Human Factors, Fatigue and Anthropometrics

1. Why This Topic Matters for the CSP Exam and for Practice

Human factors is the bridge between the technical and the human sides of safety. Most serious incidents are not caused by a single equipment failure; they are the product of people interacting with poorly designed tasks, tools, schedules and environments. BCSP places Human Factors, Fatigue and Anthropometrics inside the Occupational Health and Applied Science domain because a Certified Safety Professional is expected to design work that fits the worker rather than forcing the worker to fit the work. On the exam you will see this material in three forms: (1) conceptual questions about human error, system design and fatigue science, (2) quantitative questions involving percentiles, standard deviations and the NIOSH Lifting Equation, and (3) scenario questions asking you to choose the best control or design strategy. Mastering this topic also pays off across other domains, because fatigue, ergonomics and human error show up in incident investigation, risk management and management systems questions as well.

2. What Human Factors Is

Definition. Human factors (used interchangeably with ergonomics in most CSP contexts) is the scientific discipline concerned with understanding interactions among humans and other elements of a system, and the application of theory, principles, data and methods to design in order to optimize human well-being and overall system performance. The key phrase is fit the job to the person. The field has three traditional branches:
- Physical ergonomics: anthropometry, biomechanics, posture, repetitive motion, manual materials handling, workstation layout.
- Cognitive ergonomics: mental workload, decision making, perception, memory, human error, display and control design, situational awareness.
- Organizational (macro) ergonomics: shift schedules, staffing, communication, teamwork, safety culture, fatigue risk management.

Systems view of error. The modern position, associated with James Reason, is that human error is a consequence, not a cause. Errors occur because of latent conditions in the system (poor design, inadequate training, production pressure) combined with local triggers. Reason's Swiss Cheese Model describes layered defenses, each with holes; an accident occurs when the holes align. The practical implication for exam questions: when asked how to prevent recurrence of an error, prefer answers that change the system (design, procedures, workload, schedule) over answers that simply retrain or discipline the individual.

Taxonomy of human error (Reason).
- Slips: attention failures during a routine, well-practiced task (pressing the wrong button while distracted). The plan was right; the execution was wrong.
- Lapses: memory failures (skipping a step, forgetting to re-install a guard).
- Mistakes: the action matched the plan, but the plan was wrong. Rule-based mistakes apply the wrong rule or misapply a good rule; knowledge-based mistakes occur in novel situations when the person reasons incorrectly.
- Violations: intentional deviations from rules. Routine violations are habitual shortcuts, situational violations occur because the rule cannot be followed as written, exceptional violations are one-off departures, and optimizing violations are done for personal gain or thrill.
Slips and lapses are skill-based errors; mistakes are rule-based or knowledge-based. This links to Rasmussen's Skill-Rule-Knowledge (SRK) framework of performance levels. Remember the control implications: slips and lapses respond to design fixes (forcing functions, interlocks, checklists, reducing distraction); mistakes respond to training, decision aids and better information; violations respond to addressing motivation, supervision, realistic procedures and culture.

Other models you should recognize.
- SHELL model (Software, Hardware, Environment, Liveware, Liveware): the human is at the center and mismatches at each interface create error potential.
- HFACS (Human Factors Analysis and Classification System): Reason's model operationalized into four tiers, which are unsafe acts, preconditions for unsafe acts, unsafe supervision, and organizational influences.
- Performance Shaping Factors (PSFs): anything that influences the probability of error, such as time pressure, stress, fatigue, task complexity, experience, ergonomics, procedures and environment. PSFs are the adjustment knobs in Human Reliability Analysis (HRA) methods such as THERP, HEART and SPAR-H.
- Human Error Probability (HEP) = number of errors / number of opportunities for error. Generic values are often cited: roughly 0.001 for simple, routine, well-designed tasks, rising to 0.1 or higher under severe stress or novel conditions.

Information processing and workload. Humans perceive, decide and act. Each stage can fail. Concepts the exam may touch:
- Signal detection: hits, misses, false alarms, correct rejections; vigilance decrement occurs after roughly 20 to 30 minutes of monitoring.
- Hick's Law: reaction time increases logarithmically with the number of choices.
- Fitts' Law: time to reach a target depends on distance to and size of the target; bigger, closer targets are faster and more accurate.
- Working memory: limited to roughly 5 to 9 chunks; procedures should not rely on remembering long strings.
- Yerkes-Dodson: performance is best at moderate arousal; both boredom and overload degrade it.
- Mental workload is assessed with tools such as NASA-TLX.

Display and control design principles.
- Population stereotypes: expected direction of control motion (clockwise to increase, up or right for on, in North America). Violating them causes slips, especially under stress.
- Compatibility: spatial (controls laid out like the displays they govern), movement (control direction matches display direction), conceptual (red means stop or danger).
- Coding of controls by shape, size, color, location, texture and labeling; shape coding lets a user identify a control by touch.
- Guarding and interlocking critical controls against accidental activation.
- Displays: qualitative (check reading) versus quantitative; moving pointer on fixed scale is generally best for check reading; place critical displays in the primary visual field (within about 15 degrees of the line of sight); avoid glare; use appropriate contrast and character size.
- Alarms: avoid alarm flooding; prioritize; use distinct auditory signals for different urgency levels.

3. Anthropometrics

Definition. Anthropometry is the measurement of the physical dimensions and capabilities of the human body: stature, sitting height, eye height, reach, hand breadth, hip breadth, weight, and functional measures like grip strength. It is the quantitative foundation for designing workstations, clearances, PPE, tools, seating and vehicle cabs.

Static versus dynamic (functional) anthropometry. Static measurements are taken with the body in fixed standard postures (standing or seated). Dynamic measurements describe the body in motion (functional reach envelope, range of motion). Real tasks are dynamic, so designers apply corrections for clothing, shoes, PPE and posture (for example, add roughly 1 inch for shoe heel height; add bulk for winter clothing or chemical suits).

Percentiles and the normal distribution. Body dimensions are approximately normally distributed within a population. A percentile indicates the percentage of the population at or below that value. The exam expects you to convert between mean, standard deviation and percentiles using Z-scores:
- 5th percentile = mean minus 1.645 SD
- 95th percentile = mean plus 1.645 SD
- 2.5th and 97.5th percentiles = mean plus or minus 1.96 SD
- 1st and 99th percentiles = mean plus or minus 2.33 SD
- 50th percentile = the mean (for a normal distribution)
Worked example: Male stature has a mean of 69.0 inches and SD of 2.8 inches. The 95th percentile is 69.0 + 1.645 x 2.8 = 69.0 + 4.6 = 73.6 inches. The 5th percentile female stature with mean 64.0 and SD 2.6 is 64.0 minus 1.645 x 2.6 = 64.0 minus 4.3 = 59.7 inches.

Three design strategies (the heart of anthropometric exam questions).
1. Design for the extremes. Use the large extreme (often 95th percentile male) for clearances: door heights, access openings, leg room under a desk, escape hatches. Use the small extreme (often 5th percentile female) for reach: location of controls, shelf heights, emergency stop buttons, grasp distances. Also use the large extreme for strength limits of structures (ladder and anchorage ratings) and the small extreme for required force (a 5th percentile female must be able to actuate the control).
2. Design for adjustability. Preferred whenever feasible; typically accommodates the 5th percentile female through the 95th percentile male (seat height, work surface height, monitor height, armrests). This is the best answer when the question involves a shared workstation or a range of users.
3. Design for the average. Usually the wrong answer; the mythical average person does not exist, because someone at the 50th percentile for stature is rarely at the 50th percentile for arm length, sitting height and so on. It is acceptable only for low-criticality items where adjustability and extremes are impractical (checkout counter height, some public seating).

Why 5th and 95th and not 1st and 99th? Designing for the central 90 percent captures most users at reasonable cost; extending to 99 percent dramatically increases cost and size. For safety-critical dimensions (emergency egress, machine guard openings), designers may go wider. Note that multivariate accommodation shrinks the percentage actually fit: a design accommodating 5th to 95th percentile on several independent dimensions fits noticeably fewer than 90 percent of people.

Common anthropometric data sources. ANSUR (US Army), NHANES (CDC civilian data), CAESAR (3D scan data), and older sources such as Dreyfuss' The Measure of Man and Woman. Be aware that military data tend to be younger and fitter than the industrial population, and that secular trends (populations getting taller and heavier) make old data suspect.

Workstation design rules of thumb.
- Standing work surface height: slightly below elbow height for light work, 2 to 4 inches below elbow height for heavy work, slightly above elbow for precision work.
- Seated work: elbows at roughly 90 degrees, forearms parallel to the floor, feet flat or on a footrest, thighs roughly parallel to the floor, lumbar support, top of monitor at or slightly below eye height, monitor at about arm's length.
- Keep frequently used items within the normal reach envelope (forearm sweep with upper arm at the side) and all items within the maximum reach envelope (full arm extension) for the 5th percentile user.
- Avoid work above shoulder height, below knee height, sustained twisting, and contact stress from sharp edges.
- Provide anti-fatigue mats and sit-stand options for prolonged standing.

Hand tool design. Keep the wrist neutral (bend the tool, not the wrist), handle diameter about 1.25 to 2 inches for power grips, handle length at least 4 to 5 inches so the end does not press into the palm, avoid finger grooves that fit only one hand size, minimize vibration and pinch points, allow use by either hand, and reduce trigger force and single-finger triggering.

4. Ergonomic Risk Factors and Assessment Tools

Primary risk factors for work-related musculoskeletal disorders (WMSDs): force, repetition, awkward or static posture, duration, contact stress, vibration, cold, and insufficient recovery. Combinations multiply risk. Psychosocial factors (job strain, low control, high demand) also increase MSD reporting. Common WMSDs include carpal tunnel syndrome, tendinitis, epicondylitis, trigger finger, rotator cuff disorders, low back pain and hand-arm vibration syndrome.

Revised NIOSH Lifting Equation (1991/1994).
RWL = LC x HM x VM x DM x AM x FM x CM, where
- LC (load constant) = 51 lb (23 kg)
- HM (horizontal multiplier) = 10/H (H in inches, measured from ankle midpoint to hand center) or 25/H in centimeters
- VM (vertical multiplier) = 1 minus 0.0075 x the absolute value of (V minus 30) in inches, or 1 minus 0.003 x the absolute value of (V minus 75) in centimeters
- DM (distance multiplier) = 0.82 + 1.8/D in inches, or 0.82 + 4.5/D in centimeters
- AM (asymmetry multiplier) = 1 minus 0.0032 x A, where A is the twist angle in degrees
- FM (frequency multiplier) from a table based on lifts per minute, duration and vertical height
- CM (coupling multiplier) = 1.0 good, 0.95 fair, 0.90 poor (0.95 or 1.0 for fair depending on V)
Lifting Index (LI) = actual load weight / RWL. LI of 1.0 or less is acceptable for nearly all healthy workers; LI between 1 and 3 indicates increased risk for some workers and the task should be redesigned; LI above 3 indicates high risk for many workers. The equation applies only to two-handed, smooth, unrestricted lifts in a reasonable environment; it does not apply to one-handed lifts, lifting while seated or kneeling, lifting in restricted spaces, lifting unstable objects, high-speed lifting, hot or cold environments, or when combined with pushing, pulling or carrying. Ideal conditions (H = 10 in, V = 30 in, D less than or equal to 10 in, A = 0, low frequency, good coupling) yield RWL = 51 lb. The horizontal distance is almost always the dominant factor: moving the load closer to the body is the most effective single improvement.

Other tools you should be able to identify by name and purpose: RULA (upper limb, seated and standing postures, scores 1 to 7), REBA (whole body including legs and load, scores 1 to 15), Strain Index (distal upper extremity, multiplies intensity, duration, efforts per minute, posture, speed and duration per day), Snook and Ciriello (Liberty Mutual) tables for push, pull, carry and lift acceptable to a percentage of the population, OWAS for postural load, ACGIH TLVs for Hand Activity Level and for lifting, and OSHA checklist-based symptom surveys. Also know that OSHA has no specific ergonomics standard; citations are issued under the General Duty Clause, Section 5(a)(1).

Hierarchy of ergonomic controls: engineering controls (redesign the task, mechanical assists, adjustable workstations, tool redesign) are preferred, followed by administrative controls (job rotation, rest breaks, pacing, training, stretching programs) and finally PPE (which has limited value in ergonomics; back belts are not recommended by NIOSH as a control). Participatory ergonomics, early symptom reporting and medical management are elements of an effective program.

5. Fatigue

Definition. Fatigue is a state of reduced mental or physical performance capability resulting from sleep loss, extended wakefulness, circadian phase disruption or excessive workload. It is a physiological state, not a motivational one; it cannot be overcome by willpower, training or incentives.

Types.
- Acute fatigue: from a single period of extended wakefulness or a hard day; relieved by one good night's sleep.
- Cumulative (chronic) fatigue: builds from repeated insufficient sleep (sleep debt); requires multiple recovery nights (generally at least two consecutive full nights).
- Circadian fatigue: performance decrements that occur at predictable points in the 24-hour biological cycle regardless of sleep.
- Physical versus mental fatigue: muscular fatigue from exertion versus cognitive fatigue from vigilance or decision-making demands.

Sleep science basics. Adults need 7 to 9 hours of sleep. Sleep occurs in roughly 90-minute cycles of light sleep, slow-wave (deep) sleep and REM sleep; slow-wave sleep restores physical function, REM consolidates memory and learning. Sleep debt accumulates: losing 2 hours per night for a week produces impairment similar to a full night of sleep deprivation. The circadian rhythm, driven by the suprachiasmatic nucleus and entrained by light, produces two performance lows: the major trough between about 0200 and 0600 and a smaller post-lunch dip around 1300 to 1500. Body temperature, alertness and reaction time all follow this cycle.

Fatigue and impairment equivalence. Dawson and Reid (1997) found that after 17 to 19 hours of wakefulness, cognitive psychomotor performance is equivalent to a blood alcohol concentration of about 0.05 percent; after 24 hours awake it is equivalent to about 0.10 percent. This comparison is frequently used on exams and in safety communications.

Effects of fatigue on performance: slowed reaction time, impaired attention and vigilance, degraded memory and decision making, increased risk-taking and complacency, reduced communication, irritability, tunnel vision, microsleeps (brief involuntary episodes of sleep lasting seconds), and poor self-assessment (fatigued people reliably underestimate their impairment). Accident rates rise on night shifts, on the third or fourth consecutive night shift, during hours 9 through 12 of a shift, and roughly double after 12 hours on duty. Long commutes after night shifts create a serious drowsy-driving risk.

Shift work design principles.
- Forward (clockwise) rotation (day to evening to night) is better tolerated than backward rotation because the body adapts more easily to delays than to advances.
- Rapid rotation (2 to 3 shifts before changing) or slow rotation (3 to 4 weeks) are both defensible; the worst is weekly rotation, which keeps workers perpetually out of phase.
- Limit consecutive night shifts (preferably no more than 3 or 4), and limit shift length to 12 hours with longer shifts reserved for lower-demand work.
- Provide at least 11 hours off between shifts (10 hours is a common regulatory minimum) and at least 2 consecutive days off after a block of nights.
- Avoid quick returns (short turnarounds), excessive overtime and unpredictable schedules; limit weekly hours (fatigue risk climbs sharply above 48 to 60 hours).
- Schedule the most critical and monotonous tasks away from circadian lows.
- Regulated examples: FMCSA hours of service (11 hours driving within a 14-hour window, 10 hours off, 30-minute break after 8 hours of driving), FAA Part 117 flight duty limits, NRC Part 26 fatigue rules for nuclear workers, and API RP 755 for refinery and petrochemical shift workers.

Fatigue Risk Management System (FRMS). A data-driven, risk-based management system that treats fatigue like any other hazard. Core elements: policy and management commitment; fatigue risk assessment (hours of work analysis, biomathematical models such as FAST or SAFTE, incident data); controls at multiple levels (schedule design, staffing, workload, nap policies, lighting, task scheduling, fitness-for-duty checks); fatigue incident reporting and investigation; education for workers and supervisors; sleep disorder screening (obstructive sleep apnea is common and treatable); and continuous improvement with performance indicators. A defense-in-depth model for fatigue has five levels: adequate opportunity for sleep (schedule), actual sleep obtained (personal responsibility and education), behavioral symptoms (self and peer monitoring), fatigue-related errors (reporting and proximal controls), and fatigue-related incidents (investigation and feedback).

Countermeasures. The only true cure is sleep. Short-term aids include strategic naps (20 to 30 minutes to avoid sleep inertia, or a full 90-minute cycle), caffeine (about 100 to 200 mg, effective in 15 to 30 minutes, avoid within 6 hours of planned sleep), bright light exposure during night work and light avoidance before day sleep, physical activity, task variety, social interaction and controlled rest breaks. Workers should be trained in sleep hygiene (consistent schedule, dark cool quiet room, limit alcohol and screens before bed). Supervisors should be trained to recognize signs and to make fitness-for-duty decisions without punitive consequences for honest reporting.

Environmental contributors: heat stress, noise, vibration, poor lighting and glare all accelerate fatigue and should be addressed alongside scheduling. Whole-body vibration from vehicles and equipment is both a fatigue and a back-injury factor.

6. How the Pieces Fit Together

A fatigued worker on the night shift, using a control panel that violates population stereotypes, at a workstation sized for the average male, is a textbook stack of latent conditions. The human factors professional addresses each layer: redesign the panel for compatibility, make the station adjustable for the 5th percentile female through the 95th percentile male, schedule forward-rotating shifts with adequate recovery, and build an FRMS so the organization learns from fatigue-related near misses. Exam questions frequently present such composite scenarios and ask for the most effective or the first action; the answer is almost always the engineering or design change that removes the error opportunity rather than the administrative control that relies on vigilance.

7. Exam Tips: Answering Questions on Human Factors, Fatigue and Anthropometrics

Tip 1: Know the percentile Z-values cold. 1.645 for 5th/95th, 1.96 for 2.5th/97.5th, 2.33 for 1st/99th. Most anthropometric calculations are just mean plus or minus Z times SD. Read carefully whether the question asks for the small or large extreme, and whether the population is male, female or combined.

Tip 2: Clearance uses the large person, reach uses the small person. If the question involves fitting through, under, over or into something, design for the 95th percentile male. If it involves grasping, pressing, seeing over or reaching, design for the 5th percentile female. If the question involves a shared workstation or chair, choose adjustability. Treat designing for the average as a distractor unless the item is non-critical and no other option is feasible.

Tip 3: Classify the error before choosing the fix. Determine whether the scenario describes a slip, lapse, mistake or violation. Slips and lapses point to design and forcing functions; mistakes point to training, procedures and decision support; violations point to supervision, realistic rules and culture. Retraining alone is rarely the best answer for a slip.

Tip 4: Think system, not blame. BCSP exam items reflect Reason's systems view. When options include disciplining the worker versus redesigning the task, redesign wins. When options include adding a warning sign versus eliminating the hazard through design, design wins. Apply the hierarchy of controls to ergonomics just as you would to chemical or physical hazards.

Tip 5: For NIOSH Lifting Equation problems, write the formula first. Note the units (inches versus centimeters) because the constants change. Remember LC = 51 lb or 23 kg, that multipliers are always 1.0 or less, and that LI = load divided by RWL. Check for exclusion conditions (one-handed lift, seated, extreme temperature) that make the equation inapplicable; that can itself be the question. When asked for the most effective improvement, reducing horizontal distance (bringing the load closer) is usually the correct answer.

Tip 6: Memorize the fatigue benchmarks. 17 to 19 hours awake equals about 0.05 percent BAC; 24 hours equals about 0.10 percent. Circadian low is 0200 to 0600 with a secondary dip early afternoon. Adults need 7 to 9 hours. Forward rotation is preferred. Allow at least two full nights to recover from a block of night shifts. Naps should be 20 to 30 minutes or a full 90-minute cycle to avoid sleep inertia.

Tip 7: Watch for absolute words and best-answer phrasing. Questions often ask for the most effective, first, or best option. Several choices may be correct; pick the one that addresses root cause, is highest on the hierarchy of controls, or is most consistent with the systems approach. Options containing words like always, never or eliminates all risk are usually wrong.

Tip 8: Recognize assessment tools by scope. RULA is upper body; REBA is whole body; Strain Index is distal upper extremity; NIOSH equation is two-handed lifting; Snook tables cover push, pull and carry as well as lift; NASA-TLX is mental workload; HFACS is error classification in investigations. If a question describes pushing a cart, the NIOSH equation is not the right tool.

Tip 9: Remember the regulatory landscape. There is no OSHA ergonomics standard; the General Duty Clause applies. Fatigue rules exist in specific sectors (FMCSA, FAA, NRC, API RP 755), and ANSI/ASSP, ACGIH TLVs and NIOSH publications provide consensus guidance. A question asking which standard OSHA would cite for ergonomic hazards is testing whether you know the answer is 5(a)(1).

Tip 10: Use population stereotypes and compatibility principles for display and control questions. Clockwise increases, up is on, red is stop or danger, controls should be arranged to match the layout of what they control, and critical controls should be guarded against inadvertent actuation. Shape coding is the answer when the user must identify a control without looking.

Tip 11: Manage your own fatigue on exam day. The CSP exam is long. Sleep well for at least two nights before, avoid scheduling the exam at your personal circadian low, bring water, and use the on-screen break strategically. Apply the science to yourself.

Tip 12: Practice conversion and arithmetic quickly. Many candidates lose points on simple math under time pressure. Practice computing mean plus or minus 1.645 SD, computing each NIOSH multiplier, and dividing load by RWL until it is automatic. Write intermediate values down rather than holding them in working memory, which is exactly what the human factors literature would advise.

8. Quick Review Checklist
- Human factors fits the job to the person; error is a consequence of system design.
- Slips and lapses are skill-based; mistakes are rule- or knowledge-based; violations are intentional.
- Swiss Cheese, SHELL, HFACS, SRK, PSFs and HEP are the vocabulary of human error.
- Anthropometry: static versus dynamic; 5th percentile female for reach, 95th percentile male for clearance, adjustability preferred, average avoided.
- Z-values: 1.645, 1.96, 2.33.
- NIOSH: RWL = 51 x HM x VM x DM x AM x FM x CM; LI = load/RWL; LI above 1 means redesign, above 3 means high risk.
- Fatigue: acute, cumulative, circadian; 17 to 19 hours awake equals 0.05 percent BAC; forward rotation; FRMS with defense in depth.
- Controls: engineering first, administrative second, PPE last; no OSHA ergonomics standard, use the General Duty Clause.