Facebook recruitment: understanding research relations Prior to data collection

ABSTRACT

This article considers the multiple relations that emerge from and between Facebook commenters, as well as between commenters, researchers, and the research project during recruitment. To do so, we draw on our experiences of recruiting individuals who have concerns about or are opposed to a range of recent social and legal changes in ‘post-equality’ contexts. Understanding research as co-created rather than ‘collecting data from’ participants, we consider the researcher, commenters, and Facebook technologies as active agents, and ask how the emergent relationalities between these agents shapes the social media recruitment process. We develop thinking regarding these relationalities through an in-depth exploration of our processes that reveal key methodological considerations relevant to social media recruitment in the social sciences. As the process of recruitment is mutually constructed online through multiple relationalities across researcher/project and commenter, as well as between commenters themselves, we conclude that there is a need for dynamic, iterative, and reflexive responses and engagements rather than pre-defined frameworks.

KEYWORDS:

• Social media research methods
• online research methods
• sexuality
• gender
• LGBTI
• heteroactivism
• anti-gender

Introduction

Social media recruitment is more common in quantitative research across a range of social and health-related fields of study, and is becoming more common in qualitative social research as well (Brickman Bhutta, 2012; Cochrane et al., 2022; Forgasz et al., 2018; Iannelli et al., 2020; Kapp et al., 2013; Ramo et al., 2014; Waling et al., 2022).¹ In particular, social media is increasingly drawn upon to recruit participants for sensitive and controversial social research, including so called ‘hard-to-reach’ participants and groups (Fileborn, 2016; Gelinas et al., 2017; Jackson, 2021). Though social media has been used in sensitive social research recruitment prior to the COVID-19 pandemic, COVID-19 lockdown and social distancing restrictions further propelled social media recruitment methods amongst social researchers (Hickman Dunne et al., 2022, p. 2). To date, social researchers have considered the effectiveness of social media advertising for recruitment in educational research (Forgasz et al., 2018) and considerations for moderating negative comments and ‘trolling’ in advert comments (Waling et al., 2022).¹ In this article, we focus on the multiple relations that emerge through one aspect of social media advertising via Facebook. We draw on our experience running recruitment advertisements on Facebook for the ‘Beyond Opposition’ research project, that explored polarising social and legislative changes (www.beyondopposition.org). Whilst this project is specific, there are broader insights to be gleaned from our negotiations.

‘Beyond Opposition’ explores social divisions relating to gender, abortion, and sexuality, including topics such as same sex marriage and gender recognition. Just as feminist scholars have long conceptualised research as co-created during the data-collection stage (Rose, 1995), this article considers how the recruitment process is similarly co-created through multiple relationalities on Facebook. This includes relations that emerge between the research team and the Facebook user, as well as between Facebook users where they interact with each other on the recruitment advertisement. We further consider Facebook as a technology that operationalises said relations, including moderating, blocking, limiting, and facilitating interactions beyond the ‘control’ of the researcher.

We begin by placing our considerations of social media recruitment within the tradition of feminist understandings of the relationalities of research. We explore the iterative nature of developing our social media engagement policies and practices for the ‘Beyond Opposition’ recruitment process, that was informed by relational engagements between researchers and commenters, as well as between members of the research team. Drawing on examples from the project’s social media recruitment on Facebook, we then turn to examine the multiple relationalities that emerged during social media recruitment. We do this by considering three multi-directional relationalities that emerged through social media recruitment: first, we discuss how researchers collectively reflected on comments and iteratively developed responses. These responses required emotional labour that researchers undertook in navigating intense interactions. Second, we examine the relationality of Facebook users who interacted with the research team, as they ‘wrote back’ to the researchers. Lastly, we examine multiple relationalities that emerged between Facebook users as they interacted with each other. Together, these three relationalities offer insights into Facebook recruitment processes as co-constituted through multiple relationalities, that are co-created with non-human technologies.

Relationalities in research

We set this research within feminist traditions that have long recognised the importance of relations between researcher and researched in the creation of research data and findings (Collins, 1990/2000; Haraway, 1988; Nash, 2010; Rose, 1995; Stanley & Wise, 1993; Vanderbeck, 2005). Research practices and processes are negotiated, creating a ‘betweenness’ that shapes research spaces (Rose, 1993, 1995). Seeing research as relational – that is, created through interaction and negotiation – extends across the lifetime of a research project. These interactions and negotiations start at the recruitment phase, but the focus is often on research processes during data collection itself (e.g. during interviews, fieldwork).

Focusing on the multi-directional relationalities that emerge during the recruitment phase enables considerations of power relations and the complexities of these in the research process. Scholarship concerning relations and interactions between participants and researchers initially questioned the power of the researcher in the research process, asking for reflexive considerations and open discussions/mitigations of the power-laden creation of knowledge. Whilst these power relations remain relevant, assumptions about hierarchical power relations are also called into question, and power within research processes is aptly considered as circulating and dispersed (Bondi, 1990; Browne & Nash, 2010; Gurr & Kelly, 2019). For example, research with elite groups and individuals has shown that participants are not always ‘subject to’ the research, and can instead occupy social, economic, and other forms of privilege that are important to acknowledge and explore in the co-creation of research between researcher-participant (e.g. see Glass & Cook, 2020; Puwar, 2004).

Recruitment has often been considered as researchers ‘reaching out’ to ‘access’ participants in various ways. Yet this ‘reaching out’ contests assumptions regarding the unidirectional nature of recruitment. The ways in which recruitment is undertaken includes using gatekeepers and working through social networks, deploying specific forms of social/institutional relationalities to engage participants. It is therefore not unusual that these relations move beyond researcher-participant, as instead, potential participants can often engage with each other, discussing the research before and after their participation.

Scholarship on social media recruitment shows how it allows for increased interaction between potential participants, and further establishes ease of access to research findings (Gelinas et al., 2017, p. 11; Hickman Dunne et al., 2022, p. 3). Social media users’ ability to ‘write back’ to researchers and to engage with each other underscores a broadening out of relations at the recruitment stage, presenting new methodological considerations. As well as affording increased opportunities for interaction, the technological mediations of social media serve as important actors in the recruitment process, just as technological mediations are important actors in the relationalities of social life more broadly (Haraway, 1988; Law, 1999; Nash & Gorman-Murray, 2019). In this paper, we develop this to consider how Facebook is a technological mediation that shapes the recruitment process, alongside human actors (researchers and Facebook users).

Navigating the multiple relations of recruitment can require emotional labour, where researchers manage or suppress their emotions to meet the needs of the participant and/or project (Hochschild, 2012, p. 7; Carroll, 2013). The experiences of researchers who suffer trauma as part of the research process has received some attention (Drozdzewski and Dominey-Howes, 2015), and the shifting positionalities that surface during these experiences have been extensively debated (Vanderbeck, 2005; Bondi, 1990). Yet considerations of researchers’ emotional labour is a key and often overlooked facet of recruitment, particularly in relation to Facebook advertising and the relationalities that emerge.² These discussions are important considerations of recruitment and have specific dimensions within Facebook recruitment, particularly in projects related to ‘sensitive topics’. Following an exploration of the iterative nature of creating our social media engagement policies, we turn to explore researchers’ emotional labour as a key facet of researcher-commenter relationalities within the Facebook advertisement recruitment process.

Defining the parameters: ‘Beyond Opposition’s’ recruitment process and policies

The ‘Beyond Opposition’ research project initially (2020–2022) explored experiences in everyday spaces for those who have concerns about, or are opposed to, recent socio-legal changes relating to gender, sexualities, and/or abortion. The project focused on these people because they are overlooked and understudied, presumed to ‘disappear’ in the ‘era of equalities’. The ‘Beyond Opposition’ project is thus exploring their experiences respectfully (see Browne & Nash, 2023) in order to develop research that asks the question: ‘if not everyone’s “hearts and minds” can be changed, how do we engage with different views/opinions/positions on socio-legal changes in sexualities and genders in the 21st century?’ As the focus of this paper is on Facebook recruitment, a full discussion of the research is outside the scope of this paper, however further details can be found at www.beyondopposition.org. In this section, we introduce our strategies for recruitment via advertising on Facebook, and how we operationalised this.

The ‘Beyond Opposition’ study is a transnational study, across three countries: Canada, the Republic of Ireland (ROI), and Great Britain (GB).³ Recruitment was identified from the outset as problematic, as participants could be very nervous of speaking about their positions in these ‘post-equality’ contexts, particularly because of the positionings of project researchers, including Kath Browne (Project Investigator) and Catherine Jean Nash (Canadian collaborator) and their writings that sought to support LGBTQ equalities and lives within queer and sexualities geographies.⁴ Various recruitment mechanisms were tried both offline and online. Notably, the majority of recruitment took place remotely during the COVID-19 pandemic, where strict lockdowns and university rules meant that no face-to-face data collection was allowed.

By 2021, it became clear that Facebook recruitment conducted via advertising was the core means of engaging participants. The postdoctoral researchers ran initial advertisements on several social media websites including Twitter, Reddit, and Facebook. These platforms had similar user functions, including the ability to comment, react, and report the advertisement, and to interact with other users through comment threads. Facebook, Reddit, and Twitter also allowed for targeting specific audiences based on interests, geographic locations, and demographics.⁵ However, during recruitment for the ‘Beyond Opposition’ project, Twitter and Reddit had limited engagement, despite using similar paid advertising.⁶ As a result, no participants were recruited under these fora. For Facebook, advertisements yielded extensive engagements and participant involvement in both interviews and questionnaires. This engagement drew on and developed from comments, including debates and discussions across the 80 advertisements. At the time of writing (August 2022), 80 Facebook adverts were run by four postdoctoral researchers across three countries: Canada (46 advertisements), GB (12 advertisements), and the ROI (22 advertisements), with each advertisement running between 3 and 14 days. Through advertisements, we engaged with participants who ranged in age, sexuality, gender, and employment status.⁷ In Canada, there was a higher number of retired participants over the age of 60; this is an aspect that future researchers interested in social media recruitment via Facebook may consider, with regards to the age demographics of Facebook users. As well as age, we found there was less diversity in race, ethnicity, and religion in participants recruited from our Facebook advertisements, and this is a consideration for using this platform.

In order to create advertisements, researchers designed country/region specific images with text that introduced questions relating to the research. Careful consideration was paid to imagery, wording, and regional events taking place. Often, advertisements drew on space-based imagery (for example a bus, a parade, or a flag), that would quickly orient the Facebook user towards the focus of the research. The embedded text was typically a short question offering a sense of the research (e.g. how recent legislative changes shaped their everyday lives, or asking if they could speak to friends/family or co-workers about recent changes). The researchers also adapted advertisements in response to global or regional political, legal, and social events occurring alongside the project, utilizing what Forgasz et al. (2018, p. 267) recognize as the potential of social media to adapt to change. Once a Facebook advertisement was posted, researchers read comments (both from potential participants and non-participants), and decided if each comment could be replied to. For example, researchers replied to comments interested in the research by directing the commenters to the website for potential participation, or answered commenters’ questions about the research where appropriate. Researchers also reviewed comments to see if they violated the ethos of the project. Researchers regularly ‘checked in’ with an advertisement while it was running, and replied to comments within our agreed principles for social media recruitment (see Table 1).

Table 1: Key Principles of social media recruitment for the ‘Beyond Opposition’ project

Key Principles
1. The project always operates within its ethos
2. Abusive comments are removed regardless of the position they take on the issues
3. Responses are professional and courteous and where appropriate, seek to involve people in the research
4. Responses restate the aims of the research and its purpose
5. At times there is no benefit to be gained by responding
6. Arguments are not engaged in
7. Where conversations are developing, these are only intervened in if necessary
8. If a comment presents a threat of violence (general or specific) it needs to be reported

Scholars have noted that as part of online recruitment, researchers often develop social media strategies following challenging encounters (Fileborn, 2016; Waling et al., 2022, pp. 158–159). It has been suggested that having a strategy in place in advance could aid in avoiding adverse events. Defining a set strategy from the outset presumes that research relationalities can be known beforehand, yet discussions of the relational creation of feminist research shows that relationalities cannot be (fully) anticipated in advance. From 2020–2022, the ‘Beyond Opposition’ research team iteratively developed shared engagement principles for recruitment via Facebook advertising. Our approach to the comments began with the first GB/Ireland and Canadian postdoctoral researchers and the PI (Kath Browne), who collectively engaged in reflective discussions around individual comments, namely how they would be dealt with, what was acceptable, what was within our ethos and how this fed into our recruitment.⁸ The discussions developed into a set of ‘key principles’ (see Table 1) that were iteratively considered and applied as the team engaged with commenters over time. This lack of fixity was critical to respond to new situations as they emerged on advertisements.

Alongside the principles, the first GB-Irish and Canadian postdoctoral researchers began to develop responses to individual comments. Katie Young learned from these when she joined the project as the second Canadian postdoctoral researcher, through conversations with the first GB-Irish postdoctoral researcher and from Kath Browne. Katie continued to work with the current team (including the second GB-Irish postdoctoral researcher as they joined the research) to develop shared social media responses to common comments. The team collectively discussed and agreed-upon the best approach to handling new comments as they arose in different regional contexts. This process allowed for consistency and efficiency but also flexibility and openness key for the relationalities that are always emerging and that cannot be (fully) anticipated.

Relationalities in Facebook recruitment

Scholars (particularly those who research social justice issues) have offered insights into navigating social media recruitment spaces. For example, researchers have examined ways of ‘protecting’ potential participants from distressing comments, opting to censor or remove commenters who disagree with potential participants in pursuit of a safe research environment (Hickman Dunne et al., 2022, pp. 2–3, 7). Andrea Waling et al. (2022, pp. 161–162) experienced negative comments when recruiting older LGBTI people via Facebook adverts, opting to moderate comments that might be perceived as prejudiced, balancing free speech with ‘the need to provide a safe space for participants’. When Vera-Gray’s (2017) online recruitment for a project on men’s stranger intrusions received negative comments from a men’s rights group, she left comments public but chose not to approve a negative comment directed at a person defending the project, as the comment stood to cause emotional distress to the potential participant.

In the ‘Beyond Opposition’ Facebook recruitment process, comments were left on advertisements from all sides of the discussion, including those who disagreed with the project and who opposed those who might participate in the project. Given the diversity of social changes explored in the ‘Beyond Opposition’ project, including gender, sexuality, and abortion, commenters interested in participating in the project could disagree or oppose the various positions of other potential-participants. The diversity of positions in the Facebook recruitment space necessitated a nuanced approach to navigating multiple relationalities in the comment section.

A key initial principle for Beyond Opposition was that the project would operate within its own ethos (see ethos webpage https://beyondopposition.org/about-the-project/#AboutTheEthos; Browne & Nash, 2023 for a full discussion). This meant that we would engage with people in respectful ways, and listen to experiences seeking to understand. This was perhaps best evidenced by leaving up comments on advertisements that dealt with all perspectives around the topics of abortion, gender identities, and sexualities. There were limits to this, as we removed comments that presented a threat of violence either general or specific (regardless of the position that a commenter took on an issue),⁹ and we hid any comment where profanity was used which was in line with Facebook’s policies.¹⁰

Such an approach sees that there are multiple relationalities to be considered. Below, we focus on three relationalities that emerged during recruitment via Facebook advertisements: 1. the researcher’s role in navigating commenters’ responses on Facebook recruitment advertisements; 2. commenters’ engagement with researchers during social media recruitment; and 3. how commenters related to each other through the advertisement. Throughout, we further examine the role of Facebook as a technology in further shaping these three relationalities. In combination, we offer insights into the relational constitution of research through Facebook recruitment.

Researcher-commenters

For researchers in the ‘Beyond Opposition’ project, undertaking recruitment and responding to comments was a fluid and contextual process. Replying to comments served a dual purpose of both addressing the questions or concerns of the initial commenter, and speaking to those who read comments but did not comment themselves.

Responses to individual comments were at times very different, even where the comments were similar. For example, in Katie’s first recruitment advertisement, she received this comment: ‘If this is anti-gay rhethoric – get the F off facebook. You don’t belong here!’ Katie consulted with the research team regarding this comment, and together they collectively discussed an appropriate response. They decided that moving forward, this type of comment could be addressed directly while also outlining the aims of the project. In this instance, Katie responded by addressing their concern:

Beyond Opposition is not anti-gay at all; it is an academic research project looking at how we can tackle social polarisation around sex/gender, sexuality, and abortion. There is more information about the project and our ethos https://beyondopposition.org/about-the-project/#AboutTheEthos, and you can find out more about who we are and what our research interests are on https://beyondopposition.org/about-the-project/#AboutTheEthos page.

On a different advertisement, a commenter wrote that the researchers ‘should be ashamed’ for ‘spreading hatred and bigotry’. While appearing similar to the above comment, what constitutes spreading ‘hatred’ or ‘bigotry’ differs greatly depending on the context of the individual and how they define these terms (see Nash & Browne, 2020). It was important to consider how (potential) participants might read this comment, and we were careful not to draw a correlation between ‘hate’ or ‘bigotry’ with the experiences of potential questionnaire or interview participants. After discussing with the research team, Katie reflected on this nuance, and opted to respond to this comment differently. Katie directed the commenter to the aims and ethos of the project outlined on the website:

Beyond Opposition is an academic research project looking at how we can tackle social polarisation around sex/gender, sexuality, and abortion. There is more information about the project and our ethos https://beyondopposition.org/about-the-project/#AboutTheEthos.

The above example reflects the careful navigation and ongoing group reflection required to address diverse comments when recruiting via Facebook advertisements.¹¹ Even somewhat similar comments required different engagements, indicating that pre-determined responses are not appropriate, as responses cannot be defined in advance. Moreover, learning across the research team meant that responses can be changed even to the same comment by virtue of discussion and experience, rather than presuming pre-existing knowledge in a new area of investigation. Our responses reflected ongoing learning that developed through the project, including through participants in interviews, as well as through reflection relating to comments during recruitment. Such reflection may have evidenced to potential participants that the research project developed through learning from involvement of participants over time. Understanding research as relationally produced allows for flexibility, nuance, and ongoing learning, without negating the practicalities of shared principles developed by researchers.

Responding to comments is not a neutral activity and it requires ‘emotion management’ from the individual researcher, a coordination of ‘mind and feeling’ (Hochschild, 2012, p. 7). For Katie, this included suppressing emotions that arose during distressing encounters in service of the aims of the research (Carroll, 2013, p. 549). Katie found that replying to emotionally charged comments through the team’s predefined response often flattened the emotional intensity felt when responding. While many researcher-commenter interactions were limited to a single comment and response in close succession, Katie found that some interactions with commenters spanned across several days, weeks, or even months in rare instances. Commenters returned to contribute new thoughts or raise additional concerns to the researcher either on the same advertisement or on the project’s main Facebook page.¹² Researchers decided in these instances the purpose of ongoing dialogue with a commenter. We considered if the commenter was addressing new issues, if the specific concerns of the commenter were related to the project, or if the dialogue was serving a different purpose.¹³ Regardless of our choice to respond, the ongoing dialogue drew on the researcher’s emotional labour as a resource, at times without directly serving the aims of the project’s recruitment process. Project researchers collectively decided when conversations were no longer in service of recruitment aims and goals. Researchers’ energy, including their emotional labour, was key to this decision.

Facebook recruitment requires individual researchers to work across the multiple relationalities of potential participants on behalf of the project. Each comment requires careful and nuanced consideration, as even comments that appear similar can require deliberation amongst the team, leading to specific replies from the researcher. As well, the at-times charged and distressing nature of Facebook comments required the researcher’s emotional labour during the recruitment process. This labour can be a hidden cost that is at times an intentionally suppressed dimension of Facebook recruitment. Our contention is that it is important to be attuned to this labour, and to reserve the researcher’s emotional labour for recruitment contexts where the interactions with potential participants are purposeful and to recognise the toll of these relations. Although ‘Beyond Opposition’ put in supports as needed (for example this project offered counselling, and postdoctoral researchers developed work patterns that protected themselves), there is an excess to this in terms of emotional labour that cannot be fully accounted for or mitigated against when recruiting via Facebook.

Commenters-researchers

Research relationalities in online recruitment are co-created with commenters. In this section, we explore how (potential) participants engaged with online recruitment advertising, and how researchers responded to this, creating the dialectic relationalities of online recruitment.

Those who commented on Facebook recruitment advertisements often expressed suspicion towards the ‘Beyond Opposition’ project and/or the project researchers. They would express their perceptions of what the aims, goals, and intentions of the project might be, as well as who the researchers might be or who they might be funded by.¹⁴ This is evidenced in the below list of comments expressing perceptions of the research project across multiple advertisements:

‘a malignant authoritarian identity politics cult’; ‘leftist lunatics’; ‘dangerous forced birth propaganda’; ‘whackjob fundamentalists’; ‘a hate group’; ‘Alt-Right, UN/WEF sponsored indoctrination’; ‘a bunch of conservative people who don’t like change and progress’; ‘Gov funded propaganda’; ‘an anti-gay anti-transgender group’; ‘a spy working for [the government] making a list of pro-life people’.

The above list indicates that many commenters’ initial engagements towards the research put ‘them’ (in this case the project) against ‘me/us’ (in this case the Facebook user). Commenters often expressed concern that the project held a position that was in opposition to their own, regardless of a commenter’s expressed position on gender, sexuality, and/or abortion where such positions were expressed. The four postdoctoral researchers addressed these concerns in various ways, including outlining the project aims and ethos, as well as addressing the positionality of researchers where necessary.

As well as commenting directly on the project, commenters found indirect ways to engage with the project. These relational engagements were critical and often seeking to find issues with the project or our use of Facebook. For example, when researchers ‘targeted’ audiences based on interests or by region, commenters could request a rationale from Facebook as to why they have been targeted. This would be used by commenters to ask or critique researchers about their choices around targeting. This can be conceptualised as a relational dialogue where potential participants speak ‘back’ to researchers about their approach to recruitment, drawing on Facebook technologies to do so.

As well as commenting on the research project itself, Facebook users also commented on our approach to moderating comments. For example, some commenters noted discrepancies between the number of comments listed at the top of an advertisement compared to the number of comments viewable on the advertisement overall, suggesting that researchers were hiding comments. In doing so, commenters made clear that they were keeping track of the researcher’s actions, including hiding comments. For example, on a Facebook advertisement asking ‘Are you comfortable with recent policy changes around gender at work?’, one commenter wrote: ‘Considering all the comments are hidden, I’d guess no.’ On another advertisement, a comment had been blocked, and the commenter presumed it had been blocked by the research team, which was not the case.¹⁵ They openly wrote their concern about their comment being blocked on the advertisement, which generated discussion amongst other commenters:

Commenter 1: ‘…you also blocked me from being able to respond to notifications from this ad – but your ad keeps coming back up on my page … ’

Commenter 2: ‘Many block you, if you do not give the desired comment…’

Commenter 3: Hummm interesting

When participants post critically, and at times incorrectly, about the recruitment process and researcher practices, it can ‘undermine the understanding of other participants (and potential participants)’ (Gelinas et al., 2017, p. 11). In this instance, this comment sparked suspicion in others who subsequently commented about their uncertainty regarding the project’s ‘censorship’. Through the technological functions of Facebook that allow commenters and advertisers to block, hide, or report comments, Facebook creates a particular means of interacting that can arouse suspicion. This suspicion can contest research openness between researcher and potential participants. Thus Facebook creates social interactions through the operationalisations of the technology that is a key player in the recruitment relationalities.

As times, commenters questioned the authenticity of researchers’ interactions with commenters, implying that researchers were drawing on the technologies of Facebook to provide automated replies. For example, some commenters expressed their frustration in comments, saying that our responses appeared to be automated, coming from a ‘bot’ rather than a researcher.¹⁶ A more expressive response from the researcher may assure commenters of the ‘realness’ of the researcher’s engagement with their questions or concerns. Yet in a project that does not seek to debate or antagonise, emotional responses needed to be carefully negotiated within the project ethos. While suppressing an individual researcher’s emotional involvement may limit recruitment (especially for those who question the authenticity of a researcher’s reply), researchers’ emotional responses may not represent the project in ways that are in line with its ethos, and thus may not serve the recruitment purposes.

As with face-to-face recruitment, the power relations between researcher-researched in social media recruitment are negotiated. On Facebook, commenters can pause a recruitment advertisement by reporting it, overturning assumed power structures of the researcher-researched dynamic. During our recruitment, it was common for Facebook users to report a project advertisement, meaning that our advertisements would be temporarily taken down. Researchers would subsequently request a review of an advertisement, where the technology would assess reasons for an advertisement being reported. This usually led to the advertisement unpausing, but sometimes resulted in the end of the advertisement. The technological-human negotiations are key here. As has long been recognised from those who see technologies as agentic (Haraway, 1988; Law, 1999), technologies work as actors that not only mediate between human interactions, but also co-create the social in which they take place. On Facebook, recruitment is not only negotiated through interactions between researchers and commenters, but also through these technological tools and the norms/conditions they place on these engagements. Thus, in Facebook recruitment, power is relational and negotiated both within and beyond human interactions. These circulations of power are key in understanding and ‘using’ social media recruitment.

Commenter-commenter

During Facebook recruitment, relationalities emerged between commenters. Thus, recruitment is not fully ‘controlled’ by researchers but instead it is mediated and shaped by discussions between commenters, enabled through the technological tools of Facebook. Exploring these dynamics furthers our understanding of the relationalities that create social media recruitment, making it both feasible and also beyond researcher’s ‘control’. In this section, we show how commenters’ interactions with each other (including discourse, debate, and ‘blocking’ each other) shaped the recruitment process.

As detailed above, ‘Beyond Opposition’ works across a range of perspectives and positionalities. This resulted in recruitment advertisements becoming a point of contact and discussion between people with very different views relating to the research. Advertisements posed key research questions, and the comment section was a shared space for dialogue and debate concerning the key issues around sexuality, gender, and abortion. The discussion and debate that transpired between commenters was moderated, but not controlled by researchers. For example, in an advertisement asking the question: ‘Are you comfortable with recent changes around gender at work?’, commenters engaged with each other around issues of polarisation and opposition:

Commenter 1: Y’all in the comments here are choosing to consider ‘treating others with respect’ as ‘personal attacks against yourselves’…

Commenter 2: Respect is a two-way street. Labeling [sic] anyone who doesn’t believe in modern gender theories as a ‘bigot’ or a ‘hater’ isn’t very respectful.

Here, commenters speak to each other, rather than directly to the researcher. Whilst this is outside the researchers’ design or control, it nonetheless impacts recruitment and perceptions of the research. While we cannot speak to the experiences of commenters, and for ethical reasons did not track commenters through the data collection, the dialogue and debate that transpired between commenters may have encouraged and discouraged participation in the research.

Within our research advertising, commenters drew on Facebook as a technology to moderate each other outside of the ‘control’ of researchers. On project advertisements, users could report or block another user’s comment through functions available on Facebook. If a user reported another user’s comment to Facebook, the reported comment may no longer be viewable to others on the advertisement. The user whose comment was removed would be unable to reply to others on that comment thread moving forward, or see responses on the thread they were previously involved in.¹⁷ For example, in a Canadian advertisement, a comment that was within our ethos of respectful relationalities was reported by another commenter and subsequently blocked by Facebook. Facebook does not directly inform the researcher when commenter-commenter blocking/limiting takes place, and so researchers are not necessarily aware of how commenter-commenter moderations take shape, unless further communication is made. In this case, the original commenter wrote in a separate, new comment on the advertisement that: ‘Well I was just blocked from a civil conversation … I guess this site is Beyond Opposition!’ This user also wrote directly to the research project’s communication page, informing us that they had been blocked. In Facebook recruitment, we as researchers can not necessarily create open dialogue and ongoing communication between commenters. We are also unaware of how commenters use Facebook as a technology to limit each other’s communication on the advertisement, unless we are informed through email communication or via a new comment on an advertisement.

The full impact of commenter-commenter interrelations on the commenter cannot be known to the researchers, just as researchers are unable to know the experiences of those who read but do not engage with comment threads under an advertisement. However, we were able to gain some insights where users expressed their emotional reaction towards others’ comments through reaction emojis (for example a heart, laughing face, or crying face). The ‘tallying’ of these emoji reactions on a comment at times reflected a broader consensus of how other commenters/readers in a comment section felt about a commenter’s view or experience. Whilst it is not possible to assess how emojis are used, and they may be deployed ironically, it is notable that at times a particular comment would have what we read as an overwhelmingly affirmative reaction (‘thumbs up’ for example). At other times a commenter’s view could be undermined or disapproved of by receiving a high number of ‘laughing’ or ‘angry’ faces. The technology also created social interactions, as comments with a high number of emoji reactions from other commenters moved up in relevance on the advertisement’s comment section, and in this way would often become the first comment viewable on an advertisement. Commenters and technologies then curated the order that comments appeared, and thus co-created the recruitment process.

Alongside emoji reactions, this co-creation was also apparent in pictorial forms, including memes and GIFs to communicate with each other (see Crosset et al., 2019, pp. 950–951). GIFs tended to draw on regionally popular media content; for example, in Canadian advertisements, GIFS from the Canadian sitcom Schitt’s Creek were employed by commenters to infer a satirical disdain towards another commenters’ perspective. Schitt’s Creek GIFs were used without accompanying text in contexts where commenters expressed support for social changes relating to same-sex marriage in Canada. As Schitt’s Creek centres around LGBTQ+ relationships, commenters may have purposefully drawn on regional popular culture to imply a perspective beyond textual means. As well as GIFs, commenters shared links to external websites they believed were relevant with other users, especially when a commenter asked about a particular term, social change, or event. Commenters would subsequently engage with each other around external material within the comment thread, resulting in ongoing commenter-commenter dialogue that drew on ideas beyond the questions posed by the research. In this way, commenters (re)shaped and reconstituted the direction and discourse of the recruitment process in new and divergent ways.

As researchers, we sought to maintain the principles outlined in Table 1. However, we had to contend with the multiple relationalities at play in Facebook recruitment, many of which were not within our control. The actions of commenters towards each other, particularly reporting each other, and their ability to share external resources within the comment section, had significant consequences for recruitment and (potential) participants’ perceptions of the research. For example, the active blocking of one user’s comment by another led to the alienation of potential research participants, undermining the perceived authenticity of the project’s aims for some commenters. In this way, Facebook recruitment is co-created through the multiple relations at play, and mediated by the technological tools of Facebook.

Conclusion

Social media recruitment is a contemporary way of engaging participants. As a technological tool, we used Facebook advertising as it allowed us to reach (potential) participants across broad geographic regions, but also because Facebook advertisements allowed researchers and users to engage with each other through the comments. This technological tool mediated and shaped research relations in the recruitment space, mutually constructing the recruitment process through multiple relationalities across researcher/commenter, and between commenters themselves. Drawing on feminist discussions of the relationalities of research (Collins, 1990/2000; Haraway, 1988; Nash, 2010; Rose, 1993, 1995; Stanley & Wise, 1993; Vanderbeck, 2005), this paper showed how Facebook recruitment is mediated through technologies that play a key role in co-constituting participant recruitment.

Commenters can interact with researchers to ask questions about the research, express concerns about the research, provide information (including misinformation) about researchers or the project, and can draw on Facebook technologies to report and potentially block advertisements. Thus, recruitment relationalities are always more than researcher actions; the multiple relationalities of recruitment have implications where researchers seek to deploy their own recruitment principles, which may not align with the actions of potential participants on a recruitment advertisement, or indeed Facebook technologies. Future researchers might consider the multi-directional interactions between researcher/commenter and between commenters as they shape the recruitment process. This includes Facebook as a technology which constituted the recruitment process as an agentic actor.

Whilst the ‘Beyond Opposition’ project’s use of Facebook is specific, there are broader insights to be gleaned from our negotiations applicable to research recruitment via advertising across social media platforms. A key finding was that researchers require dynamic and reflexive responses and flexible engagements towards individual comments. This can be achieved through an iterative process within a research team, or even reflective considerations by an individual researcher. Through our iterative and collective approach to responding to new comments and questions, researchers shared responsibility in new situations during online recruitment. Katie found that sharing the responsibility of how to respond to more complex comments amongst the team eased the emotional labour of running the advertisement. As well, the iterative process used for developing key principles further allowed space to discuss difficult recruitment situations amongst the team. A shared process for responding to Facebook advertisements allows for careful and considered responses developed over time. We offer the key principles of the recruitment strategy outlined in Table 1 as a starting point for other discussions. This includes discerning when is appropriate to respond to a comment (and when it is not) and exploring ways to address unanticipated comments as they arise. This need to be considered in relation to the time-sensitive nature of this recruitment methods, including addressing aspects such as reported advertisements as well as incorrect information about the project by users.

We would also urge researchers to consider support systems and lines of communication to engage with the emotional labour of undertaking Facebook (and more broadly other social media) recruitment. Providing supports such as individual counselling for researchers was beneficial for Katie. Pausing or ending advertisements when interactions between commenters or between commenters/researchers become difficult is also an option, and support (from supervisors and others) might be needed to follow this route, and not to see it as a ‘failure’. We would urge researchers to reserve emotional labour for recruitment contexts where there is a clear purpose being achieved to avoid unnecessary labour. Finally, we hope that future research might seek to name and discuss the emotional labour of researchers in more depth, as a key aspect of undertaking social science research, both as a discussion on its own and also as we have done, as a key consideration of the various elements of research including recruitment.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This paper is part of a project that has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme [Grant agreement No. 817897].

Notes on contributors

Katie Young

Katie Young is Assistant Professor of Cultural Geography in the School of Irish Studies at Concordia University, Canada. Her research explores experiences of everyday life in Ireland, Ghana, and Canada. Her work has focussed on the intersections of everyday space and media, everyday experiences of music, and how people navigate everyday spaces through sound, including listening practices.

Kath Browne

Kath Browne (Prof. Geographies of Sexualities and Genders) research focuses on social justice, inequalities and polarisations. She has explored how LGBTQ lives can be ameliorated, including through Liveable Lives (with Niharika Banerjea). She has explored opposition to sexual and gender equalities through the concept of heteroactivism (with Catherine Jean Nash). Her current ERC project, Beyond Opposition, investigates the experiences of people who are concerned about sexual/gender/abortion socio-legal changes, and how we might engage with each other where we seek different realities/futures.

Notes

1. As is the case in other social media platforms, Facebook encompasses multiple public and semi-private online spaces, ranging from closed-communities to targeted public advertisements (Luders, 2015; Triggs et al., 2019, p. 7). Scholars have examined methodological approaches to these diverse spaces within social media platforms. While our research engaged the Facebook advertisement space, other researchers have: created closed social media groups for recruitment and data collection (Hickman Dunne et al., 2022), joined pre-existing social media groups for data collection (Schaffar & Thabchumpon, 2019), and shared research opportunities via friends-of-friends on social media (Fileborn, 2016). Different spaces offer varying levels of anonymity for users. During our recruitment, Facebook advert comments were only viewable to users who had also been targeted for the advert. While a perceived sense of anonymity can shape online behaviour, including strong, charged, or even aggressive responses (Chui, 2014), it is not possible for us to know each Facebook users’ understanding of their anonymity within the Facebook advertising space. As such, while perceived anonymity may shape charged or aggressive behaviour discussed in relation to emotional labour in this paper, it was beyond the scope of our research to explore this. Future researchers may build on existing literature in social media recruitment by exploring how users’ understandings of relationality, including their perceived anonymity, may impact how they respond to and interact with other participants, commenters, and researchers during social media recruitment.

2. The emotional engagement and experiences of social media users targeted by recruitment advertisements is beyond the scope of this paper.

3. Katie Young joined the project in October 2021 as the Canadian postdoctoral researcher. Prior to this Dr. Laine Halpern Zisman undertook this role to March 2021. In Ireland and GB, a postdoctoral researcher worked mainly on recruitment and data collection in GB from January 2020 until February 2022. From April to August 2022, a postdoctoral researcher (Dr. Carol Ballantine) undertook the recruitment and data collection in Ireland. Reflections and examples presented here are solely from Katie Young’s work, supplemented with the Project Investigator’s (Kath Browne) experiences where relevant. Where it is relevant, we acknowledge the work and contributions of the other postdoctoral researchers.

4. While beyond the scope of this article, Katie Young found that her position played a role in the emotional labour of the social media recruitment process.

5. Project researchers targeted Facebook advertisements based on interests related to the project (for example, interests in relationships, marriage, sexuality, or broader social issues). In Canada, advertisements were also targeted based on geographic regions given the geographic size of the country. The only demographic targeted was age, as participants needed to be above 18 years of age in line with our ethics. Targeting policies on Facebook shifted over time. For example, our targeting strategy was impacted by the ‘Updates to detailed targeting’ Facebook policy outlined on January 19, 2022, available here: www.facebook.com/business/help/458835214668072

6. ‘Engagement’ comprised comments on advertisements, click-through to our project survey, and signing up for interviews.

7. We recruited during the COVID-19 pandemic, which informed potential participants’ employment status.

8. We applied for additional ethical approval to keep Facebook advertisement comments to enable further reflections, including those in this paper. As part of ethical clearance, any identifiable information associated with each comment (such as the profile photos or names of commenters) were removed. As advertisements were public, users were not told directly that their comments might be used in the research.

9. Comments that presented a threat of violence were reported to Facebook in line with Facebook’s policies.

10. During our research, we found that Facebook automatically hid profane comments, without the need for researchers to intervene.

11. Careful consideration is especially important in the context of social media research, where comments are easily saved and shared by commenters.

12. Although advertisements would stop running after a set period, those who comment on an advertisement can access the advertisement after it has finished running.

13. For example, acting as a space for expressing frustration towards the researcher or the project.

14. While we did not provide our positionalities on recruitment advertisements, project researchers’ details were provided via the research website linked in the advertisement. Commenters looked up researchers and at times commented on researchers’ positions and past research on the advertisement’s comment thread.

15. The comment was likely blocked by Facebook after another Facebook user targeted for the advertisement saw and reported the comment, however Facebook does not provide this information to advertisers. As researchers, we are only able to know if this happens when users reach out to us.

16. See below for further explorations of commenters’ responses.

17. This is how Facebook functioned with regards to user comments during the time of recruitment, and this was not within the control of the researchers.