The SAGE Handbook of Online Research Methods

Nigel G Fielding,Raymond M Lee,Grant Blank

• Editorial: SAGE Publications Ltd
• Año: 2016
• ISBN:
  9781473959286
• Edición: 2
• Categoría: Sciences sociales
• Categoría: Recherche scientifique et méthodologie

Online research methods are popular, dynamic and fast-changing. Following on from the great success of the first edition, published in 2008, The SAGE Handbook of Online Research Methods, Second Edition offers both updates of existing subject areas and new chapters covering more recent developments, such as social media, big data, data visualization and CAQDAS.

Bringing together the leading names in both qualitative and quantitative online research, this new edition is organised into nine sections:

1. Online Research Methods
2. Designing Online Research
3. Online Data Capture and Data Collection
4. The Online Survey
5. Digital Quantitative Analysis
6. Digital Text Analysis
7. Virtual Ethnography
8. Online Secondary Analysis: Resources and Methods
9. The Future of Online Social Research

The SAGE Handbook of Online Research Methods, Second Edition is an essential resource for anyone interested in the contemporary practice of computer-mediated research and scholarship.

Part I Online Research Methods

1 Online Research Methods in the Social Sciences: An Editorial Introduction

Raymond M. Lee, Nigel G. Fielding and Grant Blank

Online research methods have come of age as the permeation of everyday life by information and communication technologies has grown ever more ubiquitous. Although substantial digital divides remain by country, and within countries by age, gender and socioeconomic status, the number of Internet users worldwide quadrupled between 2000 and 2014, and the current proportion of the world's population using the Internet is now said to be in excess of 40 per cent (International Telecommunications Union, 2015). Information and communication technologies have had socially transformative effects. They increasingly affect how people make and maintain social relationships, the structure of their social networks, how they go about their work, meet their partners, educate their children, how they shop, take their leisure, present themselves to the world and store their memories. Such things are, of course, of interest to social scientists in and of themselves. However, to study them also requires methods of communication, ways of harvesting and capturing information, observational strategies and tools for collaboration, not to mention analytic techniques adapted to what are often novel forms and volumes of data, all of which themselves have the capacity to be transformed by new technologies.

Introducing the first edition of The SAGE Handbook of Online Research Methods, we emphasised the newness of online methods, and the need for a cautious and critical appraisal of their use and potential. Less than a decade onwards, the terrain occupied by online research methods has changed rapidly, social researchers across a wide range of social science disciplines have become much more familiar with such methods, more adept at their use, and more attuned to the issues and challenges that they pose. As before, our primary purpose in this Handbook is to explore this terrain by highlighting across a wide range of areas the key facets of online research methods and their implications for practice. Given our focus, as was true of the first edition, we pay relatively little attention to theoretical discourses on the wider social or cultural significance of online environments. While we recognise the historically contingent and socially constructed nature of the changes wrought by development of Internet-based technologies, we leave investigation of such issues to others. So too, we take largely as a given the infrastructural ‘substrate’ (Star, 1999) that underpins online practices; the standards, protocols, mechanisms, tools and resources without which activity online would be impossible. Neither do we address in any systematic way the drivers of methodological innovation in the social sciences and the social processes that have allowed new online methodologies to be adopted, diffused and used. The Handbook, in other words, retains a pragmatic focus on the current state of the art and on the further potential of online research methods in the social sciences.

Designing Online Research

Readers will find in the Handbook comprehensive and detailed coverage of a wide range of online research methods, some possibly more familiar than others. Clearly, though, there are wider issues that crosscut the investigation of particular research problems or the use of particular research methods. For example, in designing a particular study it is necessary to assess how far one's methods and procedures meet the aims and objectives set out for the research, and researchers must attend to the ethical issues surrounding their research.

One debate that emerged early on in relation to online methods was the question of whether the ethical issues they posed were distinct and unique compared to those associated with offline methods. In their chapter on the ethics of online research, Rebecca Eynon, Jenny Fry and Ralph Schroeder argue for the essential continuity between online and offline methods in relation to research ethics, and although they recognise the importance of ethical governance frameworks they emphasise the importance of the need to make ethical judgements in a context-dependent way. They usefully address the issues that arise in a number of different research situations that include the risks and benefits involved in using online methods to gather data directly from individuals and the challenges involved in obtaining informed consent in such situations, the sometimes novel ethical questions that arise when researchers directly study social interactions in virtual environments, and the increasingly important area of how data generated by social media might be analysed in an ethically responsible way. They bring their chapter to a close by pointing to the challenges posed for online researchers by issues relating to the fluid boundaries between public and private, the potential that arises in some cases for third party reuse of data, the complexities that come with a growing interdisciplinary focus in online research and the implications for ethical practice posed by the existence of digital divides.

Although ethical and legal frameworks provide a largely inescapable context within which a given research project must be conducted, the specific methods used in the study need, of course, to be carefully weighed and considered in relation to its aims. This has, perhaps, not always been the case as far as online methods are concerned. Their relative newness has in the past prompted both unthinking enthusiasm on the one hand, or unreasoned resistance, on the other. There is merit, therefore, in taking a careful, balanced and nuanced approach to the strengths and weaknesses of online approaches.

A putative advantage of researching online is that data can be acquired quickly and often in considerable volumes. The temptation exists, thinks Karsten Boye Rasmussen, to accept the benefits this brings without a parallel commitment to scrutinise the quality of the data so produced. In his chapter Rasmussen argues the need for a systematic theoretical model of data quality as a basis for assessing the ability of online methods to generate reliable and valid data. Emphasising within this framework the importance of ‘fitness for use', Rasmussen points to opportunities for assessing data quality that arise as traditional research methods such as the survey move online. In addition, online methods provide novel sources of data with a built-in capacity for quality assessment. In both cases the potential to assess and ensure data quality is enhanced by the ‘documentality’ of online data – in other words its ability to be described via ‘metadata', as well as the ability to associate it with ‘paradata’ – the data produced as part of the process by which data are collected.

Online Data Capture and Data Collection

One can argue with probably only a little exaggeration that for much of the twentieth century direct elicitation was the method of choice for many social scientists. In other words, it was thought that the way to discover what people thought and did was to ask them directly, usually by means of an interview of one sort or another. A relatively unnoticed aspect of this was that the popularity of the interview as a method depended on a variety of technological developments including in the case of qualitative research the miniaturisation of audio recorders (Lee, 2004) and the advent of long-distance telephone lines that, in the United States at least, fostered the development of telephone survey interviewing. In the twenty-first century, there has been a decisive move away from elicitative methods. This shift has largely been fuelled by a massive extension in the availability of online communication technologies, and by a growing ability to measure more and more aspects of everyday life as and when they occur through the use of data harvested from social media sites. Where even a few years ago names like Twitter, Facebook, Instagram and the like might only have produced quizzical bemusement if not puzzlement, particularly among older social scientists, social media data available in large volumes now form an increasingly large part of the landscape of social science research.

As this Handbook indicates online research methods are very diverse. They are used across the social science disciplines and produce data, whether directly elicited and not, that manifests itself in numeric, graphical, textual and audio-visual formats. The contexts within which online data are produced range from tightly designed experiments through to looser more naturalistic approaches, the gathering of various forms of non-reactive data, not to mention simulations and games or research in virtual environments. Claire Hewson traverses this terrain in her chapter on designing online research. Emphasising the importance of maximising the trustworthiness, reliability and validity of data produced online, Hewson systematically examines the possibilities, trade-offs, constraints and opportunities researchers need to consider when generating obtrusive and unobtrusive research data online.

The machine-readable traces that our increasingly self-documenting and self-archiving world leaves behind can be thought of as ‘unobtrusive’ or ‘nonreactive’ measures, to use a term popularised by Webb et al. (1966) half a century ago. Their now classic monograph was partly meant as a rebuke to the often uncritical use of interviews and questionnaires common at the time they were writing, but it also emphasised the creative appropriation of often quite fleeting behavioural manifestations as sources of data. In his chapter, Dietmar Janetzko attempts in particular to extend the conceptual understanding of nonreactive data by examining ways in which the rather ‘thin', i.e. non-contextualised, nature of such data can be extended either through triangulating multiple sources of data or the use of newer techniques such as text mining. Janetzko also enumerates the many different sources of nonreactive data to be found online and provides a detailed guide to the complexities of using such material.

As Ayelet Baram-Tsabari, Elad Segev and Aviv J. Sharon point out in their chapter, the term ‘data mining’ is relatively new in the social sciences but has become increasingly used in the last decade, fuelled it would seem by the growing popularity of online user-generated content. Data mining involves the automated processes associated with the extraction of knowledge from large-scale databases or online repositories. Baram-Tsabari and her colleagues usefully set out how data mining approaches differ from traditional quantitative methods. They examine the characteristics that make datasets suitable for mining as well as the resources needed to analyse them. In the main part of their chapter they give state-of-the-art examples of data mining techniques in relation to studies of mainstream media, data generated by users of social media and metadata.

As do other contributors to this Handbook, Martin Innes, Colin Roberts, Alun Preece and David Rogers see the need for a discerning approach that cautiously welcomes the opportunities created by the abundance of social media data now available while at the same time critically evaluating the social and technical processes implicated in their production, consumption and use. Innes and colleagues guide readers to an understanding of social media instrumentation, providing in the process an overview of how the data available on various social media platforms might be accessed. They also provide a detailed case study of how they combined to mutually implicative effect analysis of the social media data surrounding a particular event with on-the-spot ethnographic observation taking place at the same time.

In his chapter, Jonathan Bright investigates the issues surrounding the use of ‘big data’ in the social sciences, the large volumes of data about diverse aspects of social life that have become available as the ability to store and process such volumes becomes computationally possible. Bright provides an introduction to methods for capturing big data, as well as the processes involved in rendering the material more useful for analytic purposes through proxy variables and data coding. He then goes on to point to some of the complexities surrounding the analysis of big data, taking a somewhat sceptical view of some elements of current practice. Arguing that the methods training currently available to social scientists is seldom sufficiently oriented to the skills needed to work with big data sources, Bright describes some of the specific elements that make up the toolkit that social scientists increasingly need in order to be able to deal adequately with large datasets.

The Online Survey

Survey researchers have rarely shied away from the latest technological developments available to them and, true to form, were not slow to explore the possibilities for survey deployment opened up by the Internet. In both market and academic research, the use of online surveys is now well established. Nor has development been in any sense static. Researchers have begun to adapt to newer circumstances such as the growth in the use of mobile phones, while looking forward to possibilities that currently remain on the horizon such as the use of smart televisions as survey delivery systems.

Vasja Vehovar and Katja Lozar Manfreda give an overview of the current state of the art in their chapter on online surveys. Conceptually they locate online surveys within a wider set of technologically mediated data collection methods collectively referred to as ‘computer-assisted survey information collection’ (CASIC). As Vehovar and Manfreda observe, online surveys provide some of the traditional benefits of self-completion methodologies, but with advantages over conventional paper and pencil methods that include cost and error reduction, the possibility to increase respondents’ motivation and understanding, as well the ability to use advanced design features not available within non-digital contexts. On the other hand, if researchers are to make effective use of online survey methods, they need to confront a range of issues and challenges. Among the considerations outlined by Vehovar and Manfreda are issues to do with recruitment, sampling and non-response, how design elements are used within a survey instrument and the use of post-survey adjustments. They then extend their discussion to the use of single and mixed-mode surveys as well as mixed-method approaches. Many of these topics are subsequently taken up in detail in the other chapters making up this section of the Handbook.

In his chapter on sampling methods for web and email surveys Ron Fricker swiftly but carefully rehearses the fundamentals of sampling before going on to review the applicability of a range of probability and non-probability sampling methods to online surveys. He profiles the various methods of sampling – including the use of pre-recruited panels – that might be used and looks at the issues and challenges associated with their use. As do other writers in this section, Fricker recognises that the difficulties involved in generating probability samples online encourages the use of mixed-mode surveys. Fricker concludes with a look to the future, suggesting that in the shorter term online survey sampling is likely to remain problematic, but noting that with online technologies still in their infancy it is unclear what the future might bring.

It is difficult to spend any time online without receiving a request to participate in an online survey. Low cost, ease of administration and apparent reach all combine to make survey delivery online attractive to marketeers, bureaucratic administrators and academic researchers alike. As Vera Toepoel points out in her chapter on online survey design, intriguing possibilities emerge from the move to online surveys, particularly in relation to mobile data collection, and the extension of survey materials beyond the merely textual. At every ...