The myth of the average
The simple act of summing discrete measurements and dividing its total by the number of observations was elevated to the status of a new science by Adolphe Quételet, a Belgian astronomer. While this was common practice amongst his fellow astronomers, in the mid-nineteenth century Quételet championed the application of the "average" magical properties to more worldly affairs. By the time he coined the term "average man" he had already calculated the average chest circumference of the Scottish soldiers and created the popular Quételet Index, an older parent of our well-known body mass index.
Rivaling in popularity with other nineteenth-century scientific colossi of the likes of Darwin, the Belgian astronomer's body of work had a long-lasting influence in Europe and across the pound. His principles guided the standardization of northern uniforms during the US civil war and later the design of the first fighter aircrafts. These used the dimensions of the average US soldier as a blueprint to define cockpit dimensions, following Quételet's principles to determine the one size that will fit all — or at least the most.
When the rate of aircraft crashes surged shortly before US participation in WWII, Lt. Gilbert S. Daniels, at the time a recent graduate, was called to duty with the mission of solving this obvious problem by updating the measurements on the average US fighter pilot. What Lt. Daniels found out in the process of measuring more than 4063 pilots on 140 dimensions was that, when analyzing individual data, not even one soldier was close to average in all the 10 dimensions ergonomically relevant in operating the aircraft. Thus, as some might know from the familiar struggle inside fitting rooms, to design for the average is to design for no one.
Dr. Todd Rose, a Harvard scholar and author of the book The End of Average, has been calling attention to the pernicious role that the "average" reasoning has been playing in such diverse areas as product design, social policies, and standardized testing.
"We became so used to the concept as a measuring and sorting tool, that it and its correlates — below-average, above-average — are everyday speech. We don't even question the language, although the challenges we face require a different mindset." — Dr. Todd Rose
Notable examples of this faulty mindset can still be found today. Below are examples from two areas that have been the focus of my research: Transportation and Immersive Systems (Virtual/Augmented Reality).
Designing seats that are efficient for all
The paradigmatic example from transportation comes from road vehicles passive safety systems, such as restraint systems and airbags. These have been tested primarily using crash test dummies representing the human male 50th percentile. Conveniently called EvaRID, the first female finite element dummy model, developed by Dr. Astrid Linder's team at Chalmers, was only presented as late as in 2012 — making worldwide headlines not so much because of its progress beyond the state-of-the-art, but because few thought that safety science was still at such a state.
Studies using EvaRID have been confirming that females have a 1.5 to 3 times higher risk than males in whiplash injuries from rear-end impacts. It has been hypothesized that the reason for this lies in the long history of overlooking the representativeness of different body types in the design of in-vehicle systems. The problem is not limited to female body types but also to children, elderly, and people in a temporary special condition such as pregnancy. This is a real concern as estimates indicate that 300,000 EU citizens suffer whiplash injuries annually.
With the advent of new interior concepts that might enable unconventional seat positions made possible by vehicle's higher autonomy levels, the need to disaggregate data regarding age, gender, and more importantly body type will be paramount to develop seat concepts and safety systems that are efficient for all.
Immersive systems (VR and AR) are designed by men to men
Similarly, in the realm of immersive systems development, new systems are being designed taking into account measurements of one particular group of users — the human male with a knack for gaming. Focusing on one particular group of users is a risky decision for one industry that so far failed to deliver on the promises of full technology disruption — VR only accounts for 2.4% of the global gaming market. Simulation motion sickness (MS), or the sensation of dizziness and nausea that follows the use of a VR system, may affect 25% to 40% of the users.
While the main etiological factors of simulation MS are still the subject of an interesting scientific debate, there are two things that we already know:
- One of the main reasons for MS is a conflict between what our eyes are seeing and what our vestibular system (responsible for our sense of head position and orientation) is signaling, mainly due to latencies in the processing and projection of the visual virtual environment during the user's head movements.
- Women are four times as susceptible to MS as men.
Thomas Stoffregen hypothesizes that these gender differences might be due to lower centers of gravity of some users, which causes them to sway more — which is correlated with higher rates of MS.
Nevertheless, other researchers have come forward with simpler explanations. Bas Rokers and colleagues found out that 90% of women have an interpupillary distance less than the default headset setting, while for 27% of women's the headset fit was completely off their eye-position. When this design feature is controlled, by adjusting the interpupillary distance of the device to the individual user, the MS symptoms appear to be attenuated.
Regardless of the technological substrate, UX and Usability researchers should be aware of unconscious bias when defining their user-testing samples and be watchful for how this will resonate through all the product development chain. Often we go for convenience sampling, recruiting participants from the little cosmos where we move — the university campus or the company facilities — unaware that this is just a small subset of our universe of users. Safety and usability will definitely improve if we design to accommodate the largest possible number of users instead of that intangible average one.