required permissives and current plant state, invite an independent challenge and hold the restart until the discrepancy is resolved. That makes the safe choice easier than asking the operator simply to resist familiarity or time pressure. HSE human-factors risk guidance supports designing the decision around real work.

Decision trapHow it shows upControl
Familiarity biasKnown task treated as low risk.Pre-task review of what has changed today.
Optimism biasPeople assume serious harm is unlikely.Use credible worst-case scenario and stop triggers.
Confirmation biasPeople look for evidence supporting the preferred plan.Use independent challenge and contrary evidence prompt.
Authority gradientJunior person stays silent.Structured challenge and explicit permission to question.
GroupthinkTeam agrees too quickly to preserve harmony.Assign a challenger or use pause point.
Outcome biasNo harm last time means decision was good.Review control quality, not just outcome.
Normalisation of devianceRepeated drift becomes normal work.Trend weak signals and compare to required standard.

How should risk be communicated so people understand it?

Risk communication should be specific, credible and linked to the decision people need to make.

Vague warnings such as 'be careful' do little. Effective risk communication explains the hazard, credible scenario, control that must not fail, stop trigger and escalation route.

Different audiences need different emphasis. Workers need practical cues and controls. Supervisors need decision triggers. Senior leaders need control confidence and resource decisions.

A useful message names the signal, the action and the authority to stop. “Amber pressure alarm after maintenance: hold the pump restart, confirm the valve line-up with the authorised person and record the result” is more actionable than “take care”. The supervisor should ask the operator to explain the action back and check that the alarm and procedure tell the same story. If production pressure rewards ignoring the warning, communication alone will fail. HSE human-factors guidance links behaviour to task design and conditions.

Weak messageStronger messageWhy it works
Take care around forklifts.Do not cross the loading doorway while the beacon is flashing; use the marked gate and report blocked walkways.It describes the action and trigger.
Use the permit properly.Do not break the line until isolation, drain, vent and zero-energy checks are signed and witnessed.It identifies the critical control.
Dust is bad for you.For silica cleanup, do not dry brush or use compressed air; use the assessed M- or H-class vacuum or appropriate wet method. Separately use the assessed source control, including suitable on-tool extraction where relevant, and required RPE for the dust-generating task.It gives a clear control standard.
Report near misses.Report any dropped-object, gas alarm or uncontrolled vehicle movement even if nobody was hurt.It defines weak signals.
Manage stress.Tell your manager if workload, role clarity or support is making the work unsafe or unsustainable.It connects wellbeing to work design.

What safeguards help high-risk decisions?

High-risk decisions need hold points, independent challenge, authorisation and recovery controls.

The more serious the consequence, the less the organisation should rely on informal judgement. A safety-critical decision should have a designed process, defined competence, clear evidence and a way to recover before harm occurs.

This is especially important for isolation, hot work, confined spaces, lifting, temporary works, DSEAR, process start-up, driving while fatigued and work at height.

Specify which decision cannot be made under pressure. For a bypassed interlock, the operator should have a visible hold point; a competent authority should verify the plant state and approve a documented alternative only if the applicable controls permit it. The shift handover should state the limitation and restoration owner, with a pre-start test before normal service resumes. An extra signature without an independent check adds paperwork, not a barrier. HSE human-factors guidance frames the system conditions behind error.

SafeguardPurposeExample
Hold pointStops work before an irreversible or high-risk step.No line break before isolation verification.
Independent checkReduces single-person error.Second person checks lifting plan or permit conditions.
Competent authorisationEnsures decision sits with the right authority.Technical authority approves trip bypass.
Stop triggerDefines when work must pause.Gas test fails, weather changes, unknown service found.
Recovery controlLimits harm if an error occurs.Interlock, alarm, emergency stop or exclusion zone.
Post-decision reviewLearns from unusual or high-risk decisions.Review temporary bypasses and permit deviations.

Why did this decision make sense at the time?

This question helps investigators and supervisors understand the system pressures that shaped a decision without excusing unsafe behaviour.

A mature human factors review does not start with 'why were they stupid?' It asks why the decision appeared reasonable locally. Perhaps the alarm had cried wolf before, the procedure was hard to use, the contractor wanted to finish, the supervisor was under programme pressure, or the worker had never seen the serious consequence.

This question is especially useful after permit deviations, isolation mistakes, alarm responses, driving decisions, work at height shortcuts and emergency actions. It turns blame into useful design information.

PromptWhat it revealsImprovement route
What information was visible?Whether the risk signal was clear at the point of decision.Improve displays, labels, briefings, alarms or critical-control cues.
What pressure was present?Whether time, production, client or peer pressure distorted judgement.Add hold points, escalation and leadership alignment.
What felt normal?Whether drift or normalisation had changed the accepted standard.Reset standard through supervision, evidence and worker involvement.
Who had authority?Whether the person making the decision had competence and decision rights.Clarify authorisation and stop-work routes.
What would have helped recovery?Whether a second check, interlock or alarm could catch the error.Strengthen independent verification and recovery controls.
What message did leaders send?Whether formal safety messages matched operational rewards.Align objectives, reviews and supervisor behaviours.

Weak-signal detection and recovery example. During a pump restart, an operator hears a faint change in motor tone while an amber vibration indicator flickers. Production pressure and repeated harmless alarms make it easy to normalise the cue. The safer design gives an unambiguous alarm, a second reading and an explicit hold point: the operator stops restart and informs the supervisor; maintenance checks the vibration sensor, motor and isolation state before release. The supervisor records what the worker noticed, which cue was missed or discounted, the equipment finding and the corrective action. After repair, test the alarm in realistic noise and lighting, ask operators whether they can distinguish it, and review recurrent false alarms. HSE learning guidance supports amplifying frontline weak signals; the pump scenario is an illustrative application.

Where do risk perception and decisions affect operational safety?

For an amber alarm or faint pump vibration during restart, hold the task, report the weak signal and have competent maintenance verify the sensor, pump and isolation before release. Decisions under pressure need clear cues, independent checks and a recovery path.

Risk perception becomes concrete when it is attached to a decision. The same person may make good decisions on a routine task and poor decisions under fatigue, unclear authority, ambiguous alarms or production pressure.

Use these examples for briefings and investigations. Ask what the decision point was, what made the unsafe option attractive, and what safeguard should make the safe choice easier.

Work decisionRisk-perception trapBetter safeguard
Permit-to-workPermit treated as admin because the team trusts the contractor.Verifier checks isolations and critical conditions before work starts.
Isolation and line breakingValve closure assumed because the previous job was similar.Prove, drain, vent and independent verification before breaking containment.
Alarm responseRepeated nuisance alarms make a real alarm feel low priority.Alarm rationalisation, clear response action and escalation threshold.
DrivingFatigue underestimated because the driver wants to finish the route.Journey planning, fatigue stop rule and no-penalty reporting.
Work at heightShort task feels too quick to set up proper access.Predefined access standard and supervisor stop trigger.
Emergency decisionPeople re-enter to save equipment or belongings.Drilled emergency priorities and clear incident command.

When should risk perception concerns be escalated?

Escalate when pressure, uncertainty, drift, group norms or safety-critical decisions could make poor judgement more likely.

Risk perception becomes a management issue when the work environment is shaping unsafe decisions. That may include high workload, fatigue, unclear procedures, production pressure, poor supervision, hidden chronic risks, dominant personalities or a culture that penalises challenge.

Escalation should focus on the decision system, not blaming the person. Ask what made the unsafe choice feel normal, efficient or acceptable.

SignalWhat it may meanAction
People normalise a shortcutGroup norm is stronger than the written rule.Observe task, discuss constraints and redesign controls.
Workers do not report weak signalsReporting may feel pointless or unsafe.Improve feedback, trust and manager response.
Decision made under pressureSchedule or customer pressure is shaping risk acceptance.Pause, escalate and review authority.
New starter copies unsafe practiceInduction and supervision are not controlling culture.Coach supervisor and reinforce standard.
Chronic health risks ignoredImmediate hazards are crowding out long-latency harm.Use exposure evidence and health-risk communication.
Safety-critical uncertaintyPeople are guessing beyond competence.Stop and seek competent technical advice.

What makes a critical check independent?

A competent checker establishes the critical state from defined evidence, has authority to stop the job and tests an alternative explanation rather than repeating the first person’s assumption. For the pump/permit example, a mismatch between isolation indication and the proved physical state is disconfirming evidence: retain the hold until the authorised isolation/operations owner resolves it and accepts the required test. Another signature is not proof.

A verified engineered interlock acts on the unsafe condition; an alarm still depends on detection, understanding and timely response. Record the alarm’s response owner and credible workload or nuisance-alarm failure, and test that the required action is achievable. Do not treat an alarm and a dependent human response as equivalent to an independent automatic prevention barrier.

HSE silica cleanup CN5; HSE human factors in risk assessment.

Where do risk perception and decision-making controls fail?

They fail when organisations assume awareness will overcome pressure, poor design or group norms.

Telling people about a risk does not mean they will perceive it correctly at the point of decision. The signal may be weak, the pressure may be strong, the shortcut may be normal and the consequence may feel remote.

The strongest control is to improve the decision environment: make risk visible, make the safe choice practical, create stop authority and check whether leadership messages match safety expectations.

Failure or weak practiceHow it developsWhat good looks like
Awareness-only controlTraining delivered but task pressures unchanged.Design work, supervision and recovery controls around the decision.
Production cue conflictLeaders say safety first but reward speed.Align objectives, reviews and supervisor messages.
Poor procedure usabilityWorkers bypass a procedure that is too long or unrealistic.Use task analysis and worker input to improve procedure design.
No challenge routePeople stay silent because challenge feels risky.Create structured challenge and just-culture response.
Chronic-risk invisibilityDust, noise or stress risk feels less urgent than immediate production.Use monitoring, health stories and visible control evidence.
Outcome complacencyNo incident means risk is accepted as controlled.Review controls and weak signals, not just outcomes.

What evidence shows decision-making is being managed?

Evidence should show that decisions are designed, briefed, challenged, authorised and reviewed.

Risk perception is hard to prove directly, so evidence must sample the system around decisions. Look for task observation, challenge records, briefing quality, supervision, stop-work examples, procedure usability and incident learning.

For Diploma-level learners, the key point is that human factors should be assessed like any other control system. It needs ownership, competence, monitoring and improvement.

Evidence itemWhat it should showAssurance question
Risk assessment or technical assessmentScope, hazards, affected people, controls, residual risk, owners and review triggers.Does it match the work as done?
Action trackerActions, owners, dates, priority, temporary controls and close-out evidence.Are high-risk actions verified, not merely marked complete?
Training and briefing recordsWho was briefed, on what, when and how understanding was checked.Can workers explain the controls in practice?
Inspection and monitoring recordsPlanned checks, defects, findings, trends and corrective action.Are checks sampling critical controls?
Incident and near-miss learningEvents, weak signals, investigations and improvement actions.Are repeated signals driving system change?
Management reviewTrend, assurance confidence, resources and decisions.Are leaders seeing the right risk picture?
Task observationHow decisions are made in real work.Are workers adapting around poor design?
Procedure usability reviewWhether procedures support the decision at the point of use.Are procedures realistic and used?
Stop-work and challenge recordsWhether people can raise concerns without penalty.Are concerns acted on and fed back?

How should risk perception be assured?

Assurance should test whether people understand critical risks, whether decisions are supported and whether pressure is distorting behaviour.

Assurance should include conversations, observation and review of weak signals. Ask people what would make them stop, what makes the job difficult, what shortcuts they see and what leaders respond to.

For high-hazard work, assurance should sample critical decisions: permits, isolations, bypasses, start-up, lifting, driving, hot work and confined-space entry.

Assurance checkWhat good looks likeEscalate if
Critical-risk understandingWorkers describe credible scenarios and controls.People repeat slogans only.
Decision supportChecklists, permits or hold points are usable.People work around tools.
Leadership cuesSupervisors reinforce stop-work and control quality.Speed messages override controls.
Worker voicePeople report concerns and get feedback.Concerns disappear or repeat.
Procedure fitProcedures match work as done.Procedures are ignored or unrealistic.
Incident reviewInvestigations explore performance influencing factors.Reports stop at 'did not follow procedure'.

How should the work be managed and checked?

Use an accountable plan, practical controls, worker input and checks that show whether the controls work in real tasks.

HSE’s Plan, Do, Check, Act approach connects the risk decision to implementation, monitoring and improvement. For this topic, the duty holder should name a control owner, define the evidence that the control is effective, and change the work if that evidence fails. A management-system label or certificate is not proof that an individual exposure or task is safe. HSE managing health and safety

Official guidancePractical decisionBoundary
HSE Plan, Do, Check, ActAssign the risk owner, implement the chosen control and review field evidence after change or failure.This guidance does not replace the applicable GB or NI law, task assessment or named specialist decision.

What training and competence are needed?

Match practice and assessment to the person’s decision: workers recognise concerns; managers act on controls; specialists advise within their discipline; leaders verify outcomes.

Use observed scenarios, identify the assessor and record what was demonstrated. Reassess after a changed task, failure or new legal or technical information; do not invent a universal certificate expiry.

Use a realistic choice under competing production and safety cues. The operator should identify the missing information and stop; the supervisor should show how to challenge an unsafe restart and brief the next shift; the manager should remove the pressure that rewards bypassing the hold. Record the observed decision and the design change, then retest after a procedure, alarm or staffing change. Awareness of bias alone is not proof of a safe decision system.

Role-specific competence and verification
RolePractice and assessmentAssessorDemonstrated evidence and recordReassessment trigger
WorkerDescribe cues, uncertainty and the pressure behind shortcuts; practise a clear stop route.SupervisorRecognise ambiguous cues and use the hold/escalation route; record assessor and observed decision.Confusing alarm, missed cue or near miss.
SupervisorDemonstrate pre-job briefing and support when conditions differ from the plan.Human-factors or safety leadRehearse a time-pressured alarm decision and independent check; record observed actions and feedback.Changed alarm, workload, staffing or missed hold.
Designer and engineerTest alarms, interfaces, labels and error tolerance with users under realistic conditions.Engineering leadDemonstrate cue design, alarm priority and fail-safe response with users; retain usability trial and defect log.New interface, nuisance alarms or failed trial.
Human-factors or safety specialistAnalyse safety-critical tasks, decision points and plausible failures without assuming blame.Competent human-factors leadTest alternative explanations, workload and system constraints without diagnosing individuals; retain method and limits.Conflicting evidence, repeat event or redesigned job.
Leader and auditorResource work design and test field evidence, near misses and corrective action effectiveness.Board or audit leadChallenge whether safe action is easy in real work; record sampled field decision, owner and verification.Repeat alarms, missed action or ineffective safeguard.

What do the acronyms and key terms mean?

Use these terms to explain a work decision, not as proof that a control is effective.

Acronyms and terms
TermExpansionMeaning for this topic
Risk perceptionHow a person notices and interprets a hazardVaries with context and is not a fixed personal trait.
NormalisationGradual acceptance of a deviationCan hide growing exposure.
Selective attentionFocus on some cues while missing othersDesign critical cues to be salient.
Human factorsJob, organisation, environment and individual influencesUse to redesign systems, not label people.
Work as doneReal task under actual conditionsCompare with the written procedure.
H&SHealth and safetyWorkplace risk context, not a judgement about a worker’s attitude.
RAMSRisk assessments and method statementsWritten task plan; compare it with actual cues and decisions.
HSE hierarchyElimination, substitution, engineering controls, administrative controls, then PPEA source-first order for choosing workplace controls; keep higher-order measures ahead of instructions or PPE.
PPEPersonal protective equipmentResidual protection or visibility aid; it does not replace a feasible safer route or physical separation.
ERICPDEliminate, Reduce, Isolate, Control, Personal Protective Equipment, DisciplineCompare source removal and substitution with engineered separation before work rules or PPE; Discipline means checking that stronger controls keep working.