Let’s set the scene: it’s a clear and brisk Monday morning in Los Angeles, sometime late in the 1980s.
A group of young adults and teenagers have been gathered in a nondescript conference room with the promise of $50 for a morning of their time.
They’ve spent the last three hours drinking coffee, sodas, and giving responses to questions to the smartly dressed man they’ve just met, and a woman who’s not been talking, but taking notes on her clipboard.
The questioning moves towards getting reactions from a series of photo-realistic renderings of possible designs for an upcoming range of Sony boomboxes.
As well as a range of different shapes, each body is also shown in either a traditional black, but also with casings of a bright yellow.
The consensus is clear. People like the yellow boomboxes, with people giving positive feedback that they’re more vibrant, distinct, fun, futuristic, and fashionable.
Then the group runs through an activity where they guess the price of two boomboxes, identical except for color. The yellow option receives a higher hypothetical price given the premium of it’s styling.
The smartly dressed man then thanks them for their time, and in addition to the $50 for their time, as they exit they’re presented with a stack of boxes on each side of the door, each box containing a soon to be released Sony boombox. On the left is a stack of boxes containing a black boombox, and on the right, boxes containing a yellow boombox.
The group files out picking whichever color they prefer. To the surprise of the researchers, the majority of the group takes the black boombox, with only a few people choosing the new yellow option.
What happened here?1 Why did the group act in a way that seems counter preferences that they’ve voiced only moments before?
For anyone who’s involved in the creation of new products, this is an important question to appreciate.
The customary response to explain the difference between what people say and what people do is the recognition of a range of phenomena described by people who study the human mind, and human behavior psychologists, sociologists, and economists.
These range from cute effects like pareidolia (the tendency to see faces in inanimate objects :D) to effects with potentially more serious outcomes like the bandwagon effect (our tendency to more readily agree or believe something where it’s demonstrated already share that belief), to cognitive bias of how participants (and facilitators!) think, to experimental bias like the number of people that we’re involved in the research, and how they were found.
Study of these effects is fascinating2, and awareness of these effects will undoubtably make you a better designer, but I’m uneasy using ‘bias’ as a catchall term for these effects, not only because it side-steps the bigger picture of what’s happening, but also because I see designers underestimate bias as a minor nuisance, or an intellectual curiosity.
When we perform research to learn about people, or how people use a product, we typically conduct the research within a controlled environment that assists our observation or interaction with the people we’re learning about.
Through this act of controlling this environment, we ultimately make the environment artificial, and make an unspoken premise that what people think, say and do in these artificial environments will have relevance in the greater world.
We wouldn’t observe behavior of characters in The Sims as a representation of how people act in real life, but at times will place people in situations or environments that control and impact their behavior so greatly that we might as well be running a controlled computer simulation.
To capture the significance of the difference between these research environments and the real world, I like to frame research as not something that happens directly in our world, but instead occurs in a rich simulation of reality. This then invites some healthy questions:
Here’s a specific type of simulation that’s worth it’s own mention. The acronym WEIRD is used in social sciences as a reminder that much psychology, and cognitive research is performed on undergraduates in North American universities resulting in a sample of people who are predominantly Western (white), educated, industrialized, rich, and from democratic countries (not to mention likely to be in their early to mid 20’s).
What generalizations do we learn from studying this group of people, and how do they relate to not only the wider population of the Americas, but to a global scale?
Psychologists are not the only ones to make the mistake of sampling a specific population, in design research it’s all too easy to put together a quick survey to send to friends, colleges, and family.
Given the gaps of gender, race, age, and technical literacy of people working in fields of design and technology compared to the wider population, it’s likely that studies of this kind, while quick and convenient, are also their own type of weird.
If your manage to perform research that, for your purposes, accurately simulates reality then you’re faced with a brand new dilemma.
People’s lives are wonderfully messy and even the same person’s behavior will likely change over time. How do you make sense of all these differences, and in many cases contradictions, in a useful way to guide decisions about the product you’re designing?
Ideally we’d be able to gobble up all the information and have it just make sense, but our minds don’t seem to have evolved with this ability to comprehend people much beyond the scale of the individual.
To compensate, we designers have created ingenious tools like personas, mental models, empathy maps, and activity diagrams to more easily digest, communicate, and share the information that we’ve collected.
We need these models, but using them comes with a serious cost.
These models not only help us make sense of complex data, but they also change you in that they will shape the way you see the world. With the simplify and elegance of these models, it can be temping to start to use these models as having more significance that reality itself, and to demote events or people that don’t fit our model as edge cases, or outliers.
In the same way that I believe that it helps to frame research as occurring within a simulation of reality, I believe that it helps to consciously frame these models not as reality itself, but instead as distortions of reality. A useful, but contorted view of the world.
For me, I find this it also encourages the question of how do we create models or tools which instead of attempting to explain reality (a fruitless task), prompt us to ponder our reality.
From the ever increasing capabilities of the devices we’re creating, to rapidly lowering costs, dramatically increasing adoption curves, to sea change of how we perceive digital technologies from a niche hobby to the current status quickly approaching a basic human right. In my own lifetime I’ve seen amazing change which I have no reason to doubt will continue and will take us to unimaginable places.
So, as designers & engineers working with digital technologies, I think it’s safe to say that we’re in the single industry that has the most steady and consistent change, and by association having the greatest impact on every other industry.
We seem to have a natural resistance to change, so perhaps it’s a coping mechanism of the rapid change we’re faced with, but I’m troubled by the sanctification of some of the models that we use which have served us well in the past, but may today by limiting us under the guise of a best practice.
Of the tools that trouble me, the tool that troubles me the most is the persona.
A lot has been written why designers should use this tool (and how to use them), and why you shouldn’t. Whether personas are a good design tool or not isn’t what troubles me, instead I’m worried that we don’t actively look for even better tools.
At the essence of personas I see these tools serving three main purposes:
As a general purpose tool I think that’s an impressive range of qualities from a single tool!
On the other hand, personas also risk trivializing complex situations, and perhaps even distancing us from our users by suggesting that from the creation of this artifact that we no longer need to keep spending time with out users.
Cooper created the persona as a response to some systemic issues within our industry. I encourage everyone to embrace their own inner Cooper and understand the challenges that your team are facing, and experiment with creating your own tools and models that best serve your team.