Innovations in Digital Research Methods. Группа авторов
mind, the authors conclude by outlining the ontological challenges (echoing the reservations that Elliot and Purdam set out in Chapter 3) and the technical challenges of mining text in social research settings. They note, in the case of social media, increasingly restrictive access policies, and they also consider the ethical implications of text mining used as a social research tool.
Chapter 9: Digital Records and the Digital Replay System
As many of the contributors to this book recognize, the capacity to capture behaviour through the ‘digital footprint’ that people generate as a by-product of their everyday activities has the potential to transform the practice of empirical social science. In this chapter, Crabtree, Tennent, Brundell and Knight examine how new tools for data collection and analysis make it possible to exploit this data. Their discussion focuses in particular on the development of ‘digital records’ that enable social science researchers to combine novel and heterogeneous forms of digital data, such as video, text message logs and GPS data, with more traditional and established forms, such as audio recordings and transcriptions of talk.
The authors describe the Digital Replay System (DRS), an open source, extensible suite of interoperable tools for assembling, synchronizing, visualizing, curating and analysing digital records.3 In Chapter 5, Lambert presents solutions to the data management problems attendant in the use of conventional kinds of social data such as surveys. From this perspective, DRS can be viewed as a prototype for meeting the data management and linking challenges presented by novel sources of social data. Crabtree and his co-authors provide a step-by-step exposition of several different examples; these include capturing rich accounts of people’s physiological reactions while on a fairground ride, a corpus linguistics perspective on visitors’ interactions in an art gallery, and disaster mapping and management. Collectively, these examples illustrate how the use of a system like DRS can enable the assembly of digital records capturing a wide range of interactions between people that are a by-product of their use of various digital devices, and make them available for subsequent visualization, curation and analysis. Finally, the authors consider future developments, particularly the prospects for making use of mass participation in social science research through the use of mobile devices for the crowd-sourcing of data.
Chapter 10: Social Network Analysis
The distinctive contribution of social network analysis (SNA) to social research is its stress on the importance of studying the structure of relationships between people rather than considering them as unconnected individuals. Like many of the other advances in research methods covered in this book, SNA is a mature methodological tool. Arguably, it owes its rise to greater prominence in recent years to two factors. One is that, as with many other established social research methodologies, e-Infrastructure has extended the scale and complexity of what is achievable, in this case by providing SNA with new and more powerful means to capture social network datasets, analyse them and visualize the results. The second factor is that many of the new types and sources of digital social data – such as hyperlink networks (the structures of links between websites) and social networking sites such as Facebook and Twitter – are inherently relational.
In this chapter, Ackland and Zhu review the history and methodological principles of SNA, and survey several of the research tools now available for SNA data collection, analysis and visualization. They draw on examples of studies of Facebook, Twitter, Flickr, online newsgroups and websites to illustrate contemporary and arguably the most prominent uses of SNA – to study people’s behaviour in social networking sites. Ackland and Zhu go on to discuss two key ontological questions associated with SNA as a research methodology. The first is its ‘construct validity’, an issue that has potentially major implications. Simply put, the question is: do the social structures observed in, for example, Facebook, have real-world analogies or are they properties only of the online world, entirely unrelated to its real world counterpart? If the answer is no, then arguably, for all the talk about the opportunities for social research offered by new sources of social data, the impact in terms of increased understanding of social phenomena will be very limited.
Ackland and Zhu’s second question relates to debates about the capacity of social research methodologies to distinguish between causality and correlation. Here, they offer a somewhat more optimistic prognosis, observing that data generated through people’s activity on, for example, social networking sites, is rich and time-stamped, allowing for more fine-grained analysis, while the sites themselves can be thought of as natural research instruments, ideal for carrying out large scale experiments.4 Like other contributors to this volume, they conclude with a warning about the pitfalls for researchers of relying on data sources, such as Facebook, that are proprietary and whose access is subject to terms and conditions that may change at any time.
Chapter 11: Visualizing Spatial Data and Social Media
As earlier chapters have emphasized, the social data landscape is changing at an ever-increasing pace. The ways in which data is visualized has always played an important role in its analysis and in the presentation of results, and the ever-increasing volumes of data raise new challenges for visualization methods and tools. In this chapter, following a brief history of geographic information systems (GIS), Batty and his colleagues describe new ways of visualizing social data, with a particular emphasis on mapping. They argue that Web 2.0 mash-ups, layering geographically tagged social data on top of digital maps, enable quick and simple visualization of data, presenting research outcomes in ways that can be easily understood by diverse audiences.
Many of the examples the authors present emphasize how much researchers can achieve using simple, generic technologies and services such as Google Maps and Fusion Tables. Helpfully, Batty and his colleagues at UCL’s Centre for Advanced Spatial Analysis (CASA) have packaged these services into useful tools (such as MapTube5), which not only enable the geo-mapping of datasets with a few button clicks, but also provide ways for researchers to share and re-use each other’s efforts.
Another way in which advances in visualization techniques have harnessed the increase in computer power and new sources of data is the creation of fly-through, 3D models and visualizations of, for example, urban environments. More mundanely, but perhaps of greater value to researchers and planners involved in urban science, and the latest of many research areas predicted to be transformed by the advent of big data,6 are CASA’s ‘city dashboards’, which integrate diverse sources of data to create a real-time visualization of the state of the city and its inhabitants. Example applications include visualizing in real-time the state of mass transit systems. Such tools can provide powerful and intuitive front-ends to the simulations and models presented in Chapter 6, allowing, for example, exploration of the impact of closure of parts of the system.
Batty and his co-authors stress the importance of crowdsourcing and ‘citizen science’ for creating resources accessible to the public and illustrate this with the example of Open Street Map, a free map of the world.7 They conclude with some thoughts on the future of visualization as a tool for social scientific investigation and understanding. They predict the emergence of radically different kinds of tools that make use of more abstract forms of visualization, with an increasing emphasis on the use of non-spatial data as the way forward for understanding how social systems function.
Chapter 12: Ethical Praxis in Digital Social Research
Current approaches to ethics no longer seem adequate for twenty-first century social research. We have already noted the concern registered by the authors of preceding chapters about