Share on facebook
Save on delicious
RSS Feed
In-Depth Case Studies
In-Depth Case Studies
The design and use of advanced Internet and Grid tools and infrastructure in the social, physical and computer sciences are likely to reconfigure not only how researchers produce, use and collaborate around key resources, such as data and software tools, but also how they share such resources and who can gain access to them. This reconfiguration of access raises numerous ethical concerns, legal uncertainties, and institutional conflicts. To explore these interrelated phenomena, we have redesigned our in-depth case studies to move away from a ˇ®project' focus to aim instead at the study of data sets and tool artefacts with a focus on the practices and issues that emerge as the data and tools are developed and travel between individuals, groups, communities and different stakeholders. This section explains our shift in design.
The original OeSS project plans called for 5 in-depth case studies of e-(social) science projects, including MiMeG , Digital Records , eDiaMoND , Advanced Grid Interfaces, and the Reality Grid . During the first nine months of the project, the study team completed preliminary field research and interviews on each of these projects, along with most other e-social science and e-science projects in the UK and many outside the UK. However, our first year of empirical and conceptual work led us to question the value of ˇ®projects' as the optimal object of analysis.
There were two principal concerns. First, the initial selection of project-based case studies was driven by the aim of identifying the key non-technological issues to emerge from e-science and e-social science. However, our understanding of the key issues progressed markedly since the original project plan. Our work had been informed by several social shaping studies of NCeSS, such as Chimera , which began disseminating its results since our initial proposal. Their findings provide a fairly robust outline of the key issues. Moreover, our own preliminary research covered a wide range of projects. Also we have held two issue-based workshops, a node-to-node meeting and participated in several of the NCeSS agenda setting workshops , providing further insights about existing projects. These early research efforts enabled us to refine a set of potential issues (as outlined in earlier reports). Based on our developing understanding of this field, we decided that a set of project case studies not only risked replication of research findings, but also failed to focus on the most central elements of the e-sciences ¨C the nature of the data and tools enabled by advanced e-research, both of which are likely to outlive particular projects, and raise more critical issues for the social and natural sciences.
An interim plan called for a focus on initiatives to address ethical, legal and institutional issues. For example, we might conduct critical case studies of efforts to protect the anonymity of individuals in the analysis of quantitative data. This gained support across the project, but as we examined this option it soon became clear that the scale of such fieldwork would not be feasible within the scope of the current project. For instance, there are more than five major national efforts to address concerns over anonymity of quantitative data in North America and Europe. We therefore decided that while there was great merit in this approach, it should be reserved for future, follow-on research, where the necessary resources could be marshalled. We also believe that some of these issues surrounding new initiatives will, however, be examined at a broader level than in-depth case-studies through our series of issue-based interviews, which will bring in the specialist expertise of the wider OeSS project team.
Through this iterative process of conceptualising our case-study approach we arrived at a new focus on e-science resources, focusing primarily on data sets and tools. Our preliminary research found a distinct pattern of practices and issues emerging across the different types of data (such as data on people versus objects), and tools (such as modelling and simulation versus distributed annotations of qualitative data) being produced within e-science and e-social science communities. Some projects are developing new data sets, some of which have a prior history and some of which have moved (or will move) on to other research projects. For example, the e-Diamond project assembled a data set of medical images. Some projects are creating new tools, often modifying or otherwise improving on existing software, for example, and in some cases, tool development is the major creative outcome of the project. For instance, e-Diamond research did not simply develop a new data set, but also build the software and systems for analyzing these medical images. In the case of e-Diamond, the data set was the source of particular social issues, such as privacy of medical records and the ownership of mammographic images, while the tools for compiling, comparing and analyzing visual images were a source of distinctly different issues.
Not only do different issues arise around particular data and tools, but they are also linked to the research context. For example, we have found many differences in practices to be clustered by disciplines, around qualitative and quantitative research styles, and the extent to which they are oriented towards basic or applied research and their concern with physical or human objects of research. For example, the anonymity of qualitative video-based data brings to bear a different configuration of issues different from the anonymity of quantitative survey data. Based on such considerations, we are developing a framework for selecting case-studies. A preliminary framework is summarised in the following table:
The framework will be designed to capture the main types of data and tools being produced and shared within and across e-science and e-social science. This will enable us to choose case studies that ensure a range of e-social science issues are examined. Projects become a point in the history of particular data or tools, and therefore remain significant, but not the central object of the analysis. The specific projects indicated in the table are a lens through which to observe these phenomena and form the starting point for fieldwork. The relevant PIs have been contacted and have agreed to participate in the fieldwork. In several of the cases fieldwork is already underway. However, field research, focused on the case studies of data and tools across the e-sciences, will be the major activity of OeSS over the coming year.
Within this new framework, however, the specific scientific makeup and organisational arrangements of the project will be seen as critical aspects of the context shaping the salience, importance, and tractability of the issues that arise in the utilisation and management of data and tools. Neither the intrinsic technical characteristics of the latter nor the generic attributes of ˇ°humansˇ± suffice to define the variety of problems that project managers and participants will confront in this domain. Thus, the security requirements (and access control implications) suitable for epidemiological research on federated mammography images are likely to be very different from those that are deemed appropriate in a project responsible for creating searchable databases containing the results of chemical analysis or dangerous compounds.