Thesis: Ethical Considerations

This post is one of a serialisation of key (edited) extracts of my thesis, “Digital Ethnography and a Virtual Orkney: The Role of Folklore in Creating an Online Orkney Place”, submitted and completed in 2018. You can download the full text below, with all footnotes, full data quotes, and list of references. I have left the original numerical chapter navigation in this post, for ease of reference.

3.4 Ethical Considerations

Significant consideration has been given as to how to present and discuss the results of this research. All posts and comments considered throughout the research were collated from public-facing accounts, Groups and Pages. In this context, public-facing refers to posts and comments which are, in theory, accessible via the web to anyone. This is similar to the approach undertaken by Zebracki’s study where data was collated invisibly from social media platforms, in order not to influence discussions and commentary:

The virtual ethnography, moreover, involved covert observations, where I did not have any direct interaction with online users to not influence any of the communicative threads and respect the authenticity of user-created content.

(2017, p.452)

Though to some extent the fact that all the collated data was in the public domain simplifies the collation and analysis of the data, in that data which is not publicly accessible has not been used, it nevertheless does post ethical questions relating to how the data which has been collated and subsequently analysed can be discussed throughout the research. In using data which is potentially accessible to anyone via direct weblinks or by using search engines, this means that any direct quotes may be traceable to named individuals’ personal social media accounts. The Association of Internet Researchers (AOIR) Working Committee guidelines on Ethical Decision-Making and Internet Research states that the challenge for internet researchers remains “contextual requirements with disciplinary, institutional, legal, cultural, or other constraints” (Association of Internet Researchers 2018), discussing three primary tensions relating to: human subjects; the differentiation between public and private, and; the relationship between data and persons. This research has used the AOIR Working Committee’s guidelines as a starting point for ethical considerations.

In advance of the Twitter Hour and the Hosted Hashtags, the purpose of the research was made clear, with a brief discussion regarding ethical considerations in a blog post which also drew the reader’s attention to the AOIR guidelines as a reference point for the ethics of using social media data in research. For accounts which are set to private, there was an assumption that users would not want their data to be directly cited unless they contacted the researcher to specifically stipulate they were happy for their comments to be included:

As a default, anyone who joins in the discussions but has their account set to private will not have their comments referenced in my research, unless they contact me otherwise to explicitly indicate they are happy for their comments to be included.

(Crow 2016a)

Given the manner in which the Twitter Hour and Hosted Hashtags were advertised – with clear reference to this blog post and a clear statement in the researcher’s Twitter biography regarding the research topic – and the fact the hashtags were instigated and the questions posed by the researcher, consent could be assumed for quoting the Twitter data even where not explicitly given. As a result, and noting the fact that the four hashtags themselves are searchable and easily accessible, direct quotes will be used when discussing and citing user participation in the Twitter Hour and Hosted Hashtag.

Some sensitivity is needed, however, when quoting the Facebook data, to ensure that direct quotes cannot be linked back to named individuals’ personal social media accounts, even though all comments have been made in the public domain. For the purpose of discussing and quoting data, therefore, the following differentiation has been imposed:

  • Public Accounts and Pages: where a Facebook account has been created by a business or another specifically public-facing organisation (such as a museum or an individual’s business), or a Page whose primary purpose is to advertise such a business or organisation.
  • Personal Accounts: accounts which belong to named individuals, who show no sign of using their accounts for business purposes or promotion.

Where accounts have been deemed to be associated with individual personal accounts, direct quotes have not been used. There is one primary exception which has been taken into account when discussing the research, and that relates to users who incorporated acknowledged hashtags into their posts or comments in a bid to promote their own comments and/or link with wider conversations across a single platform or multiple social media platforms. Where there are such examples in the data, direct quotes may appear in the research if it appears that promoting the individual post or comment was the primary purpose of incorporating the hashtag. Otherwise, posts and comments from personal user accounts will be described, but not directly quoted. Additionally, the specific Groups and Pages which were used throughout the study have not been specifically listed (though some of the Facebook Pages are quoted directly in the data). Though this does not mean that it is impossible to find the specific quote and individual user, it nevertheless seems an appropriate way of responding to the AOIR’s guidelines and satisfactorily responds to the need to consider the context in which the data is being discussed with reference to human subjects, the differentiation between public and private, and the relationship between data and persons.

McNeill notes that “anonymity is not simply an unavoidable result of the serial collaborative process but rather a desirable and sought-after quality to many participants” (2007, p.294-295). Though the data collated during this study may not be anonymous to the same extent as those who contributed to the types of serial collaboration described by McNeill, the principles respecting the right to anonymity or to use user data responsibly remain the same. The ethics of the Cambridge Analytica scandal relating to the harvesting of Facebook data from millions of users and using it to influence political events have, unsurprisingly, been criticised internationally (Lewis et al. 2018). This has created an environment in which the ethics of the use of social media datasets – even those obtained legitimately or in the public domain – are under increased scrutiny. In the absence of a definitive and more prescriptive ethical code of conduct for internet researchers, it is suggested that the guidelines of the AOIR should be interpreted as cautiously as possible.

The methodological approach undertaken throughout this research makes use of platform functionality and the researcher’s detailed knowledge of the platforms chosen for the three case studies, to collate and analyse datasets capable of investigating to what extent there could be said to be an online Orkney place and the role that folklore plays in relation to the creation of this place.

Leave a Reply