Designing and evaluating participatory cyber-infrastructure systems for multi-scale citizen science.

摘要

Widespread and continuous spatial and temporal environmental data is essential for effective environmental monitoring, sustainable natural resource management, and ecologically responsible decisions. Our environmental monitoring, data management and reporting enterprise is not matched to current problems, concerns, and decision-making needs. Citizen science programs create opportunities for more continuous and widespread data collection, fill data gaps, and inform decisions. These programs are increasing in number and breadth. They operate at multiple spatial and temporal scales, tackle a wide array of environmental challenges, engage volunteers from all walks of life, and create volumes of scientific data. Information management systems flexible enough to support the varied nature of these data are rare, overly technical, hard to use, difficult to understand, poorly defined, and lack effective training materials.;Flexible systems require creative attention to sustainable technology, stable institutional resources, innovative database designs, effective educational materials, and interoperable services. They require computationally efficient geospatial analysis and imaging techniques capable of handling massive amounts of data collected across vast geographic scales and they must provide effective training materials for people to learn the skills required by the citizen science process. Participatory cyber-infrastructure systems are needed to meet the needs of multi-scale citizen science programs.;This dissertation research investigated, designed, developed, implemented, tested, and evaluated a participatory cyber-infrastructure system built to support multi-scale citizen science. My objectives were to: (1) examine the art and science of multi-scale citizen science support, (2) evaluate the usability of a web mapping application created through cyber-infrastructure for invasive species citizen science programs, (3) compare the effectiveness of static and multimedia online communication approaches for training citizen scientists, and (4) offer guidelines for the development of cyber-infrastructure systems adept enough to support the needs of citizen science programs operating at multiple spatial and temporal scales in many domains. I created a participatory cyber-infrastructure system and developed a framework to situate citizen science programs based on their scope, scale, and activities. I used the cyber-infrastructure system to create a website specific to invasive species citizen science projects (www.citsci.org) and evaluated the usability of the website (n =16) to determine general perceptions, discover potential problems, and iteratively improve features. I compared the effectiveness of online static and multimedia tutorials to teach citizen science volunteers (n=54) how to identify invasive plants; establish monitoring plots; measure percent plant cover; and use Global Positioning System devices. I also continuously received feedback from citizen science organizations using the cyber-infrastructure system.;Results demonstrate that cyber-infrastructure systems can be adept enough to support the needs of citizen science projects operating at multiple spatial and temporal scales across many domains when built with a flexible architecture. Cyber-infrastructure use resulted in 27 online projects contributing 5,196 species occurrences. Features for volunteer management; communication among volunteers and coordinators; data entry; program evaluation, online analysis; and reporting integrated into cyber-infrastructure systems will improve their effectiveness. Careful attention must be given to the usability of complicated map and decision support features. Map-based and early alert tasks required a long time to complete and had low completion rates. Mean task completion rates ranged from 25 to 75% for map tasks and 0% to 33% for early alert features. Overall, the average time to complete tasks ranged from 00:01:42 to 00:02:17 and the mean completion rate ranged from 36 to 90% across all scenarios. Citizens trained online through static and multimedia tutorials provided less ( p0.001) correct species identifications (63% and 67%) than professionals (83%) across all species, but did not differ (p=0.125) between each other. The variability in percent cover estimates between static (+/-10%) and multimedia (+/-13%) participants did not differ (p=0.86 and 0.08 respectively) from those of professionals (+/-9%) and the tutorial approach had minimal influence (p=0.07) on the variability of participant plant cover estimates. Volunteers trained online struggled with plot setup and GPS skills regardless of tutorial approach. Overall, the tutorial approach did not affect the field skills and abilities learned by volunteers. The development and evaluation of a cyber-infrastructure in support of multi-scale citizen science discussed herein situates citizen science programs within a framework of their scope, scale, and activities; de-fragments data; reduces complexity; helps ensure comparability; fills data gaps; refines our understanding of web usability; improves our understanding of online educational approaches; and closes the communication gap between scientists and citizens. It increases the number and variety of people able to contribute information to address pressing environmental problems while participating in local, regional, and global environmental stewardship.