Sage Journals: Discover world-class research

Abstract

As researchers we often find ourselves grappling with social media platforms and data ‘at close quarters’. Although social media platforms were created for purposes other than academic research – which are apparent in their architecture and temporalities – they offer opportunities for researchers to repurpose them for the collection, generation and analysis of rich datasets. At the same time, this repurposing raises an evolving range of practical and methodological challenges at the small and large scale. We draw on our experiences and empirical data from two research projects, one using Facebook Community Pages and the other repurposing Facebook Activity Logs. This article reflects critically on the specific challenges we faced using these platform features, on their common roots, and the tactics we adopted in response. De Certeau’s distinction between strategy and tactics provides a useful framework for exploring these struggles as located in the practice of doing social research – which often ends up being tactical. This article argues that we have to collectively discuss, demystify and devise tactics to mitigate the strategies and temporalities deeply embedded in platforms, corresponding as far as possible to the temporalities and the aims of our research. Although combat at close quarters is inevitable in social media research, dialogue between researchers is more than ever needed to tip the scales in our favour.

Keywords

Facebook Community Pages Facebook Activity Logs social media research platforms tactics temporalities

Lacking its own place, lacking a view of the whole, limited by the blindness (which may lead to perspicacity) resulting from combat at close quarters, limited by the possibilities of the moment, a tactic is determined by the absence of power just as a strategy is organized by the postulation of power (de Certeau, 1984: 38).

De Certeau’s description of tactics will resonate with researchers’ experiences of using social media platforms and data. The issues encountered in this type of research largely ensue from the power relations between big social media corporations and researchers (often working on their own). In this article, we discuss the practical and methodological issues that we faced in two research projects which made use of Facebook as a source of data and a tool for research,¹ often fighting ‘at close quarters’. Stevie’s work examined the media coverage of the 2011 English riots from a media archaeological perspective. She used Facebook Community Pages as part of a larger media corpus to trace the media ecology of the riots. Justine’s research explored young people's engagement with social media and how these were informed by platform structures, corporate interests and the broader context of neoliberal capitalism. She incorporated Facebook Activity Logs as digital prompts during interviews to encourage discussion with her participants.

Although different, both of our projects used Facebook as a source of data about the past – be it distant (the 2011 riots) or very recent (users’ past activities) – and grappled with issues around the (re)shaping of these past events or activities by the platform itself and its temporal layers. During our research, we both became increasingly reflective about these issues and the impacts of Facebook on many aspects of our research. These reflections were furthered by our engagement with de Certeau’s (1984) seminal distinction between strategies and tactics. Such a distinction is useful to shed light on the complex struggles between social media corporations and researchers using and/or repurposing social media platforms.

By reflecting jointly on the different issues that arose in our research, we seek to elaborate on their common roots. In combining insights into different aspects of Facebook, we aim to build a greater appreciation of how and to what the extent the platform – its connective design and temporalities – informs and interferes in research practice. This is particularly valuable as social media platforms often work against research, and always confine researchers to ‘close quarters’. This kind of research dialogue will also contribute to ongoing debates around the uses but more importantly the repurposing of social media data and platforms in research.

Below, we briefly outline de Certeau’s distinction between strategies and tactics before contextualising it with reference to the emerging literature on social media and social research. We then outline our separate research projects as well as the practical issues that we encountered before discussing the commonality of the challenges we faced. We argue that we have to collectively discuss, demystify and devise tactics to mitigate the strategies and temporalities deeply embedded in social media platforms.

‘Making Do’: Social media platforms and social research

De Certeau defined strategy as the ‘calculation or manipulation of power relationships’ (1984: 35–36). A strategy has its own space which ‘serves as the base from which relations with an exteriority composed of targets or threats […] can be managed’ (1984). By contrast, tactics do not have autonomy or spaces of their own. They are negotiated and located within ‘the space of the other’:

It is a manoeuvre “within the enemy’s field of vision,” as von Billow put it, and within enemy territory. It does not, therefore, have the options of planning general strategy and viewing the adversary as a whole within a district, visible, and objectifiable space. It operates in isolated actions, blow by blow. It takes advantage of “opportunities” and depends on them, being without any base where it could stockpile its winnings, build up its own position, and plan raids. (de Certeau, 1984: 37)

In de Certeau’s account, strategies and tactics differ in terms of space and time. Strategies are located in delimited spaces and are often translated into a ‘victory of space over time.’ Tactics, because they do not have a space of their own, depend on seizing opportunities “on the wing”’ (1984: xix) – in other words, they are time-dependent.

Social media corporations enact their business strategies through the purposeful design and architecture of the platforms (spaces) they make available to users. These spaces have their own particular temporalities. As social researchers we are left with ‘only’ tactical moves to mitigate these strategies and temporalities that can often work against our research. We are dependent on and working in temporalities and spaces of which we have only a partial grasp. In this context, distance or foresight are difficult to achieve. This, as de Certeau’s opening quote vividly described, creates asymmetrical power relations between private corporations and researchers.

Since de Certeau’s writing, media scholars such as Lev Manovich (2009) have critically re-examined the distinction between strategies and tactics in the context of new media and user-generated content. The media landscape has changed dramatically with the development and spread of digital technologies in the everyday (Couldry and Hepp, 2016). Social media platforms have emerged and continue to thrive in a wider media ecology (Madianou and Miller, 2013; Strate, 2008). In this light, de Certeau’s dialectical approach of space and time has blurred as the changing media landscape has re-shaped existing spaces and temporalities and created new ones.

These are intimately linked with ongoing reconfigurations of time, such as ‘network time’ (Hassan, 2007), ‘Internet time’ (Karpf, 2012), ‘event time’ (Adkins, 2009), ‘real time’ (van Es, 2017; Weltevrede et al., 2014), and ‘social media time’ (Kaun and Stiernstedt, 2014), challenging the industrial-era dominance of mechanical ‘clock time’. Kaun and Stiernstedt (2014: 1155) used ‘social media time’ as an analytical framework to better grasp the specific ways in which the platforms organise and structure users’ time. For example, Facebook is often used to manage events, organise meetings with friends, be reminded of birthdays or as an archival repository for pictures and memories (2014). Kaun and Stiernstedt’s work provides great value in investigating users’ experiences of social media temporalities. Less has been said, however, on researchers’ experiences of these temporalities in the course of their work, which we will explore in this article.

As researchers and users of social media platforms, we are still in the dark about the details of how platforms, at the architectural level, actually work. This architectural opacity is further complicated by frequent changes to terms and conditions, the introduction of new features or the replacement of old ones (e.g. Facebook’s incorporation of live video streaming, or the introduction of the Timeline in 2011). Blackboxing persists around tools and methods in relation to social media data, including around the specific problems we encounter in research (Gayo-Avello et al., 2011; Kennedy et al., 2015). In some ways, this mirrors the problematic blackboxing of ends, operations and design elements by the platforms we collect and generate our data from (Pasquale, 2014; Paßmann and Boersma, 2017). Researchers are also vulnerable to inequalities of technical ability and resources (Rieder and Rohle, 2012). New computational tools and applications for working with digital data are constantly being developed. Many are open-source and/or freely available. But unless we have some technical understanding of how these work, we cannot expect to use them well in research.

We cannot confront platform architectures in research without also ‘combatting’ the temporalities embedded within them. Social media platforms tend to lack a sense of their own pasts (Brugger, 2015) – certainly at the level of the ‘look and feel’ of the user interfaces that most researchers interact with. Although they tend to erase traces of their own pasts, platforms like Facebook function as important sites of capture (Dean, 2010) for everything from everyday life to major events or crises (e.g. elections, terrorist attacks, disasters, scandals, riots, and revolutions, etc.). Media actively take part in the production of events rather than simply relaying them to audiences (Hoskins, 2009): events generate media data through user participation and interactions via platforms, which feed back into the events themselves. This impacts on the events while they are still happening, but also on the collective and personal recollection of them. These factors often have consequences for studying (aspects of) platforms as historical artefacts, or as sources of data about distant or recent pasts (the 2011 riots for Stevie, participants’ recent activities for Justine). Facebook’s ideology and design actively foster newness and connectivity (van Dijck, 2013), which prioritise ‘real-timeness’ (Weltevrede et al., 2014) over duration. Chun (2008) refers to the ‘enduring ephemerality’ of digital media, while Lievrouw (2012) highlights their ‘fugitiveness’. She points to the possible dangers of ‘histories written and judgements made on the basis of disconnected scraps of data’ that may not ‘survive more than a few years’ (2012: 630). We would argue that it is not necessarily objects or data disappearing completely that causes complications. It is their quickness: their propensity to fade unpredictably in or out of visibility over time, to alter in accessibility or simply to move around – to invite chase, while evading capture.

In the last decade, researchers have engaged with the epistemological and methodological questions raised by using social media platforms and more broadly Big Data in research (Boyd and Crawford, 2012; Burrows and Savage, 2014; Kitchin, 2014; Langlois and Elmer, 2013; van Dijck, 2014). Kitchin argued that the data revolution – in particular Big Data and new data analytics – is reconfiguring how research is conducted, often based on ‘new forms of empiricism that declare “the end of theory” [and] the creation of data-driven rather than knowledge-driven science’ (2014: 1). This new ‘scientific’ shift is fed by the belief that human behaviour can be turned into data and objectively tracked, quantified and analysed, often presupposing that data are collected through ‘neutral technological channels’ (van Dijck, 2014: 199). As a result, research has often failed to critically reflect on the epistemological implications of this trend and to question the companies and institutions which collect, analyse and share data. Though mostly focused on Big Data, these debates are relevant for the use and repurposing of social media platforms and data in smaller scale research.

These debates need to be understood in the context of the ‘platform capitalism’ (Srnicek, 2016) and ‘platformisation’ (Helmond, 2015) in which platforms have emerged as a dominant infrastructural and business model to extract, mine and control large amounts of data. Crucial works have examined how social media corporations using terms such as ‘participation', ‘sharing’ and ‘platforms’ have promoted the idea of a ‘social' Web and of ‘neutral’ platforms while pursuing economic vested interests (Fuchs, 2014; Gillespie, 2010; van Dijck, 2013). Furthermore, studies about and/or using social media platforms that treat platforms as ‘self-contained entit[ies] set apart from the rest of the Internet’ (Skeggs and Yuill, 2015: 1358) run the risk of overlooking the broader ecological context in which platforms thrive and generate data (e.g. Facebook plug-ins and widgets which are located on other webpages, see Helmond, 2015).

In addition, recent research has explored the complex impacts of social media designs and architectural structures on users’ engagement, and importantly on the meanings and imageries that they ascribed to the platforms (Bivens and Haimson, 2016; Bucher and Helmond, 2018). Social media architectures rely on code and algorithms which shape our engagement with the platforms while the notion and representations of algorithms affect and feed popular culture (Beer, 2017; Bucher, 2017). These contexts in which social media platforms are embedded are crucial to understanding how researchers as well as users engage with the platforms.

Moreover, researchers are also constrained by the temporalities of their own research which do not necessarily allow them to wait to ‘seize opportunities’. Such opportunities may not be available for a sustained period of time (e.g. specific features of platforms, access to data). This is often difficult as social media platforms constantly change. Over the course of our research for example, Facebook made multiple alterations to its privacy settings and conditions of use. In January 2018, following a drop in user numbers for the first time in the platform’s history (Kastrenakes, 2018), the company announced major changes to the platform’s newsfeeds and aims (Zuckerberg, 2018). Later on in the same year, the Cambridge Analytica scandal broke (Solon, 2018), forcing the company CEO to appear in front of the US Congress and the Senate in which he pledged to review the procedures for data access by third parties. This scandal demonstrates yet again the pressing need for policy-makers but also researchers to examine and challenge private corporations’ attitudes regarding users’ consent and data-mining. These issues speak to a broader lack of transparency regarding the collection, exploitation and sharing of personal data through and by platforms, as well as the unaccountable access of personal data through third-party applications by researchers. Following the scandal, Facebook announced new restrictions to accessing data via the Application Programming Interface (API). These have been criticised as a ‘quick-fix’ solution which does not address the bigger issues aforementioned, and is at risk of adversely affecting social research without opening up a discussion about best practices and the ways in which data can ethically be part of social research (Bruns, 2018).

By ‘combatting’ Facebook at close quarters and ‘making do’ with the platform, we placed ourselves in the position of reproducing the ideology of connectivity promoted by Facebook and of further normalising data-mining. However, our outsider positions can also work in our favour in repurposing Facebook in tactical ways to create new awareness, and deflect existing power relations. In this way, de Certeau’s distinction between strategies and tactics is very useful not only to explore these power relations but also to understand the struggles as located in the practice of doing research – which often ends up being tactical. The remainder of this article examines these tensions by reflecting on the practical dilemmas encountered in our two research projects.

The use of historic Facebook Community Pages

Stevie’s research explored the media ecology in which the 2011 English riots took place and applied a media archaeology approach to investigate how this had persisted over time. As means of investigating ‘new media cultures through insights from past new media’ (Parikka, 2012: 2), media archaeology can be particularly valuable in contexts where past new media and the potential insights they offer may quickly disappear (2012). Stevie sought to trace links out from, and between, samples of media coverage of the riots. She pursued these links on two levels: (1) retrospectively, across the remnants of the riotous media landscape; and (2) over the shifts in the landscape up to the present. This archaeological work involved navigating a diverse range of media content dating from the time of the riots in August 2011 up to September 2017, including a selection of Facebook Community Pages set up specifically in response to the rioting in different cities.²

Despite the prominence of Facebook and other social media in public discourse around the riots while they were happening (Fuchs, 2012), there has been little to no detailed study of the riots incorporating data drawn from Facebook. This shortfall motivated Stevie to integrate Facebook Community Pages into her research, allowing the tracing of riotous media moments, interactions and connections that derived specifically from the platform. The Community Pages functioned as a distinct space where the riots unfolded through a mix of user interactions and disclosures. In turn, this raised questions of how the Community Pages and the platform played into the riots over time in terms of citizen crisis communication (Palen and Liu, 2007), event construction and remediation (Bolter and Grusin, 1999). But it also presented difficulties because of the temporal layers involved, forcing Stevie to develop tactics for conducting the archaeological work.

Since Community Pages were introduced in 2010, anyone with an account can create and run a Community Page on almost any topic, within Facebook’s guidelines. To be visible to other users, Community Pages must first be ‘published’ by an administrator, who also controls the page settings and content (though pages can have multiple administrators). Unlike private Facebook groups, where users must be invited to ‘join’ before they can see content and take part, a user need only click to ‘like’ a page to begin interacting with it. Published Community Pages can be viewed to a limited degree by those who are either logged out or do not have a Facebook account.

The Facebook Community Page data took the form of PDF files each containing screen captures taken during the riots. Stevie was not involved in the collection process but was given access to the files when she began work. The pages were no longer accessible to her online in their 2011 incarnations, which impacted on how she approached her research – particularly from an ethical perspective. Following existing guidance on the use of social media data (Association of Internet Research (AOIR), 2012; Townsend and Wallace, 2016), she excluded any personally identifying information attached to posts or comments. In addition, Stevie ran searches while logged into her personal Facebook using the platform’s internal search bar, which confirmed the pages were no longer accessible, but also that snippets of text quoted in her written research output could not be traced back to individual users within Facebook. In total, the PDFs contained over 13,000 posts and comments from across eight Community Pages, spanning the period between the 8th and the 15th of August 2011.³ To gauge how to manage the volume of the data and address issues of data abundance (Floridi, 2012; Manovich, 2012), she first transcribed all the interactions from a single page into a database. This allowed her to note themes and patterns of interaction which she was then able to draw on in later stages of her media analysis. She then selected two further Community Pages and took smaller time-bound sub-samples. She then coded these sub-samples in NVivo, feeding iteratively into the ongoing media analysis and ethnographic work.

The main issues Stevie encountered during her research were tied to the different temporalities she had to navigate: the ongoing time of her research (which increasingly distanced her from the events she was studying); the extended time of the riots (a week in 2011); the precise time of the capture of the Community Pages; and the temporal structures and rhythms of the platform. From the outset, the Community Pages were removed from their native spatial and temporal contexts, stripped of their tactility and interactivity as a result of the PDF format. For instance, Facebook’s built-in comment count might state that there were 10 comments under a post, but not all of these may have been visible on screen at the point of capture. If the researcher taking the screenshot had not clicked to display the rest of the comments, then they were not shown in the PDF files and were out of Stevie’s reach. Facebook’s comment counter made it obvious that something else was there, or was missing, but the method of capture together with the passing of time made it impossible to know what, exactly.

In tactical terms, Stevie had to adapt to multiple temporal layers that emerged within the Community Pages. Kaun and Stiernstedt (2014) have pointed out the importance of the Timeline in ordering the everyday user experience of the platform and imposing a linear but not chronological structure. A similar visual linearity was imposed by default on the ‘wall’ of the Community Pages. This implied a tidy chronology of posts and comments, whereas the time and date stamps suggested a far messier sequence. Overall, there was no single ‘Facebook time’, only splintered and sometimes conflicting temporal versions of events. For example, a single post from the 9th of August might continue to attract comments dated days afterwards, which were threaded beneath the ‘original’ post. As a result, they appeared to precede other posts following them on the page, when in fact they came later. During the most intense periods of rioting – such as Monday, 8 August in London – a greater amount of content was being posted more quickly, creating more layers of comments, posts reacting to previous posts or comments, and new posts. The linear page structure worked to obscure the temporal messiness of both the page and the events unfolding on it by presenting the page as a coherent version of events. It encouraged an illusion of oversight – being able to see the whole page – while making it difficult to get close insight into the page, creating the ‘blindness’ de Certeau has described (1984).

Further, many of the links and connections Stevie tried to trace out from the Facebook pages unsurprisingly turned out to have broken down over time. In 2011, third-party applications like Yfrog or Twitpic were commonly used to share still or moving images of the riots by Community Page users. However, these applications have become mostly obsolete as platforms like Facebook have introduced or improved their own image-sharing solutions, and many are now defunct. Consequently, many of the images of the riots shared through these apps only exist online today as partial remnants, making it difficult to find out and verify information about them. For example, the admin of the page ‘Bring the Army to Deal with the London Riots’ posted the following link to a Yfrog video ostensibly depicting rioting in Seven Sisters in London. The link includes a thumbnail image showing a single frame from the video (see Figure 1).

In trying to follow the link out, Stevie found that Yfrog no longer operates and the video itself was inaccessible. A different problem was posed by YouTube videos shared on Facebook in 2011 which then turned out to have been removed or deleted from the former platform. Again, there was evidence of these deleted videos on the Facebook pages in the form of thumbnails. But these could provide only a glimpse of the ‘original’ content. Although this content – and the pages themselves – were no longer accessible online, this did not necessarily mean they were gone completely. The pages may have been ‘unpublished’ rather than deleted by the administrators, and importantly may persist on Facebook’s servers. This speaks to a wider struggle between social media corporations with their own strategies for data retention and researchers with differing interests.

The discussion above demonstrates three interconnected challenges that invited a tactical response. The first stemmed from the way the Community Pages had been captured. To an extent, the PDF format was beneficial, in that it allowed a partial snapshot of everything as it had been on the pages. The content (text, images, links, thumbnails, etc.) was captured together ‘in place’, which conveyed some of the multimodal richness of the page data. This differs from other technological means of capturing page data, which may return material in disaggregated forms. Although there is always a trade-off involved in any kind of data capture, it is complicated in the case of social media data by the particularities of the platform (space), the types of data available and the conditions of access to them. The second challenge arose from the multiple overlapping temporal layers embedded in the Facebook Community Pages and the platform itself. Lastly, Stevie’s research illustrates that Facebook cannot be considered in isolation from the wider media ecology it is part of, as the development of in-house image sharing and video streaming shows. Shifts across other media platforms and applications have repercussions for the development and adoption of Facebook design changes, features and uses, and vice versa. Research processes tend to take time, and therefore inevitably are affected by these shifts. Stevie was not in the position of having to rely on the continued existence of the Community Page feature to complete her research, as the data had already been collected. In contrast, Justine’s research was at ongoing risk of the Activity Log feature and its design altering during the time of her fieldwork.

Using Facebook Activity Logs in interview settings

Drawing on 32 in depth semi-structured interviews with young adults aged 20–25, Justine’s research explored their uses of social media platforms and the meanings they ascribed to these, focusing in particular on profile-checking and monitoring practices. During her interviews, she used Facebook Activity Logs as ‘digital prompts’ in order to investigate young people’s understandings of social media platforms. She asked her participants with their consent to access their Facebook Activity Logs to stimulate discussion. Activity Logs display a chronological list of one's activities on the platform (e.g. content posted, content liked and commented on, profile updates, friends added, events joined, etc.; Figure 2). Logs are a carefully selected and ranked (by Facebook) compilation of one’s activities on the platform. They do not encompass the full range of personal user data amassed by Facebook, but this ‘invisible’ data is effectively reintegrated into users' timelines, shaping the information they are exposed to.

Figure 1.

Yfrog link embedded in page post.

Figure 2.

Facebook Activity Log.

Recent qualitative studies have used aspects of Facebook in their designs, particularly the newsfeed or the timeline (see for example Duguay, 2016; Robards and Lincoln, 2017 or Boyd, 2015 for a combination of digital methods and ethnographic fieldwork). Justine drew on this literature as well as existing guidance on social media research (Townsend and Wallace, 2016). The use of Facebook Activity Logs in interview settings (as opposed to people’s activities on Facebook), however, has not been discussed in the literature. Using this feature as digital prompts raised distinct practical and ethical challenges. Research has shown that users tend to be aware of data-mining and monitoring (Lupton and Michael, 2017), but this awareness can remain abstract. It was therefore important to consider the possibility that participants might not know that their activities were systematically recorded by Facebook and/or available to them through their Activity Logs. To ensure participants' full informed consent, Justine explained the types of information their Activity Logs would contain. Participants were asked to view these on their own first and were given the opportunity to withdraw or reassert their consent at this point. In addition, to safeguard participants’ privacy and anonymity, no digital data was recorded.

It was also essential to weigh up the impacts of the Activity Log, a feature designed by Facebook, during the interviews and on the data collected. Studies have called attention to the curatorial role of Facebook in shaping how information is presented (design) and sorted (algorithms) (Bucher and Helmond, 2018; Robards and Lincoln, 2017). The company has introduced an array of features and functions such as ‘On This Day’, ‘Your Year in Review’ as well as the Timeline to facilitate access to one’s past activities as well as those of others. The introduction of these features, Robards and Lincoln argued, is part of a broader shift by Facebook increasingly positioning itself ‘as a site for reflection, nostalgia, and looking back’ (2017: 718). Such shifts and changes of design impact on how social media platforms are understood by users, while encouraging them to build, arrange and adjust their biographies in a coherent trajectory, a timeline. Using participants’ Timelines in interviews presents users with this linear construction, possibly affecting the ways in which they discuss their biographies. Similarly, the linear design of the Activity Log shaped Justine’s research in subtle ways.

Tactically repurposing Facebook Activity Logs as digital prompts created connections which would not have necessarily been made without them. As expected, digital prompts encouraged dialogue and provided material to discuss. For example, Eleanor’s Activity Log prompted discussion around implicit norms of behaviour and social expectations on Facebook:

[Looking at her Activity Log] So that is my sister sending me … what is that? Oh yeah my sister tagging me on a lot of cats pictures and things like that so I try not to hurt her feelings and I said yeah I like it too (laughs), it is a bit mean to say but she is sending too much things like that. (Eleanor, 22)

In this instance, the Activity Log encouraged discussion by providing specific examples (e.g. her sister’s tagging her in trivial pictures) in turn creating thick data regarding Eleanor’s understandings of the platform and its uses. Activity Logs were also used spontaneously by some participants to substantiate their stories, recall their past activities (including very recent ones) and/or ‘check’ their answers to factual questions (e.g. ‘what is the last thing you posted on Facebook?'). Luke, for example, used his Activity Log to remind himself of his recent activities:

Researcher: Do you have an example of the last thing that you shared on Facebook?

Luke: I may have to check (laughs) I am not one hundred per cent sure... [Checks his Activity Log] I liked a photo yesterday, it was added by one of my friend, we were away in South Ayrshire for the weekend so one of my friend posted a picture of us. I commented on a post on Saturday that someone posted on my timeline... and then before that it was Wednesday. (Luke, 21)

Furthermore, Luke presented his activities within the timeframe sequence suggested by the log (i.e. yesterday, Saturday, Wednesday) which helped him to re-contextualise them. This illustrates how his recollection was remediated by the design of the log and its presentation as a timeline. Participants often relied on the data of the Activity Log, either spontaneously or when presented with it, to re-contextualise their activities (e.g. according to date, or previous and subsequent entries). The majority of participants did recall each of the activities listed on their logs: crucially, however, they would not have necessarily done so without making use of the record. The display of participants’ activities during the interviews thereby created a new setting in which participants retrospectively reconstructed the meanings and contexts of their activities, coinciding with work about the mediatisation of memory in digital settings (Hoskins, 2009).

In addition, the design of the platform, generally encouraging connectivity and newness, and in particular the layout of the Activity Log itself, condensed participants’ activities in one place as a continuous list (see Figure 2). This meant that participants’ activities appeared as a detailed list of entries on their screens, which overwhelmed some participants at first sight. The design also encouraged participants to move quickly from one entry to the next, scrolling down through the log limitlessly. This affected the ways in which participants discussed their activities as Emma’s excerpt demonstrates:

[Looking at her Activity Log] Yeah okay so that was just one of my classmate's birthday today so I just said ‘happy birthday'. That is from the group. here we were discussing where to meet for the group project... Here one of my friends was saying that she is beginning her fourth year for her degree.... ‘Page Name’, that is one of my friend who is doing... well actually he is another person from my internship and I did friend him as well but we were in different departments so I don't know him that well but I just friended him. So he just sent me a request to like his photography page so I did that... err... right that is one of my friend and her boyfriend was saying that they are together and that he is really happy, something like that so I just liked it. (Emma, 22)

Emma listed every entry appearing on her screen, providing only very short explanations and succinct elements of contexts for each before moving on. Justine also felt that the Activity Log design restricted the space and time she had to ask follow up questions and discuss some of the entries in more depth. In this way, the strategy of Facebook encouraging speed, connectivity and newness significantly shaped the possibilities for ‘tactically’ repurposing the Activity Log in her research.

Furthermore, while the data presented on the Activity Log is selected and subjective (see above), participants perceived it as a ‘true’ or ‘systematic' recording. This partly reflects how the ideology of platforms as neutral technical structures (Gillespie, 2010) has pervaded understandings of social media, which in turn impacted on the dynamics of the interviews. For example, embarrassment and unease arose when participants' logs seemed to directly contradict something they had previously said in the interviews (e.g. ‘I don’t use Facebook very much’), and affected how they discussed their engagement with the platform. Aaron, for instance, appeared embarrassed when his Activity Log showed that he had requested and friended someone the day prior to his interview; earlier, he had emphasised that he would very rarely send friend requests. As a result, he distanced himself from his Activity Log by listing succinctly the data appearing on the screen and talking about himself in the third person (which is also the viewpoint of the Activity Log):

[Looking at his Activity Log] Right … ‘Aaron sent James a friend request', ‘Aaron liked a photo' ‘Aaron liked a photo', my brother Dean posted a link to my timeline. Aaron and Carry are now friends', ‘Aaron and Spencer are friends’... yeah I've been meeting a lot of people since I am here. (Aaron, 20)

This not only caused awkwardness as his log contradicted what he had previously said but had a closure effect on the interview. As a researcher, Justine also struggled to negotiate this seeming disparity which created an unintended situation of individual scrutiny, in which she was potentially perceived as ‘checking' on participants and on the veracity of their previous answers.

As the discussion above shows, repurposing Facebook Activity Logs as digital prompts in interviews encouraged discussion and in many ways informed Justine’s research questions. However, it created closure effects and unwelcome situations of scrutiny which affected the data collected, the dynamics of the interviews and worked against her research aims. Using such prompts might also have reinforced normative understandings of technology and data as neutral and legitimate ways of grasping personal experiences – what de Certeau would see as the strategy of the platform. Although these side-effects can never be entirely removed from research, it is important to weigh up the pros and cons of tactically repurposing social media platforms as tools and to reflect on the extent to which research might be grounded in corporate strategy.

Conclusion

As our discussion demonstrates, we ‘made do’ with Facebook and its specific temporalities. We both worked with features designed for purposes other than research and with incomplete data generated and moulded by Facebook. Justine was dependent on the ways in which Facebook recorded, ranked and presented participants’ personal past activities on the Activity Log. Similarly, in attempting to reconstruct versions of events during the riots, Stevie had to rely partly on the (biased) construction of events that had been imposed by Facebook. These experiences both feed into wider questions around the ‘truth’ of accounts of everyday lives or political and social events based on remnants captured and sorted by Facebook. Furthermore, we both encountered practical dilemmas closely connected to the various temporalities embedded in our research and in aspects of Facebook. In Stevie's case, the events she was interested in were unfolding long before she began her research, and were partly recorded by and through Facebook at the time of the riots. These recordings were shaped in multiple ways by the platform’s temporalities (e.g. pages missing, only ‘traces’ of the riots). Both research projects show that the platform defies the easy excavation of itself, and to varying extents, the events and activities it captures. This is in contradiction to the strategic data mining that corporations themselves and third parties like Cambridge Analytica can and do conduct on a grand scale. This is possible due to the bulk of personal information they are able to access and generate.

Moreover, research activities themselves become part of this process. Justine’s repurposing of participants' Activity Logs – an ostensibly visible part of users’ algorithmic identities – as digital prompts reincorporated participants’ recent activities in the present time of the interviews. This fed back into the interview dynamics as well as the data collected. In addition, this created a tension between participants’ recollections and presentations of their activities during the interviews and the activities as recorded by the platform. Meanwhile, the searches Stevie ran while logged into Facebook were automatically recorded by Facebook, and became part of the data Facebook holds about her as a user. In the course of using social media or other digital platforms in our research we may ourselves unwittingly leave ‘digital footprints’ (Hepp et al., 2018) – which in turn affect the construction of our algorithmic identities in ways we might not fully realise (cf. Horning, 2017).

Although our research projects used different aspects of Facebook, the practical challenges that we encountered as well as our tactical approaches bore striking similarities. In this way, de Certeau’s distinction between tactics and strategies is useful to understand the asymmetrical relations of power between social media corporations and researchers. We are dependent on social media platforms as spaces which were not originally designed for research, and that we repurposed in order to generate, collect and access data for research. Social media platforms have also their own temporalities which work actively against these purposes. As social researchers interested in small-scale data as well as Big Data we have to develop tactical approaches to mitigate social media corporations’ strategies – in particular regarding their data-mining design, their specific temporalities, and their ideologies around connectivity, speed and newness. This means creating novel ‘ways of operating’ within these platforms corresponding to the temporalities and aims of our research. As de Certeau stressed, a tactical approach does not ‘imply a return to individuality’ (1984: xi) and combatting at close quarters does not have to be alone.

This article is a step forward to creating a space for dialogue, with the aim to share individual tactics and learn from them collectively so we can better foresee and take advantage of ‘the possibilities of the moment’ to use and repurpose social media platforms in research. Cambridge Analytica has shown that such dialogue between researchers using social media data and/or platforms is more than ever needed, especially if we want to tip the scales in our favour.

Footnotes

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research,authorship,and/or publication of this article.

Funding

The authors received no financial support for the research,authorship,and/or publication of this article.

Notes

References

Adkins L (2009) Sociological Futures: From Clock Time to Event Time. Sociological Research Online 14(4): 8. DOI: 10.5153/sro.1976.

Association of Internet Research (AOIR) (2012) Ethical decision-making and internet research. Available at: http://aoir.org/reports/ethics2.pdf (accessed 8 June 18).

Beer

(2017) The social power of algorithms. Information, Communication & Society 20(1): 1–13.

Bivens R and Haimson O (2016) Baking gender into social media design: How platforms shape categories for users and advertisers. Social Media + Society October–December: 1–12.

Bolter J and Grusin R (1999) Remediation: Understanding New Media. Camridge, MA: MIT Press.

Boyd

(2015) Making sense of teen life: Strategies for capturing ethnographic data in a networked era. In: Hargittai

Sandvig

(eds) Digital Research Confidential: The Secrets of Studying Behavior Online, Cambridge, MA: MIT Press, pp. 79–102.

Boyd

Crawford

(2012) Critical questions for Big Data. Information, Communication & Society 15(5): 662–679.

Brugger

(2015) A brief history of Facebook as media text: The history of an empty platform. First Monday 20(5): . Available at: http://firstmonday.org/ojs/index.php/fm/article/view/5423/4466 (accessed 12 January 2018).

Bruns A (2018) Facebook shuts the gate after the horse has bolted, and hurts real research in the process. Internet Policy Review. Available at: https://policyreview.info/articles/news/facebook-shuts-gate-after-horse-has-bolted-and-hurts-real-research-process/786 (accessed 8 June 2018).

10.

Bucher

(2017) The algorithmic imaginary: Exploring the ordinary affects of Facebook algorithms. Information, Communication & Society 20(1): 30–44.

11.

Bucher

Helmond

(2018) The affordances of social media platforms. In: Burgess

Poell

Marwick

(eds) The SAGE Handbook of Social Media, London: SAGE Publications, pp. 233–253.

12.

Burrows

Savage

(2014) After the crisis? Big Data and the methodological challenges of empirical sociology. Big Data & Society 1(1): 1–6.

13.

Chun

(2008) The enduring ephemeral, or the future is a memory. Critical Inquiry 35(1): 148–171.

14.

Couldry

Hepp

(2016) The Mediated Construction of Reality, Cambridge: Polity Press.

15.

Dean

(2010) Blog Theory: Feedback and Capture in the Circuits of Drive, Cambridge: Polity.

16.

De Certeau

(1984) The Practice of Everyday Life, Berkeley: University of California Press.

17.

Duguay

(2016) “He has a way gayer Facebook than I do”: Investigating sexual identity disclosure and context collapse in a social networking site. New Media & Society 18(6): 891–907.

18.

Floridi

(2012) Big data and their epistemological challenge. Philosophy and Technology 25(4): 435–437.

19.

Fuchs

(2012) Social media, riots, and revolutions. Capital & Class 36(3): 383–391.

20.

Fuchs

(2014) Social Media: A Critical Introduction, London: Sage.

21.

Gayo-Avello D, Metaxas P and Mustafaraj E (2011) Limits of electoral prediction on Twitter. In: Proceedings of the Fifth AAAI International Conference on Weblogs and Social Media, Barcelona, Spain, 17–21 July 2011, pp. 490–493.

22.

Gillespie

(2010) The politics of ‘platforms’. New Media & Society 12(3): 347–364.

23.

Hassan R (2007) Network Time. In: Hassan R and Purser RE (eds) 24/7: Time and Temporality in the Network Society. Stanford, CA: Stanford University Press, pp. 37–61.

24.

Helmond

(2015) The platformization of the web: Making web data platform ready. Social Media + Society 1(2): . DOI: 10.1177/2056305115603080.

25.

Hepp A, Breiter A and Friemel TR (2018) Digital Traces in Context: An Introduction. International Journal of Communication 12(2018): 439–449.

26.

Horning R (2017) Sick of Myself. Real Life, 17 May. Available at: http://reallifemag.com/sick-of-myself/ (accessed 30 January 2018).

27.

Hoskins

(2009) The mediatisation of memory. In: Garde-Hansen

Hoskins

Reading

(eds) Save as: Digital Memories, Basingstoke/New York, NY: Palgrave Macmillan, pp. 27–43.

28.

Karpf

(2012) Social science research methods in internet time. Information, Communication and Society 15(5): 639–661.

29.

Kastrenakes J (2018, January 31) Facebook usage falls in the US as it begins to tinker with the News Feed. The Verge, 31 January. Available at: https://www.theverge.com/2018/1/31/16956818/facebook-q4-2017-earnings-50-million-hours-reduced-usage (accessed 8 June 2018).

30.

Kaun

Stiernstedt

(2014) Facebook time: Technological and institutional affordances for media memories. New Media & Society 16(7): 1154–1168.

31.

Kennedy

Moss

Birchall

et al. (2015) Balancing the potential and problems of digital methods through action research: methodological reflections. Information, Communication and Society 18(2): 172–186.

32.

Kitchin R (2014) Big data, new epistemologies and paradigm shifts. Big Data & Society April–June: 1–12.

33.

Langlois

Elmer

(2013) The research politics of social media platforms. Culture Machine 14: 1–17.

34.

Lievrouw

(2012) The next decade in internet time: Ways ahead for new media studies. Information, Communication & Society 15(5): 616–638.

35.

Lupton

Michael

(2017) ‘Depends on Who’s Got the Data’: Public understandings of personal digital dataveillance. Surveillance & Society 15(2): 254–268.

36.

Madianou

Miller

(2013) Polymedia: Towards a new theory of digital media in interpersonal communication. International Journal of Cultural Studies 16(2): 169–187.

37.

Manovich

(2009) The practice of everyday (media) life: From mass consumption to mass cultural production? Critical Inquiry 35(2): 319–331.

38.

Manovich

(2012) Trending: The promises and challenges of big social data. In: Gold

(ed.) Debates in the Digital Humanities, Minneapolis: University of Minnesota Press.

39.

Paßmann

Boersma

(2017) Unknowing algorithms: On transparency of unopenable black boxes. In: Schafer

Van Es

(eds) The Datafied Society: Studying Culture Through Data, Amsterdam: Amsterdam University Press, pp. 139–146.

40.

Palen L and Liu S (2007) Citizen communications in crisis: anticipating a future of ICT-supported public participation. In: Proceedings of the SIGCHI Conference on Human Factors in Computing Systems, San Jose, CA, USA, 28 April–3 May, pp. 727–736.

41.

Parikka J (2012) What is Media Archaeology? Cambridge: Polity Press.

42.

Pasquale

(2014) The Black Box Society, Cambridge, MA: Harvard University Press.

43.

Rieder B and Rohle T (2012) Digital Methods: Five Challenges. In: Berry D (ed) Understanding Digital Humanities. Basingstoke: Palgrave, pp. 67–84.

44.

Robards

Lincoln

(2017) Uncovering longitudinal life narratives: Scrolling back on Facebook. Qualitative Research 17(6): 715–730.

45.

Skeggs

Yuill

(2015) The methodology of a multi-model project examining how Facebook infrastructures social relations. Information, Communication & Society 19(10): 1356–1372.

46.

Solon O (2018) Facebook says Cambridge Analytica may have gained 37m more users' data. The Guardian, 4 April. Available at: https://www.theguardian.com/technology/2018/apr/04/facebook-cambridge-analytica-user-data-latest-more-than-thought (accessed 8 June 2018).

47.

Srnicek

(2016) Platform Capitalism, Cambridge: Polity Press.

48.

Strate

(2008) Studying media AS media: McLuhan and the media ecology approach. Media Tropes 1: 127–142.

49.

Townsend L and Wallace C (2016) Social media research: A guide to ethics. Available at: http://www.gla.ac.uk/media/media_487729_en.pdf (accessed 8 June 2018).

50.

Van Dijck

(2013) The Culture of Connectivity: A Critical History of Social Media, Oxford: Oxford University Press.

51.

Van Dijck

(2014) Datafication, dataism and dataveillance: Big Data between scientific paradigm and ideology. Surveillance & Society 12(2): 197–208.

52.

Van Es K (2017) Liveness redux: on media and their claim to be live. Media, Culture and Society 39(8): 1245–1256.

53.

Weltevrede

Helmond

Gerlitz

(2014) The politics of real-time: A device perspective on social media platforms and search engines. Theory, Culture & Society 31(6): 125–150.

54.

Zuckerberg M (2018) Personal Facebook post. Available at: https://www.facebook.com/zuck/posts/10104413015393571 (accessed 1 February 2018).