Researching Transition Spaces

Part 1

In this project, we will study two sets of complimentary research methodologies. Part 1 will focus on research methods that are rooted in the scientific tradition, especially those that deconstruct phenomena into their objective observable characteristics… Part 2 will focus on research methods that are rooted in the ethnographic tradition, especially those that acknowledge the interrelation of the observer and the observed.

–Project Briefing, Nassim JafariNaimi

This project was an assignment for the fall 2018 Discovery & Invention course in Georgia Tech’s Digital Media program. Joining me were classmates Anuraj Bhatnagar and Christina Bui.

Selecting a Transition Space

There were six researchers at the start of this project, but only the initial meeting brought all players to the table. The idea was to choose a common object, then to split into subgroups of three, allowing each to pursue research questions independently. Among dozens of brainstormed possibilities, the group narrowed down our options for transition spaces to just four.

  1. Virtual Reality
  2. Restaurants
  3. Dog Parks
  4. Crosswalks

We set about critiquing each of these finalists. The transition space from real to virtual is largely internal, making it difficult to quantify. Restaurants resist observation, since they create private spaces at each table and booth. The dog park transition from leashed to unleashed makes for an interesting paradigm. However, animal behavior is unpredictable, and it is often difficult to correlate canine response to specific stimuli. By process of elimination then, we came to crosswalks. The transitions space there is literal, highly public, and easily observed. More importantly, it is an information-dense artifact, with an ecosystem of interactions taking place on a literal timer.

The pedestrian scramble crossing at 5th & Spring near the GAtech campus would be the focus of our study.

With our artifact chosen, the sub-groups split and we got to work.

The Scientific Method(s)

Deconstruction and construction are key features Part 1 of this project. The group’s aim was to break down our observations into discrete data and then rebuild that information into a coherent theory. The team selected a trio of research methods with this thinking in mind, with an ultimate aim of returning hard data.

Rejecting fly-on-the-wall observation as redundant (it is easy to remain unobtrusive in a bustling city street) and task analysis as impractical (good luck interviewing truck drivers at a stop light), we implemented the following methods in the following order.

  1. Participant Observation — This method allowed the team an entry point into the data-rich design space of the crosswalk. In short, we got out and walked the space repeatedly, experiencing the activity and context of use firsthand. This ethnographic strategy may seem like an odd choice for research meant to focus on the objective and the observable. However, it was a necessary first step in the process, serving as our primary data-gathering strategy and focusing our attention on specific aspects of the crosswalk. The method included personal experiences, video and photography, field notes, and dialogue with our fellow crosswalk users. This activity provided the raw material for analysis.

    Researcher Anuraj Bhatnagar takes careful note of the elements of the crosswalk. These observations would later be organized through the AEIOU method.

  2.  AEIOU — An organizational method, this acronym stands for Activities, Environments, Interactions, Objects, and Users. With worksheets in hand and notes and images from our participant observation in front of us, the team began to break down our many observation into rigorously organized categories, drawing larger themes of behavior from patterns in the data.

    Organizing our observations in Activities, Environments, Interactions, Objects, and Users helped to separate the quantifiable from the subjective. 

  3. Behavioral Mapping — Crosswalks are a location-based artifact, with user pathways crossing and intersecting in every iteration of the built-in 15-second timer. This method provided a tool for analysis, offering data visualization and a ready-made framework for converting chaotic motion into logical, repeatable behaviors.

(Un)Quantifiable Analysis

People are not logical. While the team was able to identify cause and effect in some cases (e.g. shade-seeking behavior on hot days), other information was less accessible.

Our shotgun approach to observation yielded several observable behaviors. The data suggested interesting themes like pregreening, jaywalking, and the contrasting activity of groups versus solo pedestrians. However, the team landed on the behavior of “personal vehicles” as a question worthy of deeper analysis.

Bikes, scooters, and skateboards seemed to act like pedestrians in one moment and motor vehicles the next. We theorized that the decision to ride or walk a vehicle across the space might correlate with the pedestrian countdown timer. From the first behavioral map, however, we saw that this was idea was inaccurate. Starting at roughly the same time, one cyclist chose to ride his bike across the crosswalk, while another walked her vehicle across. Clearly, there was more at work in this interaction than a simple cause-and-effect interaction between timer and rider. There were motivations at work that numbers could not show, no matter how many times we counted and mapped them.

Despite their identical start times, the red cyclist and the blue cyclist chose different methods of locomotion during a pedestrian crossing, riding and walking their bikes respectively. The purple scooter rider also chose to ride.


Part 2

Methodological Errors

In Part 2 of this project, the team began with a mistake. Seeking to explore our existing question of personal vehicular behavior, we chose methods that would focus response, zeroing in on experiences of that personal vehicle aspect of crosswalks. What we had not taken into account is the nature of ethnography. The paradigm is characterized by a multiplicity of perspectives, and the limitations of knowledge (including researcher knowledge) is an assumed part of the methodology. By framing our research to fit a single point of inquiry, we nearly lost the rich data of lived experience.

Ethnographic Methods

Seeking to repeat the structure of our first research, the group settled on a linear progression of ethnographic methodologies. In the same way that we began with participant observation before, we thought to focus on ourselves first for the ethnography, building outward from those structuring experiences to broader perspectives.  Rejecting questionnaires and surveys as overly-quantitative and word clouds as imprecise, we decided to structure our inquiry in the following order.

  1. Role-Playing — We created five scenarios related to personal vehicles in the space of the crosswalk. With the help of a pair of additional volunteers, we then took turns playacting our responses. Scenarios were built around confrontation and argument in order to foster rich interactions. For example, in one scenario Participant A would portray a mother whose child was nearly hit by a bus, while Participant B would portray the bus driver. We followed up each scenario with round table discussions, jotting down our thoughts and then breaking them down into discernible themes. These included ignorance of the law, rule-following, culpability, road rage, and attitudes towards different demographics of pedestrian.

    Researchers Colin Stricklin and Anuraj Bhatnagar dodge imaginary traffic in this role-playing scenario.

  2. Directed Storytelling — From the themes we identified in the role-play, we next set about crafting a series of interview questions tied to those themes. Because we could not stop traffic to perform interviews, we elected to seek interviewees beside a bike rack near our intersection. Our thinking was that, since personal vehicles like bicycles were at the center of our inquiry, this would be the ideal place to isolate those individuals most familiar with the issues at hand.
  3. Graffiti Walls — There were two distinct version of the team’s graffiti board. In the first, we sought to continue the process of winnowing down our area of inquiry. This board featured images of wrecked cars and bikes in intersections. It featured the prompt, “That one time I nearly got hit in a crosswalk…” and was seeded with anecdotes from the teams’ own close call experiences. The intent was to gain more insight into the existing themes of culpability and rule-following. On the advice of Prof. JafariNaimi, however, we changed to a more open-ended board (see below). This one featured a greater diversity of images, as well as a more generic prompt.

    The diverse imagery and open-ended prompt yielded a greater variety of responses than the focused version. Researchers brought the board to different classes across campus, asking for responses from classmates and instructors.

Ethnographic Analysis

The task of coding feedback did not end with the role-playing method. Rather than enriching existing themes, the second version of the graffiti board served to identify new themes. These included banal observations and non sequiturs, thoughts on redesigning crosswalks, and crosswalks as social space. Crucially, these themes were not ones that the design team could have invented on its own. By foregrounding our own experiences in role-playing, and then using those preliminary conclusions to devise our guided interview questions and graffiti board prompt, we risked blinding ourselves to other insights.

A number of observations concerning local rainbow crosswalks opened up the possibility of using the space for expressive purposes.

In particular, observations of Atlanta’s rainbow crosswalk at the intersection of 10th Street and Piedmont Avenue were a recurring theme. If we were to pursue this design inquiry further, we might look at a way to redesign the crosswalk as a social space rather than a simple safety feature. Much like the findings of this research, such a space could be based on a multiplicity perspectives. What might an expressive and dynamic crosswalk look like? How could it reflect the identity of its particular neighborhood? If designed properly, such an artifact could create an unexpected and extraordinary space within its community.

Part 3

Project 3 will focus on using the methods discussed in class toward framing a problem and devising a design intervention in response. You are not limited in scope or aim of the design intervention. 

–Project Briefing, Nassim JafariNaimi

For this portion of the project, we mixed the crosswalk subgroups as a means of broadening perspective. Joining me in Part 3 were classmates Sammi Hudock and Jordan Chen.

Framing the Problem

Combining our respective groups’ previous insights, the team hit upon four key considerations for crosswalk design.

  • Aesthetic: Crosswalks can serve a decorative function, e.g. our local rainbow crosswalk at 10th and Piedmont.
  • Practical: Crosswalks currently serve a utilitarian purpose of shepherding pedestrians from one side of a street to another while ensuring their safety throughout the process.
  • Communal: Crosswalks are public spaces where groups of pedestrians gather and perform the transition together until their eventual dispersal.
  • Political: The effective use of a crosswalk relies on the cooperation of multiple parties on an established set of traffic laws. At the same time, the establishment of the crosswalk itself relies on meeting expectations of a crosswalk as a safe functional space as determined by multiple political offices and organizations.

Observing that crosswalks are most usually thought of in terms of their practical and political aspects, the team decided to stage our redesign from an alternative perspective. Our guiding principle could be stated in the form of a question: How might we create a crosswalk that serves as a creative space for its community without sacrificing core functionality?

The team drew inspiration from existing creative crosswalk projects from around the world.

Researching Creative Crosswalks

Since we knew that we could not safely and legally implement changes to an existing crosswalk given the project’s timeframe, we re-conceptualized our space as one that guides pedestrian traffic through a space. Drawing inspiration from theme park design as a place that shepherds pedestrians using creative centerpieces and infrastructure design, we looked at this image of Discovery Island at Disney World. The design incorporates pathways which guide pedestrians around a central tree if they want to get from the top right to bottom pathways. We could not implement exactly the same physical restrictions on movement, but we  could mimic them in our design.

Screenshot 2018-12-11 at 12.20.30 PM

Discovery Island at Disney World offered us insight into physical barriers for traffic control, and also suggested our eventual centerpiece. 

Space and Design 

We decided to create a makeshift roundabout centered on an interactive installation outside the north side of Clough Commons. We chose this space because it is an area that experiences a high density of foot traffic as well as vehicular traffic from scooters and bikes. We wanted to see if introducing a roundabout would provide a sense of order to the normal chaos of traffic there, especially during periods of high traffic like class change. We also wanted to observe whether the interactive piece would impact the ways people go about their transition through the space, and whether they chose to interact with it or not.

In designing the roundabout, we chose to delineate space for both pedestrian and cyclist traffic with the use spray chalk lines. The orange triangles indicate traffic cones intended to provide dimensionality along the z-axis and to visually indicate points of entry and exit.

In purely practical terms, the creative centerpiece could have been anything. It merely had to serve as a physical barrier, and to invite people in to interact. Recall that we were coming at this design from the aesthetic and communal angles though. Some of our initial ideas included player pianos or a 3D graffiti board structure, but we wanted a design specific to the place and time of the intervention. The Thanksgiving tree conceit won out due to its novelty, seasonal appropriateness, and potential for interactivity. It also fit the practical concern of visibility, being conspicuous enough to attract attention.

Screenshot 2018-12-11 at 12.20.20 PM

With a conventional roundabout design placed in an unconventional space, the team sought to impose a familiar order in a chaotic intersection.

Methods and Execution

Our materials list included the following:

  • Spray chalk, white and green
  • Foam core stencil 
  • 1 tree branch
  • 9 traffic the cones
  • Christmas tree stand
  • Paper leaves w/ hole punch, ribbon, and markers
  • Small table

We implemented the design the Tuesday before Thanksgiving, increasing the relevance of the centerpiece. Installing the piece before the first classes of the day, we gathered data for four hours, from 8:00 am to noon. Team members alternated carnival barker duties for the duration, calling in passersby to “hang a leaf.”

We spoke with users throughout the experiment. Recording their reactions as well as our own first-hand experiences and impressions was our primary mode of data gathering.

The team also watched for traffic behavior, both pedestrian and cyclist, in relation to the constructed roundabout. We estimate that 30% of cyclists and 15% of pedestrians took note of the installation and altered their behavior. We did not note any significant traffic issues caused by the installation. 


Users would write what they were thankful for on the leaves, then hang them on the “Thanksgiving Tree” at the center of our roundabout.


Where we had previously observed a space of pure utilitarian use—pedestrians with heads bowed and phones out, busily making their way to class—we now saw a space with a mixed use. The tree received nothing close to 100% use, but the amount of leaves gathered in four hours suggest a space that is at least more communal that it had been.

Significantly, the leaves weren’t the only feedback the team received. During our conversation with pedestrians, our recorded notes revealed a few major categories: traffic safety, the tree itself, and the pressing need to get to class.

Pedestrians and cyclists alike used and ignored the crosswalk by turns. While the precise behavior of a roundabout never emerged, users did slow down. As it turned out, we were not the only people on campus interested in that kind of behavior.

Screenshot 2018-12-11 at 12.20.02 PM

A selection of user feedback shows typical comments from users.

Bicycle Infrastructure Improvement Committee

When we started this inquiry, one of the first things we did was to contact the City of Atlanta Planning Department, hoping to gain some perspective on the ways crosswalk interventions are actually staged in the community. Once we’d decided to implement our project on campus, however, that early research became practically irrelevant. As it turned out, however, there was an on-campus group with an interest in this work.

Members of the Georgia Tech Bicycle Infrastructure Improvement Committee introduced themselves during the intervention, and invited the team to attend one of their regular meetings. In contrast to our aesthetic-focused work, this was a group with a primarily practical mindset. 

We disseminated our findings at the meeting, and received feedback from the various stakeholders in the community. One particularly interesting point of comparison came from a landscape architects, who suggested replacing several of the site’s grid intersections with xeriscaping or rough cobble stones as a means of slowing down bicycle traffic. “Sure,” he said. “You get rid of the fun interactive elements, but this will solve your problem.”


This sketch from a Gerogia Tech landscape architect show grid intersections filled with xeriscaping or rough cobblestones, intended to slow cyclist traffic.

It is a valid approach, but it does not have to be a totalizing one. Could the rough cobbles be made to look like a chess or go board? Could these areas become dedicated sidewalk art spaces? Once the practical problem is solved, why not return to the alternative mode of the aesthetic?

The meeting served to highlight a key insight from the intervention. When designing a solution to a problem, aspects of the practical, the aesthetic, the communal, and the political all have their place. What we found is that it’s important to know when to use the right lens, but equally important not to let one lens dominate the others.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.