By the end of this topic, you should be able to...
identify where the 5th, 50th and 5th–95th percentiles are appropriate for a design scenario.
Guiding Question
How do ergonomic considerations influence the design of a product?
Did You Know?
The US Air Force in 1950 faced a crisis. Pilots were losing control of aircraft at alarming rates despite no mechanical failures being found. Lieutenant Gilbert Daniels was tasked with investigating. His hypothesis: the cockpit — designed in 1926 around the average of ten body measurements from male pilots — no longer fit anyone well as the pilot population had diversified. Daniels measured 4,063 pilots across ten dimensions. He then counted how many fell within the "average range" — the 30th to 70th percentile — across all ten dimensions simultaneously. The answer: zero. Not one pilot out of 4,063 was average on all ten measurements (Daniels, 1952). The Air Force's response was to mandate adjustable cockpits — and accident rates fell significantly. The lesson was foundational: designing for the average person serves almost nobody.
Why This Topic Matters
Understanding percentiles is the difference between making an educated, evidence-based sizing decision and making an assumption. Every dimension on every product carries an implicit percentile decision — whether the designer made it consciously or not. This topic gives you the conceptual tool to make that decision deliberately, justify it with data, and recognise when a single percentile is insufficient and a range — or adjustability — is required.
What Is a Percentile?
A percentile describes how a data point compares to all data in a set, divided into 100 equal parts (IB DT Glossary, 2024).
In the context of anthropometrics, a percentile tells a designer what proportion of a population has a body dimension at or below a given value.
The 5th percentile for standing height means: 5% of the population is at or below this height. This is a small person.
The 50th percentile for standing height means: 50% of the population is at or below this height. This is the median person.
The 95th percentile for standing height means: 95% of the population is at or below this height. This is a large person.
The percentile range defines the upper and lower limits of the population a product is designed to serve. For a given demographic — gender, race, age — the 50th percentile is the median (IB DT Glossary, 2024).
The Normal Distribution of Anthropometric Data
Anthropometric data, like most biological measurements, follows a normal distribution — a symmetrical bell curve. This means:
The majority of the population clusters around the mean (50th percentile)
Progressively fewer individuals exist at the extremes
The 5th and 95th percentiles sit approximately 1.65 standard deviations below and above the mean respectively
This mathematical property is what makes the 5th–95th percentile range a practical design target — it encompasses 90% of the population within a statistically defined and manufacturable dimensional range.
UK Adult Male Standing Height — Illustrative Data:
Percentile | Approximate Height |
|---|---|
5th | 1,642mm |
50th | 1,740mm |
95th | 1,855mm |
(Pheasant & Haslegrave, 2006)
The range between 5th and 95th percentile is 213mm — a 213mm adjustment range is required to accommodate 90% of adult male stature. No single fixed height can serve this range adequately.
The Three Percentile Design Decisions
The core skill in this topic is not reading a percentile table — it is identifying which percentile to design for, and why, based on the type of measurement and the safety and usability implications of getting it wrong.
There are three fundamental decisions:
1: Design for the 5th Percentile (Small User)
Use when:
The measurement is a reach, minimum force, or minimum capability dimension — where failure to accommodate the smallest or weakest user creates a usability or safety failure.
Logic:
If the 5th percentile user — the smallest, shortest-reaching, or weakest — can successfully use the product, then every larger, taller, or stronger user can also use it.
The constraint is the small user.
Identified design scenarios:
Scenario | Measurement | Why 5th Percentile |
|---|---|---|
Emergency stop button on industrial machinery | Reach distance to button | If the smallest-reaching user can reach it, everyone can. An unreachable emergency stop is a safety failure |
Light switch height in a public building | Vertical reach height | Must be reachable by the shortest user in the population |
Fire exit push bar | Operating force | Must be operable by the weakest user — a child, an elderly person, or a person with reduced strength |
ATM keypad height | Forward and downward reach | Must be reachable by seated wheelchair users (5th percentile eye height and reach) |
Stair handrail height | Grip reach height | Must be reachable by the shortest user requiring support |
Cockpit controls in aircraft | Forward and lateral reach | If the shortest pilot can reach all controls, taller pilots can certainly reach them |
Safety principle: In any scenario where failure to reach, grip, or operate a control creates a safety hazard, design conservatively — for the smallest or weakest user. Never design for the average and assume it is safe enough.
2: Design for the 95th Percentile (Large User)
Use when:
The measurement is a clearance, maximum load, or minimum space dimension — where failure to accommodate the largest user creates a usability or safety failure.
Logic:
If the 95th percentile user — the tallest, widest, or heaviest — fits through, sits in, or operates within the space, then every smaller user also fits. The constraint is the large user.
Identified design scenarios:
Scenario | Measurement | Why 95th Percentile |
|---|---|---|
Doorway height in a public building | Vertical clearance | If the tallest user passes without ducking, all shorter users pass freely |
Legroom in a vehicle or aircraft seat | Knee-to-seat-back clearance | If the longest-legged user fits, all shorter users fit |
Helmet interior volume | Head circumference | If the largest head fits, all smaller heads fit (with padding adjustment) |
Seat width | Hip breadth | If the widest user fits, all narrower users fit |
Structural load capacity of a chair or platform | Body mass | If the heaviest user is safely supported, all lighter users are safely supported |
Overhead clearance on a production walkway | Stature + hard hat | Must clear the tallest worker in the population without risk of head strike |
MRI scanner bore diameter | Shoulder width + body depth | Medical imaging equipment must accommodate the largest patients requiring scanning |
Safety principle: In any scenario where insufficient space, insufficient structural capacity, or physical obstruction creates a safety or usability failure, design for the largest user. Never design for the average and assume large users will manage.
3: Design for the 5th–95th Percentile Range (Adjustable or Range of Sizes)
Use when:
A single fixed dimension cannot simultaneously serve both the smallest and largest users within the target population — meaning a single percentile decision would exclude a significant proportion of the population.
Logic:
Where the dimension must be both large enough for a large user AND small enough for a small user — or must be positioned correctly for users of very different sizes — neither the 5th nor the 95th percentile alone solves the problem. The solution is adjustability or a range of sizes.
This scenario arises when:
The measurement has both a minimum and a maximum functional value
Designing for the 5th percentile would make the product too small for the 95th percentile user
Designing for the 95th percentile would make the product non-functional for the 5th percentile user
Identified design scenarios:
Scenario | Measurement | Why 5th–95th Range |
|---|---|---|
Office chair seat height | Popliteal height | Too low: large user's knees above hip level, poor posture. Too high: small user's feet don't reach floor. Must be adjustable |
Car seat position | Hip-to-pedal distance | Too far: short driver cannot reach pedals safely. Too close: tall driver's knees strike steering column. Sliding adjustment required |
Bicycle saddle height | Leg extension to pedal | Correct saddle height is specific to each rider's leg length. Must be adjustable |
School desk height | Seated elbow height | A class of 30 students spans a significant anthropometric range. Fixed single height serves none of them optimally |
Rifle stock length (military) | Arm length and shoulder geometry | A fixed stock length disadvantages both short and tall soldiers. Adjustable stocks are now standard |
Keyboard tray height | Seated elbow height | Must be adjustable to prevent both elevated and depressed shoulder posture across the user population |
Workbench height in manufacturing | Standing elbow height | A fixed workbench height creates musculoskeletal risk for workers significantly taller or shorter than the design target |
When adjustability is not feasible — Range of Sizes:
For products where continuous adjustability is impractical or too costly, a range of sizes — a selection of sizes that caters for the majority of a market (IB DT Glossary, 2024) — is used to approximate the coverage of adjustability.
Product | Range of Sizes Strategy |
|---|---|
Cycling helmets | XS / S / M / L / XL — each covering a head circumference range |
Personal protective equipment (gloves) | S / M / L / XL — each covering a hand breadth range |
Clothing | Numerical or letter sizing systems covering stature and girth ranges |
Children's furniture | Age-banded sizing (e.g. 3–5 years, 6–9 years) based on age-stratified anthropometric data |
The Problem with Designing for the 50th Percentile
The 50th percentile is the median — it describes the midpoint of the population distribution, not a "typical" or "average" user in any meaningful design sense. Designing to the 50th percentile:
For clearance: excludes the top 50% of the population (all users larger than the median)
For reach: excludes the bottom 50% of the population (all users smaller than the median)
For both simultaneously: serves only users close to the median and systematically excludes everyone at the extremes
As Daniels' 1952 Air Force study demonstrated, designing to the 50th percentile across multiple dimensions simultaneously produces a product that fits almost nobody within the actual user population (Daniels, 1952).
When is the 50th percentile legitimately used?
The 50th percentile is appropriate in limited, specific scenarios:
Scenario | Rationale |
|---|---|
Economic reference point — one-size products where cost prevents adjustability and a range of sizes | A single non-adjustable product inevitably excludes some users; the 50th percentile minimises the average displacement from optimal for all users |
Comparative reference in research and data analysis | The 50th percentile is the median — useful for comparing populations or tracking changes over time, not for primary design decisions |
Products where deviation from the median has symmetrical and minor consequences | A coat hook designed to the 50th percentile height is inconvenient for very tall or very short users — not dangerous |
The key principle: The 50th percentile is almost never the correct primary design target for a safety-critical dimension. It is a statistical reference point, not a design solution.
Worked Example
Scenario
A designer is specifying the height of an emergency stop button on an industrial packaging machine. The target user population is adult workers in a mixed-gender UK manufacturing environment.
Relevant static measurement
Forward functional reach (vertical — minimum reachable height from standing position)
Data (UK adult population, mixed gender, Pheasant & Haslegrave, 2006)
Population | 5th Percentile Forward Reach | 50th Percentile | 95th Percentile |
Adult female | 655mm | 720mm | 790mm |
Adult male | 710mm | 780mm | 855mm |
This is a reach dimension — the button must be reachable by the smallest user.
Design for the 5th percentile — specifically the 5th percentile female, as women represent the smaller end of the combined adult population distribution.
Safety-critical — the emergency stop must be reachable by every worker. This confirms the conservative 5th percentile female decision.
Design specification
Emergency stop button positioned at a maximum height of 655mm from floor level to button centre.
Verification
At this height, all workers with a forward reach of 655mm or greater can operate the button. Since 655mm is the 5th percentile female value, 95% of female workers and virtually all male workers can reach it.
Case Study
London's Crossrail (Elizabeth Line)
Transport for London's Elizabeth line (Crossrail), opened in 2022, provides a rich documented example of deliberate percentile decision-making applied across multiple dimensions in a single infrastructure project (TfL, 2018).
Platform-to-train gap and step - Clearance and Reach
The platform edge height and train floor height were coordinated to minimise both the vertical step (clearance decision) and the horizontal gap (clearance decision).
Both were designed to accommodate the 95th percentile user — including wheelchair users, who require zero-step, minimal-gap boarding. The design target was a maximum 50mm gap and 15mm step — achievable for 95% of rolling stock positions (TfL, 2018).
Ticket gates — Clearance
Gate width was designed for the 95th percentile shoulder width — wide enough for the broadest user in the population to pass without turning sideways. Accessible gates were designed for 95th percentile wheelchair width — a separate, larger clearance dimension that is also the ADA and UK Building Regulations minimum.
Handrail heights — Reach
Vertical handrails on escalators and platforms were positioned to the 5th percentile grip reach height — ensuring the shortest users can comfortably grip the rail without over-reaching, while taller users grip the same rail at a lower point on their arm, which remains ergonomically acceptable.
Seating — Adjustability not feasible; 50th percentile compromise
Fixed platform seating cannot be adjustable. TfL specified seat heights close to the 50th percentile popliteal height — acknowledged as a compromise that serves the median user adequately and creates only minor inconvenience (not safety risk) for users at the extremes.
The Elizabeth line case demonstrates the systematic nature of percentile decision-making in complex design projects. Different dimensions within the same product or environment require different percentile decisions — and those decisions must be explicitly justified against the safety and usability consequences of each dimension, not applied uniformly.
Decision Guide
Design Dimension Type | Design For | % Population Served | Consequence of Wrong Decision |
|---|---|---|---|
Clearance (space, height, width, load) | 95th percentile (large user) | 95% accommodated | Large users excluded, risk of entrapment, injury, or inability to use product |
Reach (access to controls, handles, switches) | 5th percentile (small user) | 95% accommodated | Small users excluded, safety-critical controls unreachable |
Bidirectional (must fit AND reach) | 5th–95th range (adjustable or range of sizes) | 90% accommodated | Either large users cannot fit OR small users cannot reach — both failure modes exist simultaneously |
50th percentile reference only | Median user | 50% above, 50% below | Acceptable only for low-consequence, non-safety-critical dimensions |
Key Vocabulary
Term | Definition |
|---|---|
Percentile | A term that describes how a data point compares to all data in that set, divided into 100 equal parts |
Percentile range (upper and lower limits) | That proportion of a population with a dimension at or less than a given value; for a given demographic, the 50th percentile is the median |
Adjustability | The ability of a product to be changed in size, commonly used to increase the range of percentiles for which a product is appropriate |
Range of sizes | A selection of sizes a product is made in that caters for the majority of a market |
Anthropometrics | The aspect of ergonomics that deals with body measurements |
Static data | Human body measurements when the subject is still |
Dynamic data | Human body measurements taken when the subject is in motion |
Clearance | The physical space between two objects |
Reach | The range that a person can stretch to touch or grasp an object from a specified position |
Workspace envelope | A 3D space that is typically physical and/or virtual that needs to have defined permissible boundaries of movement and operation |
Ergonomics | The application of scientific information concerning the relationship between human beings and the design of products, systems, and environments |
Practice Questions
Question 1
A designer is specifying the height of a handrail for a public staircase. Identify the appropriate percentile to use for this dimension and explain your reasoning. (2 marks)
Question 2
A designer is specifying the interior headroom of a passenger car. Identify the appropriate percentile to use and explain why designing to the 50th percentile would be insufficient. (3 marks)
Question 3
Identify the appropriate percentile for the following dimensions on a public ATM cash machine, and justify each decision: (6 marks) (a) Screen height (visual access) (b) Keypad reach distance (c) Card slot clearance width
Sources
IB Design Technology Guide (First Assessment 2027)
Daniels, Gilbert S. The "Average Man"? Technical Note WCRD 53-7, Wright Air Development Center, Wright-Patterson Air Force Base, Ohio, 1952.
Pheasant, Stephen, and Christine Haslegrave. Bodyspace: Anthropometry, Ergonomics and the Design of Work. 3rd ed., CRC Press / Taylor & Francis, 2006.
Rose, Todd. The End of Average: How We Succeed in a World That Values Sameness. HarperCollins, 2016. (Accessible account of Daniels' research and its implications for design and beyond.)
Transport for London. Crossrail Accessible Design Guide. Transport for London, 2018, www.tfl.gov.uk/corporate/publications-and-reports/crossrail-design.
Rose, Todd. The End of Average. HarperCollins, 2016. (Highly recommended for students — accessible, compelling account of why designing for the average person fails almost everybody.)
Pheasant, Stephen, and Christine Haslegrave. Bodyspace. 3rd ed., CRC Press, 2006. (Chapter 2 provides the definitive practitioner guide to percentile application in product design.)
HFES/ANSI 100. Human Factors Engineering of Computer Workstations. Human Factors and Ergonomics Society, 2007. (Professional standard demonstrating percentile-based workstation design in practice.)
Linking Questions
How are user-centred research methods used to collect human factor data? (A2.1)
Which aspects of ergonomics are appropriate for user-centred design (UCD) practice? (B1.1)
How does ergonomics affect modelling and prototyping of potential design solutions? (B2.2)
How important is ergonomics to inform effective inclusive design? (C1.2)