In many fields, outliers are seen as a nuisance. We run tests to justify ignoring them; we explain them away; we resent their intrusion on our neat results. Design thinking, however, asks us to do the opposite—to forgo the blinkers that constrain us to staring at the centre of the bell curve, and to take a good hard look at the outliers.
We don’t do this merely to be contrary: extreme users show us the full range of human interaction with a service or product. As IDEO says in its Field Guide to Human-Centered Design, “without understanding what people on the far reaches of your solution need, you’ll never arrive at answers that can work for everyone.” Also, because extreme users may find that existing solutions fit them poorly, they often design creative workarounds.
But categorizing users as either extreme or mainstream is itself a fraught exercise. Some groups of users come more easily to mind than others—because of stereotypes, prior experiences, varying visibility, or population differences—yet the very creation of those groups involves applying a mental filter to the world, highlighting some features and suppressing others. This automatic reductionism runs counter to the goal of human-centred design, which aspires to see individuals in as much complexity as possible.
The more complexity we include, the harder it becomes to define a norm. For example, consider a story told by Dr. Todd Rose, the director of Harvard’s Mind, Brain, and Education program, in his book The End of Average. He described the United States Air Force’s attempt to design a cockpit for the average pilot—and their subsequent discovery that the average pilot was a mediocre model, because he did not exist.1 It’s relatively easy to find a pilot who is average on one trait, just as it’s easy to find a student with, say, the average GPA. But the more traits we include—the higher-definition a picture we build—the greater the chance that this average-GPA student will be non-average on some trait. Eventually, nobody is average.
Focusing on users we perceive as outliers, then, also helps us see beyond the spectrum that designated them as outliers in the first place. We may find that they have many needs in common with the ‘norm’; these common needs suggest areas for improved design. We may find students vary along some spectrum we never considered, in a different way than they vary on the original spectrum. Indeed, even though Innovation Hub projects have sometimes focused on specific groups of students, our results revealed common themes across projects, and we’ve shared findings that synthesize these themes, most recently in a report we prepared on student mental health.
We are all non-average in many ways, and we may all even be extremes—in some ways and at some times. In her book Mismatch, Kat Holmes, Director of User Experience Design at Google, argued that our ability to interact with the environment fluctuates. Wheelchair users need curb cuts—the slopes that blend sidewalk into street—but so do people with laden shopping carts. One need is more persistent than the other, but that doesn’t change the immediate situation: easily crossing the street or parking lot. A similar scenario is described in the Innovation Hub’s report on accessibility at Convocation (publicly available soon): in one persona, we tell the story of a student who broke his leg, which led to him hobbling down the stairs on crutches when he couldn’t find the elevator.
So, we find diversity when looking at the entire student body, but we also find diversity within ourselves. We interact with the environment one way, in one moment, that may be more difficult in the next. Designing for the full range of collective human experience also helps us design for the full range of individual human experience.
Of course, we cannot keep in mind all 91,286 UofT students separately.2 At some point, individualization becomes overwhelming: there must be a way to design for all without holding 90,000 individual consultations. That’s where needs and personas come in. Breaking the lines that artificially divide groups allows us to see needs that transcend those lines. But needs don’t impact all people equally or in the same way—that’s the very idea of individuality. So we develop personas, detailed stories of individual, imagined users that we compile from our data, to show how the need might manifest in different ways. (To read the student personas created by the Innovation Hub, see our Persona Compendium.)
This way of working subscribes to IDEO’s vision of design thinking as a repetitive cycle of divergence and convergence.3 We diverge to see participants as individuals, then converge upon new, universal themes and tensions. No one convergence point tells the whole story; by converging, we discard details. But iterative cycles of divergence and convergence, in which we find multiple different convergence points, builds our understanding by adding new facets to our model.
Hopefully, this process moves us beyond a simplistic idea of outliers. We don’t design for a few groups of students, then, as an addendum, add on needs for other groups. Instead, we recognize that ‘outlier’ needs are our needs too. In our recent Transforming the Instructional Landscape report, we called this principle “Accessibility Comes First.” We don’t mean accessibility solely for students with specific needs; we mean accessibility for everyone. In one example from the report, wider aisles help wheelchair users, but they also help those of us who prefer not to “squish … through rows with a backpack on your back, umbrellas/bags on the floor, coffee cups/tablets on the tables.”4 We may all be unique, but that uniqueness is a commonality too.
1It was 1950; while women served in the Air Force through the Women in the Air Force program, they would not be accepted as military pilots until 1976.
2Based on Fall 2018 enrolment.
3See the aforementioned Field Guide.