Analyzing Collaborative Activity – Representing field research for understanding collaboration, CSCW 2002 Workshop Position Paper
Supporting Informed Debate of Urban Planning Alternatives
A.J. Brush, University of Washington, Seattle, USA
Alan Borning, University of Washington, Seattle, USA
Introduction
Planning for land use and transportation infrastructure in a major metropolitan area, for example modifying zoning or expanding roads, is a complicated task. Debates regarding alternatives are often politically charged, with the stakeholders, from professional urban planners to citizens, having very diverse opinions. The goal of the UrbanSim project [Urb] is to allow people to simulate the future effects of different policies on urban growth and development, transportation usage, resource consumption, and the environment. We hope comparison between alternatives will foster informed civic debate and deliberation.
UrbanSim is a state-of-the art land use modeling system that includes a transportation model. Users can interact with UrbanSim to create scenarios that model policy alternatives and then run the simulation into the future to explore outcomes. UrbanSim also supports analyzing and comparing the results of different scenarios. UrbanSim has been applied in Eugene, Salt Lake City, and Honolulu. The current interface is aimed at urban planning professionals; its output is primarily tabular results, which can then be explored using GIS and other software.
While the software development efforts to this point have focused on developing a state-of-the art simulation system available to experts, two longer-term goals have been supporting a more diverse user group, and evolving the system to more directly support the inherently collaborative activity of planning and analyzing urban development. We are using the Value Sensitive Design methodology [FKB02] to ensure human values are explicitly accounted for in our design process.
Currently, we are in the process of gathering data through interviews and observation to better understand the urban planning process and role of simulation in that process. This data will be used in two ways: (1) as empirical data to help determine the important stakeholder values to support when presenting simulation results, and (2) to design interfaces to UrbanSim for comparing and discussing alternative scenarios by groups, particularly for mediation purposes.
As we are gathering and analyzing data this fall, the focus of this workshop is very relevant for our project. As we discuss in more detail below, we plan to initially use approaches from Contextual Design [BH98] and Personas [Co99] to represent and analyze the data we collect. We expect multiple representations of our data will likely be necessary to capture the complexity of the urban planning process. At the workshop, I will share our initial representation of the data, and hope to discuss alternative approaches and theories that might help us to better understand and represent the data.
Research Plan:
In this section, we outline in more detail the techniques we will use to collect data and our planned approach for analyzing and representing the data. While we are generally interested in better understanding the urban planning process, our focus on stakeholder values and determining ways that UrbanSim simulation results could better support group decision making will guide our data collection and analysis efforts.
Data Collection
We will collect data in two main ways, through interviews and observations of meetings. The interviews will be with a variety of representatives from different stakeholders in the urban planning process, including staff from metropolitan planning agencies, elected officials, and members of advocacy groups. They will be semi-structured in the style of contextual inquiry. Our focus in the interviews will be understanding an interviewee’s role in land use and transportation planning, the values that are important to them and their stakeholder group, what results or indicators (e.g. the number of days air pollution exceeds EPA guidelines, the cost of new space for startup businesses) they would like from the simulation, and observing their current use of simulation results if applicable.
In observing meetings of groups involved in planning, decision-making, and discussion of land use and transportation alternatives, we will take a more ethnographic approach. By observing meetings, we hope to gather additional data about the values important to the group to verify and to expand information gathered in interviews. Meeting observations will also be important for gaining a sense of the culture of the group. In addition, we will observe meetings between representatives from a variety of stakeholder groups, with a particular focus on how different groups interact, both in collaboration and as adversaries. As necessary, we may also use surveys and other questionnaires with members of the stakeholder groups to collect additional information.
Data Analysis
When analyzing and building representations of the data, we plan to meld two existing approaches: contextual design and personas, taking into account the goals of Value Sensitive Design. From contextual design we expect three work models: the flow model, culture model, and artifact model, to be particularly important. The flow model will help us characterize the collaborations and communication occurring, the culture model will help us understand important values, and the artifact model will be important input to redesigning the UrbanSim interface.
One major challenge is that we will probably need several different sets of work models. One set to represent interactions between all the stakeholders in the urban planning process (e.g. communication between metropolitan planning organizations, advocacy groups, and citizens) and another set of work models for the internal workings of each stakeholder group (e.g. how a particular advocacy group operates).
Analyzing interview data to determine values important to each stakeholder group is an important part of our commitment to the Value Sensitive Design methodology. Value Sensitive Design employs a tripartite methodology with conceptual, empirical and technical investigations. In our conceptual investigations, we have already identified three values (fairness, accountability, and democracy) that will be explicitly supported in the interface. The stakeholder values we identify (e.g. pedestrian-friendly neighborhoods or minimal commute time) will then all be supported equally in the interface as far as possible, allowing each stakeholder to select the values important to them when simulating policy alternatives. At this point, our plan is to represent stakeholder values in the culture model for that group.
We view personas as an important step in our data representation task. Developing a persona (or multiple personas) for different stakeholders will help us characterize the many different participants in the urban planning process. By defining and refining the personas, we will explore whether multiple interfaces to UrbanSim will be necessary and if so how many (e.g. one for professional planners, one for concerned citizens). Personas will also be valuable for focusing our development team and prioritizing new features.
Conclusion
Land use and transportation planning is a complex and politically charged process. We are excited by the opportunity to support informed comparison of different policy alternatives using the UrbanSim modeling system. Our initial efforts to understand the values of different stakeholders groups and characterize the interactions within and between groups are critical in developing interfaces to the UrbanSim system that will foster debate and deliberation. The Analyzing Collaborative Activity Workshop at CSCW 2002 will be a valuable forum for sharing our data representation efforts and receiving feedback and ideas about other representation and analysis techniques that would be help us understand the complicated activity of urban planning.
References
[BH98] Hugh Beyer and Karen Holtzblatt, Contextual Design, Defining Customer Centered Systems. San Francisco: Morgan Kaufmann, 1998.
[Co99] Alan Cooper, The Inmates are running the Asylum: Why High-Tech Products Drive Us Crazy and How to Restore the Sanity. Indianapolis: SAMS, 1999
[FKB02] Batya Friedman, Peter H. Kahn, and Alan Borning, Value Sensitive Design: Theory and Methods, Submitted for publication, Available at:
[Urb] UrbanSim web page,