Methods and Ethics of Internet Research
Simon Moreton
The increased availability of ICT has offered new opportunities to researchers. These include new ways of writing and documenting work (word processing etc), recording information (digital Dictaphones, video cameras, mobile phone cameras), and sharing research findings (blogs, social media, online journals). These changes support new epistemological practices of writing, creating, and sharing knowledge. These technologies also generate new social practices, as we have seen outlined in this document so far. The rise of Internet-based communities, technological mediated interpersonal communications, and myriad digital platforms through which to perform everyday life and self-making all offer new phenomena to research.
There is a clear imperative to better understand how we engage with research practices online. This section examines some literature about the ethical challenges and methodological approaches adopted when researching through online media.
Methods
Beaulieu considers “a number of ethnographies for the ways in which ‘objects’ are made, and the role explicitly and implicitly attributed to technologies in their constitution” (2004 p. 143). Many online research methods are analogous with those in the offline/onsite research. Surveys, for example, will be constructed in similar fashions but differ in methods of transmission; ethical approaches to interviewing or gathering data through personal communications are applied similarly to online and offline practices, though again the transmission medium might vary e.g. letter vs. email. However, the multitude of platforms available through which to conduct research on the Internet, means that we must be constantly attuned to developing new approaches for research (Beneito-Montagut, 2011).
Beneito-Montagut (2011 p. 725) refers to an ‘expanded ethnography’. This takes into account the need for numerous, cross-cutting methods to capture the complexities of everyday life:
“It is expanded in the sense that the research strategy adopted is and necessarily has to be as complex as the object of study itself. It also recognizes the fact that groups and relationships on the internet have diverse and complicated organizational settings, and they do not always take place in the public space or at the same site.”
There are three components to pursuing this methodology: the recognition of multi-situated actors, e.g. those using numerous types of ICT/connection strategies; the recognition of online/offline interplay e.g. where relationships mediated or forged online are used to create ‘in-person’ meetings and vice versa; and flexible multimedia techniques to engage with participants and produce a rich empirical account of their activities – these include interviews, diaries, multimedia tools etc. Of note to our study, are the use of research diaries and other methods to create a set of reflexive depictions of everyday interaction with communication technologies
What is important in this approach to methods is a recognition that “True involvement of research subjects presupposes open-mindedness and methodological flexibility on the part of the researchers. Subjects did not come aboard simply to refine the language of our questions to them” (Bakardjieva and Feenberg, 2000 p. 238).
Another study that ‘follows’ the activities of online subjects is that offered by Oulasvirta et al. (2010). Their analyses of microblogging practices among users of the microblogging service ‘Jaiku’ over its first 10 months was undertaken to understand the motivations behind, and meanings derived from different forms of microblogging practices. The work centred around an analysis of what sort of postings were sent, for example those relating to food, work, news or other extra-ordinary events. They unpack user platforms, such as telephones or desktops, the frequency of enagements and response to those engagements from other users. The aim is to understand the temporalities and spatialities of microblogging in relation to more established forms of self-making on the internet, such as weblogs.
The researchers acquired “all public postings and comments sent in Jaiku between October 26, 2006, and June 29, 2007” (Oulasvirta et al., 2010 p. 140, emphasis in original). Their method was to take a random sample of 1,600 ‘jaikus’ (analogous to tweets) and assess what was said in these short utterances. They coded these tweets in accordance with a rigorous method to create categories, such as ‘Direct Addressing’ (e.g. a greeting to another user), a ‘Factual Claim’ e.g. a piece of software is not working, ‘weather’ and ‘transient state’ e.g. hungry, tired etc. This taxonomy then provided the backdrop for their analyses, categorising each of these into further groupings, such as activity, experience, informational and conversational (ibid p. 242). They also tracked responses to individual Jaikus, that is messages that were aimed at specific users in response to an original posting.
This study does raise ethical questions of the sort identified already regarding public and private data. Although the authors point out the public nature of the ‘jaikus’, the tendencies they note in the microblogging practice – towards self-exposure, revealing of personal information etc. – suggest users might not conceptualise the act of microblogging as a wholly public one. This leads towards questions of a right to non-exploitation on the part of researchers: are these users likely to be happy with analyses of their microblogging utterances? Because consent was not given, but assumed due to the public nature of microblogging, does resistance matter? Do the jaiku users have ‘ownership’, or moreover, something to say, about their utterances that they have been unable to contribute to the research process because of the way the dataset has been managed?
One of the challenges associated with constructing an online-informed ethnography is a tendency for research to focus on the way Internet platforms are used to perform tasks such as work and communication, rather than the embeddedness of ICT use in everyday life. Beneito-Montagut (2011) for example suggests that of the work carried out on Internet use, there is a significant proportion that tackles the social effects of Internet usage and communities that arise out of that use. But, she notes, “studies are less numerous when we look at uses of the internet in a holistic way, scrutinizing its role in everyday life as a crucial part of communication processes and interpersonal relations ... Moreover, the role of emotions in online communication is still underexposed” (Beneito-Montagut, 2011 p. 717).
Another study that adopts this position is Valentine and Holloway (2002). They document the increasing ubiquity of ICT and comment how many early studies “uncritically celebrated on-line worlds as disembodied spaces, in contrast to the materiality of "real-world" environments” (ibid. , 2002 p.304). In particular they point out how virtuality – seen as the inauthentic or unreal – is something that somehow invades or compromises ‘the real’. In relation to this struggle to understand the ‘virtual’ in opposition to ‘the real’ Beaulieu (2004) notes that the way in which many social scientists have approached online ethnography has been underpinned by a belief that social phenomena on the Internet, such as the ability to create community are illusory. Calhoun (1998) and others (Driskell and Lyon, 2002) demonstrate how debates about the nature of Internet communities have a tendency to focus on the ‘communities of interest’ and ‘categorical identities’ that form online. These are perceived to lack even the groundedness in place from which to begin a programme of research.
Ethics
Constructing this methodology requires more than modifying existing methods to document the use of the Internet as a tool by individuals, groups, agencies or other bodies. This approach only finds new ways to describe activities that are theorised as being analogous, or at least relative to, activities that take part in ‘offline’ settings. It is instead a task of finding away to integrate an understanding of connectivity into how we understand the practice of everyday life. This enables a reflexive engagement with these practices to show they are intertwined.
Beneito-Montagut (2011 p. 717) argues that “everyday life takes place on the internet: there is no difference between online and offline interpersonal communication (IC)”. Another challenge pertains to rights of individuals over their public image on the Internet.
These new forms of technological practice thus offer new tools with which to conduct research. This might include participation in chatrooms, Bulletin Boards, or Multi-User Dungeons. It might include using email or video conferencing technologies to conduct research interviews. It might involve an analysis of data logs, communication transcripts and other digitally aggregate records of behaviour. Within all of these methods there are new ethical considerations to make. Madge (2007), writing about Internet research in the discipline of Geography, notes that there are five areas of research practice that need to be rethought in relation to Internet methods.
These are informed consent, confidentiality, privacy, debriefing and netiquette. Considerations around consent include being able to properly inform participants of the nature of the research and unintended outcomes of participating e.g. embarrassment, emotional distress; difficulties in verifying the identity of online research participants; the ethical challenges of researcher presence in online fora, specifically the acceptability of ‘covert observation’ such as in public chatrooms – and arguments that behaviours in chatrooms would change if the researchers self-identified; the Association of Internet Researchers (henceforth AoIR) have published guidlines for this type of research
Data protection and confidentiality raises very obvious questions. As with ‘situated’ research, online research has its on demands on adequate data storage practices, securing anonymity for participants etc. Madge (2007) points out the different conditions of online survey-based work – with organisations providing secure services for this practice, automatic anonymisation etc. – and participant-led, interview/focus group/chatroom based work. For these forms of practice, the onus is placed on the researcher to store data appropriately – such as through the use of encryption etc. – and to be aware of the nature of the subjects with which they are engaging. Madge points out that AoIR make distinctions relating to the nature of the work being carried out: in chatrooms for example, or with individual users, offering and maintaining participant confidentiality is critical; in more public fora, such as with authors and bloggers, it is less so, given the nature of public writing on the Internet.
Privacy also has its own set of concerns relating, but not reducible to, those of data protection and confidentialiy. The challenge of the private/public subject in Internet use is already a complex terrain (e.g. Curry, 1997) where traditional notions of private/public sphere are dissolved in the spaces of connectivity, information sharing and personal expectations of how technology operates. Bakardjieva and Feenburg (2000 p. 233) describe the problem as arising from: “the blurred distinction between the private versus public domain, the ease of anonymous and pseudononymous communication, and its global reach”. Solutions are not easily come by, but a rethinking of the validity of private/public dichotomies seems to be a good starting point, but also that fluidity and improvisation in methods be considered because “the scenario in the ethics of research on online communities that emerges from our analysis implies the necessity of more stages and iterations in the elaboration of the research design than are usual in off-line research.” (ibid. p. 239). Privacy, relating to new forms of subjectivity and objectivity, thus needs to be conceptually redrawn while paying attention to the same types of practices associated with offline research, such as pseudonyms, encriptions, participant engagement in transcripts etc.
Debriefing is also an established element of good research practice in offline/onsite research. This term refers to meeting, or otherwise communicating, with participants to disseminate findings, check on the status of the participant, and otherwise become attuned to the impact – if any – the research has had upon those directly involved. In the potentially impersonal/distanced aspects of online research run the risk of diluting the efficacy of debriefing processes.
Related to the question of debriefing and participation is content. Consent needs to be seen as more than permission.
Participation within online networks, such as chatrooms and message boards, is also an important element of Internet ethnography. One of the key practical challenges emerges from understanding the position of the researcher in that context. Does the researcher self-identify at the beginning of the process and risk alienating members, but ensure transparency? Does the researcher self-identify at the end of the observation process, thus maintaining the data integrity but compromising research relations? Or does the researcher never state their position as such? These three approaches risk alienation, exploitation of subjects and compromise the premise of ‘debriefing’, depending on how they are adopted.
Bakardjieva and Feenberg (2000 p. 238) recognise that “doing research online presents us with new ways to involve the virtual subject as collaborator in our project. Furthermore, we would argue, this circumstance suggests new possibilities for elaborating a situated ethical approach elegantly combining research objectives and methods with subjects’ right to non-alienation.” This concept of non-alienation is key in the distance world of Internet and technologically mediated communications.
The increased availability of ICT has offered new opportunities to researchers. These include new ways of writing and documenting work (word processing etc), recording information (digital Dictaphones, video cameras, mobile phone cameras), and sharing research findings (blogs, social media, online journals). These changes support new epistemological practices of writing, creating, and sharing knowledge. These technologies also generate new social practices, as we have seen outlined in this document so far. The rise of Internet-based communities, technological mediated interpersonal communications, and myriad digital platforms through which to perform everyday life and self-making all offer new phenomena to research.
There is a clear imperative to better understand how we engage with research practices online. This section examines some literature about the ethical challenges and methodological approaches adopted when researching through online media.
Methods
Beaulieu considers “a number of ethnographies for the ways in which ‘objects’ are made, and the role explicitly and implicitly attributed to technologies in their constitution” (2004 p. 143). Many online research methods are analogous with those in the offline/onsite research. Surveys, for example, will be constructed in similar fashions but differ in methods of transmission; ethical approaches to interviewing or gathering data through personal communications are applied similarly to online and offline practices, though again the transmission medium might vary e.g. letter vs. email. However, the multitude of platforms available through which to conduct research on the Internet, means that we must be constantly attuned to developing new approaches for research (Beneito-Montagut, 2011).
Beneito-Montagut (2011 p. 725) refers to an ‘expanded ethnography’. This takes into account the need for numerous, cross-cutting methods to capture the complexities of everyday life:
“It is expanded in the sense that the research strategy adopted is and necessarily has to be as complex as the object of study itself. It also recognizes the fact that groups and relationships on the internet have diverse and complicated organizational settings, and they do not always take place in the public space or at the same site.”
There are three components to pursuing this methodology: the recognition of multi-situated actors, e.g. those using numerous types of ICT/connection strategies; the recognition of online/offline interplay e.g. where relationships mediated or forged online are used to create ‘in-person’ meetings and vice versa; and flexible multimedia techniques to engage with participants and produce a rich empirical account of their activities – these include interviews, diaries, multimedia tools etc. Of note to our study, are the use of research diaries and other methods to create a set of reflexive depictions of everyday interaction with communication technologies
What is important in this approach to methods is a recognition that “True involvement of research subjects presupposes open-mindedness and methodological flexibility on the part of the researchers. Subjects did not come aboard simply to refine the language of our questions to them” (Bakardjieva and Feenberg, 2000 p. 238).
Another study that ‘follows’ the activities of online subjects is that offered by Oulasvirta et al. (2010). Their analyses of microblogging practices among users of the microblogging service ‘Jaiku’ over its first 10 months was undertaken to understand the motivations behind, and meanings derived from different forms of microblogging practices. The work centred around an analysis of what sort of postings were sent, for example those relating to food, work, news or other extra-ordinary events. They unpack user platforms, such as telephones or desktops, the frequency of enagements and response to those engagements from other users. The aim is to understand the temporalities and spatialities of microblogging in relation to more established forms of self-making on the internet, such as weblogs.
The researchers acquired “all public postings and comments sent in Jaiku between October 26, 2006, and June 29, 2007” (Oulasvirta et al., 2010 p. 140, emphasis in original). Their method was to take a random sample of 1,600 ‘jaikus’ (analogous to tweets) and assess what was said in these short utterances. They coded these tweets in accordance with a rigorous method to create categories, such as ‘Direct Addressing’ (e.g. a greeting to another user), a ‘Factual Claim’ e.g. a piece of software is not working, ‘weather’ and ‘transient state’ e.g. hungry, tired etc. This taxonomy then provided the backdrop for their analyses, categorising each of these into further groupings, such as activity, experience, informational and conversational (ibid p. 242). They also tracked responses to individual Jaikus, that is messages that were aimed at specific users in response to an original posting.
This study does raise ethical questions of the sort identified already regarding public and private data. Although the authors point out the public nature of the ‘jaikus’, the tendencies they note in the microblogging practice – towards self-exposure, revealing of personal information etc. – suggest users might not conceptualise the act of microblogging as a wholly public one. This leads towards questions of a right to non-exploitation on the part of researchers: are these users likely to be happy with analyses of their microblogging utterances? Because consent was not given, but assumed due to the public nature of microblogging, does resistance matter? Do the jaiku users have ‘ownership’, or moreover, something to say, about their utterances that they have been unable to contribute to the research process because of the way the dataset has been managed?
One of the challenges associated with constructing an online-informed ethnography is a tendency for research to focus on the way Internet platforms are used to perform tasks such as work and communication, rather than the embeddedness of ICT use in everyday life. Beneito-Montagut (2011) for example suggests that of the work carried out on Internet use, there is a significant proportion that tackles the social effects of Internet usage and communities that arise out of that use. But, she notes, “studies are less numerous when we look at uses of the internet in a holistic way, scrutinizing its role in everyday life as a crucial part of communication processes and interpersonal relations ... Moreover, the role of emotions in online communication is still underexposed” (Beneito-Montagut, 2011 p. 717).
Another study that adopts this position is Valentine and Holloway (2002). They document the increasing ubiquity of ICT and comment how many early studies “uncritically celebrated on-line worlds as disembodied spaces, in contrast to the materiality of "real-world" environments” (ibid. , 2002 p.304). In particular they point out how virtuality – seen as the inauthentic or unreal – is something that somehow invades or compromises ‘the real’. In relation to this struggle to understand the ‘virtual’ in opposition to ‘the real’ Beaulieu (2004) notes that the way in which many social scientists have approached online ethnography has been underpinned by a belief that social phenomena on the Internet, such as the ability to create community are illusory. Calhoun (1998) and others (Driskell and Lyon, 2002) demonstrate how debates about the nature of Internet communities have a tendency to focus on the ‘communities of interest’ and ‘categorical identities’ that form online. These are perceived to lack even the groundedness in place from which to begin a programme of research.
Ethics
Constructing this methodology requires more than modifying existing methods to document the use of the Internet as a tool by individuals, groups, agencies or other bodies. This approach only finds new ways to describe activities that are theorised as being analogous, or at least relative to, activities that take part in ‘offline’ settings. It is instead a task of finding away to integrate an understanding of connectivity into how we understand the practice of everyday life. This enables a reflexive engagement with these practices to show they are intertwined.
Beneito-Montagut (2011 p. 717) argues that “everyday life takes place on the internet: there is no difference between online and offline interpersonal communication (IC)”. Another challenge pertains to rights of individuals over their public image on the Internet.
These new forms of technological practice thus offer new tools with which to conduct research. This might include participation in chatrooms, Bulletin Boards, or Multi-User Dungeons. It might include using email or video conferencing technologies to conduct research interviews. It might involve an analysis of data logs, communication transcripts and other digitally aggregate records of behaviour. Within all of these methods there are new ethical considerations to make. Madge (2007), writing about Internet research in the discipline of Geography, notes that there are five areas of research practice that need to be rethought in relation to Internet methods.
These are informed consent, confidentiality, privacy, debriefing and netiquette. Considerations around consent include being able to properly inform participants of the nature of the research and unintended outcomes of participating e.g. embarrassment, emotional distress; difficulties in verifying the identity of online research participants; the ethical challenges of researcher presence in online fora, specifically the acceptability of ‘covert observation’ such as in public chatrooms – and arguments that behaviours in chatrooms would change if the researchers self-identified; the Association of Internet Researchers (henceforth AoIR) have published guidlines for this type of research
Data protection and confidentiality raises very obvious questions. As with ‘situated’ research, online research has its on demands on adequate data storage practices, securing anonymity for participants etc. Madge (2007) points out the different conditions of online survey-based work – with organisations providing secure services for this practice, automatic anonymisation etc. – and participant-led, interview/focus group/chatroom based work. For these forms of practice, the onus is placed on the researcher to store data appropriately – such as through the use of encryption etc. – and to be aware of the nature of the subjects with which they are engaging. Madge points out that AoIR make distinctions relating to the nature of the work being carried out: in chatrooms for example, or with individual users, offering and maintaining participant confidentiality is critical; in more public fora, such as with authors and bloggers, it is less so, given the nature of public writing on the Internet.
Privacy also has its own set of concerns relating, but not reducible to, those of data protection and confidentialiy. The challenge of the private/public subject in Internet use is already a complex terrain (e.g. Curry, 1997) where traditional notions of private/public sphere are dissolved in the spaces of connectivity, information sharing and personal expectations of how technology operates. Bakardjieva and Feenburg (2000 p. 233) describe the problem as arising from: “the blurred distinction between the private versus public domain, the ease of anonymous and pseudononymous communication, and its global reach”. Solutions are not easily come by, but a rethinking of the validity of private/public dichotomies seems to be a good starting point, but also that fluidity and improvisation in methods be considered because “the scenario in the ethics of research on online communities that emerges from our analysis implies the necessity of more stages and iterations in the elaboration of the research design than are usual in off-line research.” (ibid. p. 239). Privacy, relating to new forms of subjectivity and objectivity, thus needs to be conceptually redrawn while paying attention to the same types of practices associated with offline research, such as pseudonyms, encriptions, participant engagement in transcripts etc.
Debriefing is also an established element of good research practice in offline/onsite research. This term refers to meeting, or otherwise communicating, with participants to disseminate findings, check on the status of the participant, and otherwise become attuned to the impact – if any – the research has had upon those directly involved. In the potentially impersonal/distanced aspects of online research run the risk of diluting the efficacy of debriefing processes.
Related to the question of debriefing and participation is content. Consent needs to be seen as more than permission.
Participation within online networks, such as chatrooms and message boards, is also an important element of Internet ethnography. One of the key practical challenges emerges from understanding the position of the researcher in that context. Does the researcher self-identify at the beginning of the process and risk alienating members, but ensure transparency? Does the researcher self-identify at the end of the observation process, thus maintaining the data integrity but compromising research relations? Or does the researcher never state their position as such? These three approaches risk alienation, exploitation of subjects and compromise the premise of ‘debriefing’, depending on how they are adopted.
Bakardjieva and Feenberg (2000 p. 238) recognise that “doing research online presents us with new ways to involve the virtual subject as collaborator in our project. Furthermore, we would argue, this circumstance suggests new possibilities for elaborating a situated ethical approach elegantly combining research objectives and methods with subjects’ right to non-alienation.” This concept of non-alienation is key in the distance world of Internet and technologically mediated communications.
