Psychology › Section M · Industrial-organizational psychology › Topic M.4

Human factors and workplace design

When a nurse picks the wrong setting on an infusion pump or a pilot pulls the wrong lever, the easy verdict is human error. Human factors psychology asks a better question: what about the design made the error likely, and how can it be redesigned so that people get it right?

Fitting the job to the person

Human factors psychology, the third area of I-O psychology and often called ergonomics, studies how people interact with machines, tools, displays, software and their surroundings, and designs them to fit human abilities and limits. Its guiding idea is to fit the job to the person rather than force the person to fit the job. Physical ergonomics deals with bodies: reach, posture, lifting and strain. Cognitive ergonomics deals with minds: attention, perception, memory, workload and decision-making. Both rely on task analysis, breaking a job into its steps. The field grew from the Gilbreths' motion studies and expanded rapidly in the Second World War, when complex aircraft and radar pushed operators to their limits.

Design lessons from the cockpit

In the 1940s, psychologists studying wartime aircraft found that many crashes blamed on pilot error had a design cause. In some aircraft the switches for the landing gear and the wing flaps looked and felt alike and sat side by side, so tired pilots sometimes raised the wheels instead of the flaps after landing. Alphonse Chapanis proposed shape coding: a small wheel-shaped knob on the landing-gear control and a flap-shaped one on the flap control, so pilots could tell them apart by touch. Paul Fitts and Richard Jones, analysing hundreds of pilot accounts, agreed: confusing controls invite errors that design can prevent.

Physical ergonomics

Physical ergonomics designs workstations, tools and tasks around the human body. Designers use anthropometry, the measurement of body dimensions, and typically aim to fit the middle 90% of users, from the 5th to the 95th percentile, by making chairs, desks and controls adjustable. For computer work, common guidance places the top of the screen at or slightly below eye level and about an arm's length away, with elbows bent at roughly a right angle and feet supported. Poor design contributes to musculoskeletal disorders, such as back pain and repetitive strain injuries, so ergonomists also redesign lifting, reaching and repetitive motions.

Attention and vigilance

Many jobs require monitoring for rare signals: airport baggage screeners, radar operators, lifeguards and security analysts. In the 1940s Norman Mackworth found that people watching a clock-like display for occasional double jumps of a pointer missed more signals as time went on, with a clear decline within about 30 minutes. This vigilance decrement is a reliable finding: sustained attention is effortful and fades with time on task. Designers respond with task rotation, regular breaks and automated alerts. Too many alarms create the opposite problem, alarm fatigue, in which people begin to ignore warnings that are usually false.

Signal detection: hits and false alarms

Signal detection theory describes decisions made under uncertainty. When a signal, such as a weapon in a bag or an attack on a computer network, may or may not be present, there are four outcomes: a hit (signal present, detected), a miss (present, not detected), a false alarm (absent, but reported) and a correct rejection (absent, not reported). An operator's criterion, how much evidence they need before saying "yes", trades misses against false alarms: a lenient criterion, saying "yes" on weaker evidence, catches more real threats but raises more false alarms. Organisations must decide which error is costlier.

Design principles for everyday things

Don Norman popularised principles for designs that people use correctly without instructions. Affordances are the actions an object allows, and signifiers are cues that show where and how to act, such as a flat plate on a door that says "push". Mapping is the relationship between controls and their effects; natural mapping arranges controls to match the layout of what they control. Feedback tells users immediately what their action did. Constraints limit possible actions, such as a plug that fits only one way. When designs break these principles, people make predictable errors, and blaming the user misses the real cause.

Worked example: four burners, four knobs

A kitchen hob has four burners arranged in a square, but its four knobs sit in a straight line along the front. Cooks often turn on the wrong burner. Analyse and fix the design. 1. Identify the principle: mapping. A row of four knobs gives no natural link to a square of four burners. 2. Predict the error: users must guess, so wrong-burner errors are expected even among careful cooks. 3. Weak fix: printed labels, which help but must be read every time. 4. Better fix: arrange the knobs in a square matching the burners. 5. Check: with natural mapping the correct knob is obvious at a glance.

Human error and the Swiss cheese model

James Reason distinguished slips, where the right plan is carried out wrongly, such as pressing the wrong button, from mistakes, where the plan itself is wrong. His Swiss cheese model pictures an organisation's defences, such as design, training, procedures and supervision, as slices of cheese, each with holes that represent weaknesses. Usually a hole in one layer is blocked by another, and an accident happens only when holes in several layers line up. The model shifts the question from "who is to blame?" to "which defences failed, and how can they be strengthened?" It is widely used in aviation, healthcare and the nuclear industry.

Worked example: Three Mile Island

In March 1979 a reactor at the Three Mile Island nuclear plant in Pennsylvania partly melted down. Analyse the human factors as cause and effect. 1. The fault: a relief valve opened as intended but then stuck open, letting cooling water escape. 2. The display: a control-room light showed only that the signal to close the valve had been sent, not whether the valve had actually closed. 3. The operators' belief: reading the light, they assumed the valve was shut and misread other indicators, even reducing emergency cooling. 4. The aftermath: the Nuclear Regulatory Commission required better instrumentation, operator training and control-room design. 5. Lesson: displays should show the system's actual state, not merely the command given.

Checklists, fatigue and automation

Checklists protect against lapses of memory and attention. Aviation checklists became standard after a 1935 crash of a complex new bomber whose pilot forgot a locking control. The World Health Organization's Surgical Safety Checklist (2008) was followed in a 2009 study of eight hospitals by a fall in deaths and complications. Fatigue rules cap how long operators may work; US rules for airline pilots now require rest periods that include a chance for at least eight hours of uninterrupted sleep. Automation can reduce workload, but it can also leave people out of the loop: operators who rarely intervene may trust the system too much and struggle when it fails.

Common mistakes to avoid

First, stopping at "human error" instead of asking what in the design made the error likely. Second, designing for the average person: almost no one is average on every dimension, so designers aim for adjustability across a wide range. Third, mixing up misses and false alarms in signal detection. Fourth, assuming people can stay fully alert on long monitoring tasks; the vigilance decrement says otherwise. Fifth, thinking more alarms always mean more safety; alarm fatigue can make people ignore them.

Key takeaways

Human factors psychology fits tools, displays and tasks to bodies and minds. Wartime cockpit research, such as Chapanis's shape coding, showed that design causes many errors. Physical ergonomics uses body measurements and adjustability. Cognitive ergonomics studies the vigilance decrement and signal detection's hits, misses, false alarms and correct rejections. Norman's principles, including mapping and feedback, make correct use obvious. The Swiss cheese model treats accidents as failures of several defences. Three Mile Island showed that displays must show actual states; checklists, fatigue rules and careful automation add layers of safety.

Flashcards

Flashcard

Human factors psychology

Designing tools, displays and tasks to fit human abilities and limits

Often called ergonomics.

Test your understanding

Question 1 of 5

A cooker's four knobs are in a row, but its four burners form a square, and users often turn on the wrong burner. Which design principle is being broken?

Question 2 of 5

A security analyst flags a harmless login as an attack on the network. In signal detection terms, this is:

Question 3 of 5

Airport screeners miss more prohibited items as their shifts go on. Which response best fits human factors research?

Question 4 of 5

According to Reason's Swiss cheese model, when does an accident occur?

Question 5 of 5

What was the key display problem at Three Mile Island in 1979?

← Previous topicNext topic →