Prevent and monitor accidents at work
Classify human and system errors, evaluate Fox et al.’s token economy and Swat’s accident-monitoring study, and design reporting that does not hide risk.
Section OPTION-OR-4.4.3 · 4.4.3 Health and safety
Original StudyForA teaching mapped to Cambridge 9990 Organisational Psychology for 2028–2030. Example studies are independent learning support, not Cambridge material or official mark schemes. Findings are limited to their samples, measures and settings.
Before you begin
- Use the study design and sample to decide how far a workplace finding can be generalised. A change in productivity, satisfaction, reported incidents or accidents is a different outcome and needs its own evidence.
Work on paper. Try each question before revealing help, and explain why your method works.
Quick readiness check
A worker follows the correct steps but performs them in the wrong order. What error is this?
What you will learn
- Human errors of omission, commission, sequencing and timing; system errors in operator–machine systems
- Token economies for reducing work accidents; Fox et al. (1987)
- Monitoring accidents and risk events; Swat (1997)
Classify how an accident pathway can fail
A human omission is a required action not carried out; a commission is an action carried out incorrectly. A sequencing error places correct actions in the wrong order, while a timing error performs an action too early, too late or at the wrong moment. The labels describe how an action deviated from the intended task; they do not by themselves explain why it happened.
A system error can arise in the operator–machine interface, including poorly designed or confusing controls and displays. If controls look similar, labels are unclear or important information is hard to see, the system can make error more likely. Accident analysis should inspect the task, equipment, training, workload and environment instead of stopping at individual blame.
A useful investigation traces what was expected, what happened, what information and controls were available, and what conditions shaped the action. More than one error or system condition can contribute to the same event.
Fox et al. used tokens to reinforce safety behaviour
Fox, Hopkins and Anger (1987) introduced a token economy using trading stamps at two open-pit mines. Workers earned stamps for specified safety-related actions, including working without lost-time injuries, making adopted safety suggestions and taking action that prevented an accident. The programme was followed by reductions in lost-time injuries, days lost and accident costs.
The intervention ran in particular mining operations and bundled several contingencies, so it does not show that one reward alone caused every change or that the same programme will work in another industry. Rewarding injury-free outcomes can also create pressure not to report harm; the original programme itself included a loss of awards for failure to report injuries or accidents.
An ethical safety programme rewards hazard correction, safe procedures and reporting, not silence or an accident-free statistic. Keep a confidential, non-punitive route for injuries and near-misses, fix system hazards, and check whether reporting completeness changes when incentives are introduced.
Swat showed why incident monitoring needs detail
Swat (1997) examined 83 accidents across four industrial plants with a total of 2,964 workers. At least 45.8% of those studied involved housekeeping violations. A follow-up at one meat-processing plant found that 95% of incidents and minor injuries were not reported.
The study illustrates how recording circumstances and minor injuries may reveal recurring hazards that a count of severe accidents misses. The figures come from selected plants and an additional study in one meat-processing site; they are not universal rates for all workplaces.
A monitoring system should record what happened, task and equipment context, injury severity and near-misses, then look for repeated patterns. Make reporting accessible and protect workers from punishment for honest reporting. Track the reporting rate as well as injury counts, because fewer reports can mean less reporting rather than improved safety.
Sources for this lesson
- Cambridge International AS & A Level Psychology 9990 syllabus, 2028–2030
- Cambridge International AS & A Level Psychology 9990 Reference List, 2028–2030
- Fox, Hopkins and Anger (1987), The long-term effects of a token economy on safety performance in open-pit mining
- Swat (1997), Monitoring of accidents and risk events in industrial plants
Pause and explain
What is the safest interpretation of the 95% figure in Swat (1997)?
Worked example
After introducing a reward for ‘zero reported injuries’, a warehouse records fewer incidents. Use Fox et al. and Swat to evaluate whether safety improved.
Show the worked solution
- A reduction in reports does not distinguish fewer hazards from less willingness to report.
- Fox et al.’s programme in two mines was followed by fewer lost-time injuries, but it rewarded injury-free outcomes and included penalties for failed reporting.
- Swat found substantial underreporting in one meat-processing follow-up; monitor near-misses, minor injuries and reporting access as well as severe injuries.
Answer The warehouse cannot infer safer work from fewer reported incidents alone. The reporting incentive may suppress reports, and Swat’s findings show why minor incidents need to be captured. Audit hazards and safety controls, provide a non-punitive reporting route, and evaluate injury and near-miss rates together with reporting completeness.
Try a topic test
Choose Easy, Normal or Difficult. Save your answers and open worked review after finishing.
Test this topic →From guided practice to a new situation
Write your answer and reasoning first. Use a hint only if you are stuck. The model solution is for self-checking; your written work is not automatically marked.
A worker presses a clearly marked emergency-stop control after noticing a hazard, but a supervisor later says the worker should have acted sooner. Explain how timing differs from commission and one system question to investigate.
Give me a hint
Classify the action separately from the equipment or procedure that shaped it.
Compare my reasoning
- A timing error concerns when an action occurs relative to the task or hazard.
- A commission error means carrying out an incorrect action; pressing the emergency stop may instead be the correct action taken at a disputed time.
- Check whether the warning, control placement, response protocol and workload made timely action possible.
The concern is about timing, not necessarily commission: the worker used the intended control, and the dispute concerns when. Investigators should examine warning visibility, control design and the response procedure before assigning blame.
Look for these in your work
- Error type distinguished
- Action judged with context
- System factors investigated
Compare one strength and one limitation of Fox et al. (1987) when evaluating a safety reward scheme for a hospital.
Give me a hint
Match the study setting and intervention to the proposed application.
Compare my reasoning
- The study measured workplace safety outcomes over time in two operational mines, making the intervention relevant to real work.
- It bundled several rewards and contingencies in mining; it cannot isolate one component or guarantee transfer to hospital work.
- Hospital measures should reward hazard reporting and safe practice rather than zero reported injuries.
Fox et al. provide applied evidence that a multi-part token economy was followed by fewer lost-time injuries and costs in two mines. The context and bundled programme limit causal precision and transfer to hospitals; any local scheme should protect reporting and measure patient and worker safety directly.
Look for these in your work
- Study strength specified
- Design and transfer limits stated
- Safe adaptation proposed
Design a monitoring plan for repeated minor injuries in a workshop that protects reporting and helps identify system risks.
Give me a hint
Record the circumstances and near-misses, then make reporting safe and usable.
Compare my reasoning
- Create an accessible confidential route for workers to report injuries, near-misses and hazards without penalties for honest reporting.
- Record task, timing, equipment, controls, housekeeping, injury severity and response, using consistent categories.
- Review recurring patterns with workers, correct hazards and monitor both event rates and reporting completeness over time.
A useful system captures circumstances and small events as well as severe injuries, protects honest reporting and feeds results back into hazard correction. Track report completeness and repeat risks so falling counts cannot conceal silence.
Look for these in your work
- Near-misses included
- Context recorded consistently
- Reporting safety and follow-up specified
Common mistakes
- Blaming an operator without investigating controls, displays and conditions
- Treating omission, commission, sequence and timing as complete causal explanations
- Claiming Fox et al. tested a reward in every industry
- Rewarding ‘zero reports’ or discouraging honest incident reporting
- Equating fewer recorded events with safer work
What can you explain now?
Why should an organisation track reporting completeness as well as accident counts?
Compare with the explanation
A fall in recorded events may reflect underreporting rather than fewer hazards; completeness helps interpret the trend.
Separate the underlying event rate from whether events are recorded.
After trying it yourself, choose your next review. This is your self-assessment.
Your review choice appears on Today. Sign in to sync it across devices.
Make it stick
Classify omission, commission, sequencing and timing errors; add system factors when examining a workplace incident. Explain why accident counts need reliable reporting.
If you needed the solution, close it and solve the problem again from a blank page. A correct answer is stronger when you can explain the reason for each step.