![Sample our Politics & International Relations journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/5947/TOC-Subject-Sample-2021-Politics-International-Relations.jpg"})
Abstract
This article reports an evaluation of the quality and scale of interactions during an online field experiment. A large number of individuals (n= 179) worked with an online public participation geographic information system (PPGIS) platform during a month-long field experiment about regional transportation improvement decision making in the central Puget Sound area of Washington. The system platform logged more than 120,000 client–server interaction events. We developed a geovisual analytic technique called a grapevine to evaluate the quality and scale of public participation using event log data. The grapevine 4D space–time geographic information system (GIS) visualization helps distinguish productive clusters of analytic-deliberative process and for guiding content analysis of the user-generated discussion. Comparison of the nature and content of participant message exchanges before and after GIS-based analytic activities revealed a significant shift in focus. We characterize this shift in the focus of deliberation as the result of participants sharing their lay expertise, in the form of simplifying assumptions, to cope with the technical details of the GIS-based analysis and move the large group toward agreeing on a transportation package for the region. The article concludes by extending the implications of the research with a three-part framework called participatory interaction modeling, wherein geographically distributed networks of designers and developers, participant users, and social and behavioral science evaluators learn how to create PPGIS capabilities that can better address societal goals.
En este artículo se informa sobre una evaluación de la calidad y escala de las interacciones registradas durante un experimento de campo a través de la web. Un gran número de individuos (n = 179) trabajaron en red con la plataforma de un sistema de información geográfica de participación pública (SIGPP) durante un experimento de campo de un mes de duración, acerca de la toma de decisiones para mejora del transporte regional en el área central del Puget Sound del estado de Washington. La plataforma del sistema cargó más de 120.000 eventos de interacción cliente-servidor. Utilizando datos de eventos de entradas, desarrollamos una técnica analítica geovisual denominada grapewine(“viña”) para evaluar la calidad y escala de la participación pública. La visualización espacio-tiempo 4D de grapevine del sistema de información geográfica (SIG) ayuda a distinguir aglomeraciones productivas de proceso analítico-deliberativo y a guiar el análisis de contenido de la discusión generada por el usuario. La comparación de la naturaleza y contenido de los intercambios de mensajes del participante antes y después de las actividades analíticas basadas en SIG revela un cambio significativo de foco. Caracterizamos este cambio en el foco de la deliberación como el resultado de la decisión de los participantes de compartir su experticia de legos en forma de suposiciones simplificadoras, para competir con los detalles técnicos del análisis basado en SIG y mover al grupo mayor a concordar sobre el paquete de transporte para la región. El artículo concluye extendiendo las implicaciones de la investigación con un marco tripartita denominado modelado de interacción participativa, en el cual las redes de diseñadores y desarrolladores distribuidos geográficamente, usuarios participantes y evaluadores de ciencia social y comportamental aprenden cómo crear capacidades SIGPP que puedan aproximarse mejor a las metas de la sociedad.
Acknowledgments
This research was supported by National Science Foundation Information Technology Research Program Grant No. EIA 0325916. Support of the National Science Foundation is gratefully acknowledged. We would also like to acknowledge the PGIST research team and especially Piotr Jankowski, Michael Lowry, Arika Lingmann-Zielinska, Michael Patrick, Kevin Ramsey, Martin Swobodinski, Zhong Wang, Matthew Wilson, Jie Wu, and Guirong Zhou for their contributions to a completed system and the approaches we discussed relative to evaluation. We thank the reviewers for their thorough critique of the manuscript drafts. The authors are solely responsible for the content.
Notes
1. There are four general settings for any field experiment, derived by cross-tabulating how participants interact in time (synchronous or asynchronous) and space (presence or telepresence-online) as outlined by CitationJankowski and Nyerges (2001). These four general categories of experimental setting have appeared in many different forms in the literature; see Miller (2007) for other approaches.
2. For comparison, an ongoing survey of more than 2,000 individuals in the United States by the 2008 Digital Future Project reported that the average Internet user to date spends about 15.3 hours every week online (Center for the Digital Future 2009).
3. The MH test uses the chi-square distribution and is useful for distinguishing changes in questionnaire response due to experimental treatments in pre- and post-test sample designs when assumptions of normality cannot be met. In most cases, a parametric paired-sample t test would be the appropriate test for a pre–post study. Because variables in a questionnaire are self-reported ordinal ranks, differences between reported values are arbitrary, so use of a nonparametric test is more appropriate.
4. Any MH test result for paired questions of less than 0.05 (asymptotically significant, two-tailed < 0.05), which measures the probability of obtaining a chi-square as extreme in repeated samples if the difference was merely random, would mean that participant's responses before versus after using the LIT PPGIS platform were significantly different.
5. Compared with their expectations before using the LIT PPGIS platform, after using the LIT PPGIS platform participants responded more negatively to (1) having an interest in these kinds of discussions on the Internet (asymptotically significant, two-tailed = 0.022, Std. MH statistic = 2.286), (2) believing that discussions they have with other participants in the LIT Challenge will help them understand transportation problems and proposed improvements in the central Puget Sound region (asymptotically significant, two-tailed = 0.000, Std. MH statistic = 3.922), and (3) expecting that their own opinions about transportation issues in the central Puget Sound region might be shaped by participating in the LIT Challenge (asymptotically significant, two-tailed = 0.000, Std. MH statistic = 3.923).
6. For preliminary purposes, we decided that any other event type than those indicating the five listed deliberative activities would be provisionally classed as an analytic activity.