Arguments for and field experiments in democratizing digital data collection : the case of Flocktracker

Abstract

Data is becoming increasingly relevant to urban planning, serving as a key input for many conceptions of a "smart city." However, most urban data generation results from top-down processes, driven by government agencies or large companies. This provides limited opportunities for citizens to participate in the ideation and creation of the data used to ultimately gain insights into, and make decisions about, their communities. Digital community data collection can give more inputs to city planners and decision makers while also empowering communities. This thesis derives arguments from the literature about why it would be helpful to have more participation from citizens in data generation and examines digital community mapping as a potential niche for the democratization of digital data collection.

In this thesis, I examine one specific digital data collection technology, Flocktracker, a smartphone-based tool developed to allow users with no technical background to setup and generate their own data collection projects. I define a model of how digital community data collection could be "democratized" with the use of Flocktracker. The model envisions a process in which "seed" projects lead to a spreading of Flocktracker's use across the sociotechnical landscape, eventually producing self-sustaining networks of data collectors in a community. To test the model, the experimental part of this research examines four different experiments using Flocktracker: one in Tlalnepantla, Mexico and three in Surakarta, Indonesia. These experiments are treated as "seed" projects in the democratization model and were setup in partnership with local NGOs.

The experiments were designed to help understand whether citizen participation in digital community mapping events might affect their perceptions about open data and the role of participation in community data collection and whether this participation entices them to create other community datasets on their own, thus starting the democratization process. The results from the experiments reveal the difficulties in motivating community volunteers to participate in technology-based field data collection. While Flocktracker proved easy enough for the partner organizations to create data collection projects, the technology alone does not guarantee participation. The envisioned "democratization" model could not be validated. Each of the experiments had relatively low levels of participation in the community events that were organized.

This low participation, in turn, led to inconclusive findings regarding the effects of community mapping on participants' perceptions and on the organizations themselves. Nonetheless, numerous insights emerge, providing lessons for the technology and how it might be better used in the future to improve digital community mapping events.