Taking the Measure of Our Experiences
How BEworks quantifies the human experience in public spaces.
When you think about the public spaces you frequent — parks, public transit, streets, or museums — what comes to mind? Do any strong feelings or memories surface, such as the joy of watching your kids play at a sunny neighborhood park, or a feeling of dread when you think about boarding a subway car at rush hour?
Since public spaces are such an important part of people’s everyday lives, they often bring up strong feelings. For those in the kyu Collective who design physical spaces, these strong feelings are important to hear to ensure we design spaces that resolve pain points, meet user needs, and provide a great overall experience. As behavioral scientists, we also know that people’s judgments of spaces can be highly biased, which is something that design researchers should be aware of and account for in their work, if they aren’t doing so already.
If we ask a user to tell us about their opinions of a public space, their judgments will be influenced by the most intense experiences they’ve had there, as well as their most recent experiences. For instance, people may rate the park where they had their first kiss as more beautiful, well-designed, and user-friendly than another, very similar park simply because there is a positive mental “halo” around that memory. Similarly, people may rate their experience of a transit service negatively because they’ve read about safety incidents.
People over-weight these intense experiences in memory for the same reason most people intuit that riding in an airplane is more dangerous than riding in a car (it’s not, statistically); evolution has selected those of us whose memories are highly sensitive to reward and loss. But, this availability heuristic, as it’s called, distorts people’s perception of spaces and could prevent design researchers and physical space designers from getting accurate and representative information about their users’ experiences. The same is true for the recency bias, which describes how people over-weight their most recent experiences when making judgments.
At BEworks, we strive to understand how humans truly think, feel, and behave in the most objective way possible. While we always want to hear about edge cases, we want to consider those in light of the facts that are representative of most experiences. By understanding how most people truly experience a space, we can design and improve spaces that drive the widespread behavior change that our clients want to see — whether that is reducing the number of transit riders who block doorways and stairs, facilitating the orderly exit of concert-goers from a venue, or enhancing productive negotiations at a climate change summit.
So, what are some of the less biased methods we use to understand the experience of physical spaces? A recent project, measuring the pro-social and anti-social behaviors in some of the busiest subway stations in downtown Toronto, allowed us to see this in action.
In the figure above, we have mapped different research methods onto the types of user insights those methods provide. Some of these methods — like observed behavior, perceptions, and computer vision analysis — we’ve used in past casework — while others, like heart rate and skin conductance, portable EEG, and acoustic event analysis — are methods many of BEworks’ scientists have used in a lab but have not yet employed in fieldwork. Our behavioral scientists are actively exploring these methods and seeking opportunities to leverage them in support of our clients’ questions. That said, we always want to ensure that we select the most appropriate research method for our client’s ask.
Here are sample use cases for the different methods:
Sample client ask:
How can we help people move faster through this space?
Research method:
Group-level patterns: Computer vision analysis can measure human movement at scale.
Sample client ask:
How much fun are people having in this space?
Research method:
Group-level patterns: Acoustic event analysis can detect the frequency of certain predictable sounds, like laughter.
Sample client ask:
How well are people getting along in the space?
Research method:
Observable social behaviors: Field researchers can make on-site structured observations to look for the frequency of prosocial behavior (like eye contact) and self-interested behavior (like blocking stairs).
Sample client ask:
Does the space fit with people’s expectations?
Research method:
Cognition and motivation: Portable EEG can detect when people’s expectations don’t match their environment, even if they can’t articulate why.
Sample client ask:
Is it easy for people to get where they want to go in this space?
Research method:
Cognition and motivation: Wearable eye trackers can reveal how much confusion people feel during navigation by measuring visual searching.
Sample client ask:
Do people feel comfortable and welcome in this space?
Research method:
Emotional responses: Heart rate and skin conductance monitors can compare the stress levels of different user groups to reveal unarticulated stressors.
Sample client ask:
Are people behaving how we want them to within the space?
Research method:
Observable individual behaviors: Field researchers can make on-site structured observations to look for the frequency of desired behaviors, such as putting waste in a bin or interacting with installations.
Sample client ask:
What are the best and worst things about the space?
Research method:
Perceptions and opinions: Subjective opinions can reveal bright spots and pain points that help tell the story of the space, but these non-representative findings should be considered in combination with less biased research findings
Reducing Problematic Subway Behaviors in Toronto
Our client, the Toronto Transit Commission (TTC), handles approximately 141 million annual riders and proactively identified a future need to handle challenging passenger volume, especially at three of the busiest subway stations in their network: St. Andrew (in the downtown core), St. George (a university campus and transfer station), and Dundas (the largest shopping mall and a cultural center).
BEworks conducted on-site observations of how riders moved through these stations and observed a few problematic behaviors that created bottlenecks and crowds. Those behaviors included individuals running on the platform, blocking train doors, and stopping at the bottom of the stairs. We observed that some of these behaviors stemmed from issues with the stations themselves, such as dimly lit platforms causing people to bunch up by the stairs. Riders also did not appear to follow social norms around how to walk up and down stairs, or where to wait on the platform, making it difficult for people entering the station to “go with the flow.”
To reduce these undesirable behaviors — and we know we’ve all been guilty! — we produced audio announcements, behavioral posters, floor decals, and new fare gate signage to improve passenger flow within the existing station design. These interventions compensated for the lack of clear social rules about how to move through the station by providing explicit instructions for riders on where they were expected to stand and walk.
To measure the effect of these interventions on problem behaviors, we partnered with Invision AI, an AI software company specializing in 3D modeling of spaces and computer vision. We measured passenger movement patterns over a two-week baseline period, and then implemented our interventions in the three stations. Rider movements were captured on CCTV feeds that were never stored by BEworks or Invision AI; individual riders were converted to dots on a spatial map and analyzed in real time so that any single dot could never be attributed to any individual. We then looked for any change in passenger movement patterns in the two weeks after these alterations.
What we found was interesting. We were able to reduce the number of people standing in front of the subway doors by 17 percent with floor decals instructing people to stand to the side. We also reduced the number of people stopping at the top and bottom of the stairs by up to 21 percent using “Keep Moving” decals. Together, the interventions were able to get up to 27 percent more people off of platforms in any given hour, which will help the TTC manage capacity into the future. After the study, the TTC decided not to take down the interventions, so if you find yourself in Toronto in 2024, you may be able to spot them in stations yourself!
Measuring Human Experience to Design Better Spaces
We recognize that the user experience, no matter how robustly measured, will need to be considered alongside many other design factors such as scale, sustainability, and permanence, as well as the broader social and cultural context in which the project is being built. And as we further explore technology to help us better understand the human experience, issues of privacy, informed consent, and public interest will remain top priorities for us.
We hope that psychological and behavioral research methods become more commonly used so that designers are equipped with the most accurate and credible insights into how users truly experience and behave in physical spaces.
Jennifer Weeks is Vice-President of Strategy and Pierre-Jean Malé is a Senior Strategist at behavioral change firm BEworks.