A design theory for observational crowdsourcing

点击翻译按钮获取中文摘要

Crowdsourcing, in which activities normally undertaken inside an organization are expanded to engage ordinary people in contributing to activities of an organization, has become increasingly popular with the development of online software platforms that support these activities. Crowdsourcing involving data collection (data crowdsourcing) is one of the most important forms, as it makes it possible to collect data on a scale that is otherwise impractical for many organizations. Most data crowdsourcing efforts to date have focused on simple and well-defined data collection tasks (microtasks). However, a new and more powerful form of data crowdsourcing has emerged, requiring data collection for complex, ill-defined and evolving uses (observational crowdsourcing). However, despite major differences in the nature of the task and the contributors, the primary approach for designing software platforms for observational crowdsourcing currently mirrors that of microtask crowdsourcing. This limits the extent to which observational crowdsourcing can be used effectively. We propose to develop and test a design theory for observational crowdsourcing. This design theory will be expressed in the form of design principles that link features of a software platform with the distinct characteristics of observational crowdsourcing projects. These principles will be derived and justified based on core underlying theories that connect design features to desired outcomes. This theory will then be used to design a new crowdsourcing platform that will host observational crowdsourcing projects. In this process, specific design choices will be made following design guidelines extracted from the theory. The platform and design theory will be evaluated by assessing the quality of data collected. Key expected outcomes are that the platform will enable the collection of accurate, precise, and diverse data, which will be useful to the consumers of observational crowdsourcing data. The significance and impact of this work are two-fold. First, the platform will support observational crowdsourcing projects in Canada and beyond. Such projects stand to benefit project sponsors and other data consumers, while engaging data contributors in contributing to real societal challenges. Second, the proposed design principles will be available to be adapted and used by other developers of observational crowdsourcing and related technologies.