Blurred Public-Private Boundaries and Ethical Practices Through a Case Study on Instagram

’- “by problematizing simple understandings of what constitutes “publicly available data,”” ⤴️

  • “thereby complicating common “consent waiver” approaches” ⤴️
  • “in Lauren Berlant’s concept of “intimate publics” to” ⤴️
  • “the different understandings of “publics” ⤴️
  • “the case for a more reflexive approach to social media research ethics that builds on the socio-techno-ethical affordances of the platform to address difficult questions” ⤴️
  • “Based on the assumption that social media users have knowingly published the content in the public domain” ⤴️
  • “ethics is embedded in the research methods; it must be an ongoing concern during data collection, analysis, and dissemination.” ⤴️
  • “the boundary between public and private (Markham et al., 2018)” ⤴️
  • “privacy is instead about users’ sense of control over what information is shared with whom and in which context. Con” ⤴️
  • “some scholars argue that modifying people’s images is as unethical as showing the full image (Gross, Katz, & Ruby, 2003).” ⤴️
  • “even when researchers make attempts to protect users’ privacy by hiding the data source or de-anonymizing participants, we can never guarantee that users cannot be traced and identified via digital traces (Zimmer, 2010).” ⤴️ p41
  • “Zimmer outlines respecting contextual understandings of privacy and developing data anonymization strategies as important practices (2010) and Highfield and Leaver (2016) argue that instead of focusing on public–private binaries, researchers should consider the potential to do harm,” ⤴️
  • “If data are deemed to be public, or “naturally occurring” (Silverman, 2007), seeking informed consent from users is not necessary according to most institutional ethics committees.” ⤴️
  • “users might be revealing more information than expected through the image’s background” ⤴️
  • “Or they might be sharing content of others that has been acquired illicitly such as “creepshots,” that is, nonconsensual sharing of pictures (Highfield & Leaver, 2016).” ⤴️
  • “relying on a simple understanding of “publicly available” is not sufficient for social media research to be ethical (see also Locatelli, 2017 and Tiidenberg, 2018b for a discussion of this).” ⤴️
  • “Roland Barthes locates the blur of public and private at the very advent of photography. He writes that “the age of photography corresponds precisely to the explosion of the private into the public, or rather into the creation of a new social value, which is the publicity of the private: the private is consumed as such, publicly” (Barthes, 1980, p. 98).” ⤴️
  • “Marwick and boyd (2010) have suggested the term “imagined audiences” to conceptualize how social media users engage in self-presentations” ⤴️
  • “Rather than opposing the intimate and the public, Berlant (2008) conceives of the “intimate public,” often produced by media and marketing, as “constituted by strangers who consume common texts and things” (2008, p. viii). For Berlant, this public can be characterized by emotion and personal life, in ways that unsettle more traditional ideas of “public life” that emphasize collective action and civic and democratic debate (Habermas, 1962; Sennett, 1977).” ⤴️
  • “people often presume they know not only the common feeling but also the bounds of their audience.” ⤴️
  • “We used a snowball sampling strategy to broaden our search from these popular hashtags to ensure that diverse family representations were also included,” ⤴️
  • “digital researchers who may read a series of posts as performances of political identity when they were not intended as political, or indeed as data for research.” ⤴️
  • “As scholars, we may work with shared definitions of “the public,” or with legal or institutional cutoffs for what such notions of “public” will allow, but this does not mean that the spaces and people we are researching share this view or knowledge.” ⤴️
  • “When revisiting the data in 2018, some posts were no longer searchable, because the user had deleted the post, changed the account name, changed the privacy settings, or deleted the profile altogether. For ethical reasons, we decided not to include these posts in our sample” ⤴️
  • “what is the intended public here? And often, as our own efforts indicated, this is very difficult to discern without directly asking the owner of the account.” ⤴️
  • “our guidelines were to exclude posts no longer publicly available at the point of starting the analysis process (coding),” ⤴️
  • “Dobson et al. argue that, ironically, “public intimacies are not public enough in the sense that participants have little control over what platforms do with their intimate relationships” (2018, p. 22). Scholars scraping “intimate data”” ⤴️
  • “Berlant’s notion (2008) of “intimate publics” illuminates, people have different understandings of what they feel as being “their” public. We therefore decided to seek permission from account” ⤴️
  • “We conceptualize the challenges we faced as “ethically important moments” (Guillemin & Gillam, 2004). These are moments that capture unanticipated and challenging situations that emerge in the daily practice of doing social research, but are often overlooked or simply not reported. The” ⤴️
  • “a link to our research blog, as a means to verify our researcher identities. This” ⤴️
  • “We explored ways of creatively representing images in a way that de-identified the people pictured such as “cartoonifying” the” ⤴️
  • “Despite their technically “public” nature, the posts appeared to publicize “private” content, aimed at friends and/or family members, or specific intimate publics.” ⤴️
  • “when reengaging with the data in 2018, we found some users had set their accounts to private after we contacted them about the project1. This made us wonder how many of the users in our sample were actually aware that their privacy settings allowed not only family and friends to see their posts but also “complete strangers” like us,” ⤴️
  • “Ultimately, the question that arose for us was whether “publicly available” also means “available for academic scrutiny” (or commercial purposes, for that matter; cf. Highfield & Leaver, 2016)” ⤴️
  • “some accounts appeared to be professional profiles used to promote various commercial endeavors. These users were mixing family-related posts with posts about their business, thereby posing other questions about account holders’ own deliberate blurring of public/private boundaries” ⤴️
  • “In our efforts to discern what the account holder considered to be their “public,” we examined details such as the account name and description, other posts, and any links to external websites. This research was helpful in unpacking some of the complexities around how people determine and frame the audience for their Instagram posts, but ultimately it was too difficult to ascertain the purpose without speaking directly to the user.” ⤴️
  • “like Locatelli (2017) we did not want to represent users who had not granted permission” ⤴️
  • “we suggest thinking about the socio-techno-ethical affordances of the platforms we are working on, that is, drawing actively on the inbuilt functions in apps and platforms” ⤴️
  • “We are not suggesting an approach that requires “informed consent for everything” (see Salganik, 2018 for examples of cases when informed consent represents risks to participants or is impossible to obtain);” ⤴️
  • “This becomes particularly salient in cases like the present one where the data—Instagram posts—to a large extent include children.” ⤴️
  • “Our approach means that we are only able to engage with the data at the aggregated level (as Highfield & Leaver, 2015, also suggest), but not reproduce the large majority of the posts in our sample.” ⤴️
  • “This sometimes requires omitting revealing details despite their analytical lure.” ⤴️
  • “As Tiidenberg suggests, “Just because something is technically accessible and collectable, doesn’t mean it should be accessed and collected” (Tiidenberg, 2018b, p. 472). Rather than trying to de-identify posts as much as possible, as Highfield and Leaver (2015) suggest, we argue that we must engage with users and ask for consent before reproducing and representing their content” ⤴️