Difficult data: reflections on making knowledge claims in a turmoil of competing subjectivities, sensibilities and sensitivities

ABSTRACT

Much care has been paid to reflexivity, including considering emotional sensitivities in relation to face-to-face interviews. However, less attention has been paid to the impact of sensitivities on the process of data analysis and the presentation of research findings. This paper focuses on what constitutes ‘the data’ to be analysed, who analyses and interprets the data, what is silenced and what is made public, and what this might mean for feminist research practice in areas considered to be ‘sensitive’. It examines these issues by focusing on research on abortion decision-making in which the views and emotions of the researcher were profoundly challenged by women’s constructions of narratives containing a theme of abortion regret. The paper demonstrates how theoretical perspectives – in this case, feminist standpoint theory – can be used in sensitive research, to allow the researcher to resolve such tensions by providing a means through which greater attention can be paid to the context of participants’ narratives.  

KEYWORDS:

• Reflexivity
• sensitive research
• feminist standpoint
• emotions

Introduction

This research paper adds nuance to understandings of sensitive research through an exploration of the emotion of distress when conducting research on abortion. The distress was experienced by both the participant and the interviewer (myself), but in very different ways. The participant’s distress was visibly shown through sorrow and verbally expressed through guilt. I experienced empathy with her sorrow, but my distress was dominated by anger, an emotion rarely discussed in research literature, and an emotion concealed in the research encounter.

At the heart of this paper is a reflexive exercise about a researcher–participant interaction that occurred during a qualitative interview for a research project on young women and unintended pregnancy and involved a participant – a young mother – discussing her experiences of an abortion, during which she repeatedly talked about ‘killing my baby’. Whilst I was upset at her expressions of distress and guilt, I was also distressed and angry that she should feel this way. Maybe rather naively, I had not anticipated such a virulent reaction to her recent abortion. Upon considerable reflection, I understood that I had not challenged my own assumptions adequately and that I experienced intersecting tensions that related to three identities brought into the research. As a feminist pro-choice activist, I had spent many years campaigning for abortion rights and was also aware that ‘killing my baby’ is a core element of anti-abortion discourse and rhetoric. As an applied social researcher, I was seeking to maintain a professional distance and minimise subjectivity. And, as a feminist researcher committed to grounding knowledge in women’s experiences and voices, the dilemma that I knew would come later was how to interpret and represent these unwelcome data.

I begin by outlining relevant methodological issues: in particular, sensitivity and reflexivity; emotions and empathy; and subjectivities. I then describe the research encounter, and the consequent data analysis dilemma. Finally, methodological implications of these issues are discussed, focusing on the possibility for multiple interpretations of the data, for competing subjectivities and for the potential of feminist standpoint to resolve these dilemmas.

Sensitivity, reflexivity and emotions

Researcher reflexivity is mostly concerned with how researcher–participant interactions unfold, and the implications of this for knowledge generation (Mao & Feldman, 2019). Whilst establishing rapport and encouraging open dialogue is seen as critical in qualitative interviews (Charmaz, 2014; Rapley, 2004), it is widely acknowledged that the researcher does not remain neutral in this process. Qualitative researchers actively shape the whole of the research process, constructing the collection, selection and interpretation of data (Finlay & Gough, 2008; Finlay, 2008). Reflexivity is arguably particularly relevant in areas viewed as sensitive research, for sensitive research may invoke researcher emotions that can affect knowledge generation.

Reflexivity can thus take different forms, one of which Finlay (2008) describes as ‘intersubjective reflection’ which captures the emotional investment that researchers have in the research. This involves recognising qualitative interviews as inter-subjective emotional encounters, as well as allowing researchers to ensure that they act in accordance with those values (Rodríguez-Dorans, 2018).

Feminist research in practice: embracing subjectivities

The ‘myth’ of value-free research has long been challenged by feminist researchers’ values with interpretations being acknowledged as central to the research process (Roberts, 1981). Similarly, feminist research has long advocated reflexivity as an essential practice that may facilitate greater understanding of research encounters as social encounters (Ramazanoglu & Holland, 2002).

It has also been argued that feminist researchers should do more than simply reveal themselves, through reflexive research, but openly reflect on the significance of their own identities for their research, indicating how this may have influenced their behaviour in the field (Ryan-flood & Gill, 2010). This can be extended to data analysis and dissemination, for the researcher is an active interpreter of what is heard and then reported from the interview (Johnson & Rowlands, 2012). All interpretations and voices are thus subject to conflict and dispute (Denzin, 1997).

Feminist research that invokes an open acknowledgement of subjectivity has been subject to considerable scholarship (Ramazanoglu & Holland, 2002). The feminist critique of positivism has involved developing a critical awareness of research processes, with a particular focus on researcher/researched relationships. This entails a challenge to the positivist perception of an objective, neutral observer who leaves the field without influencing the data. This has been characterised as a myth (see for example, Miller et al., 2012; Ribbens & Edwards, 1998). An interesting alternative approach is presented by Rolls and Reif (2006) who argue that if researchers increase their own reflexive capacity this can increase their understanding of the phenomena and thus improve objectivity.

Undertaking applied social research invariably involves encounters with positivism, as evidence-based policy and practice does not easily accommodate competing subjectivities. It is a particularly hazardous activity for feminist academics who wish to remain true to feminist principles whilst simultaneously seeking to influence policy and practice (Gillies & Alldred, 2012). Bringing together feminist political struggle and research has long been a significant strand in feminist work (Henwood et al., 1998), but the various sensitivities involved in doing this have rarely been acknowledged. This paper focuses on political sensitivity and clashing value frameworks.

Standpoint feminism rejects the view of objectivity as ‘value-free’ yet makes the claim that feminist research may be more objective than androcentric traditional research because it produces less distorted knowledge (Harding, 1987). Standpoint feminism further makes the case that, because women’s lives and roles in almost all societies are significantly different from men’s, women hold a different type of knowledge. Harding (1987) argued that knowledge grounded in women’s experience of struggles against male domination can produce a more complete knowledge of gendered social lives than that based only on men’s experiences. Feminist standpoint theory therefore sought to develop new feminist knowledge of gendered social lives through women speaking their truth (Hartsock, 1997).

Feminist standpoint theory does make knowledge claims, albeit partial, and yet political, and thus remains an attractive methodological approach for feminist political activists who are also engaged in applied social research. This was the methodology adopted in the research project I am reflecting upon now. In that project the original research team attempted to ground the knowledge produced in women’s experiences, and produce research that would be politically useful, as well as policy-relevant. As I show in the account below, this approach was to prove extremely challenging.

The research encounter

In subjective accounts of fieldwork, qualitative researchers are concerned to explore how their biographies interact with their interpretations (Finlay, 2008), and this is what I will do here. The relevant biographical information concerns my political activism from the 1980s onwards, to defend and improve abortion policy and provision in response to anti-choice campaigns that had sought to limit abortion. An important self-motivation for undertaking research on abortion was to understand what could help improve women’s experience of abortion in the UK.

The research encounter presented here occurred during a research project on teenage pregnancy (Hoggart, 2012; Hoggart and Phillips 2011). In one of the early interviews, I interviewed Cara (pseudonym). Cara had experienced what she described as three unintended pregnancies: the first at 16 ended in miscarriage. The second time she was pregnant, 2 months after her miscarriage, she had considered the option of having an abortion but decided to proceed with the pregnancy, a decision she explained by drawing on the concept of responsibility: ‘in my mind I was like, if I’m a big enough girl to lie down with a man, then I’m big enough to take the responsibility’. When I interviewed Cara, 3 years after becoming a mother, she had recently had an abortion. When we talked about her decision-making for her most recent pregnancy, her narrative, whilst making it clear that it had been a difficult decision, was initially focused on her own social and economic circumstances, incorporating those of her daughter and current partner (not her daughter’s father):

I’m still going college, he’s working, there’s not really enough space for a baby, that’s the only reason why I had the abortion, because it wasn’t right and I didn’t want to bring another baby in the world not working and not being settled, not knowing what I was going to do with my own life, I thought it would be unfair

This complex mixture of influencing factors (considerations of child(ren), partner, own education, living conditions, own future life) is a common theme in abortion research (Rowlands, 2008). What is not so common, but this may be because it is not reported, is what Cara told me a little later in the interview when she was talking about her abortion experience:

I’m not ever going to do it again, it’s a horrible experience, like gonna be hanging over my head for a long, long time, that I killed a baby, the baby didn’t ask to be here but the baby’s here, I should have kept the baby and then I’m thinking I couldn’t think like that because I’m just going to bring myself down and I’m going to hurt even more, I’m gonna regret it, I’m regretting it and I just lost my mind for a while.

Within this quote there is a narrative theme of regret, a theme that is central to anti-abortion propaganda; and pro-choice political activists are wary of the concept of post-abortion regret. I was simultaneously empathising with and upset for Cara, but also angry and upset that she could feel this way; as well as immediately dismayed by the dilemma I knew the research team on this project would face: what would we do with these data? It was a theme that was repeated, though not so intensely, in a few other interviews.

My subjective response of shock and anger was repressed. Two people in any kind of conversation affect each other, in ways that may not be fully understood (Bourne and Robson 2015; Bondi, 2014). My practice had always been to seek to minimise the extent to which my own views may affect participants’ responses. Maybe this was why I had also failed to prepare for the possibility of such a clash of subjectivities through a critical consideration of the potential for conflicting views, a consideration that should have been a key element of feminist research practice. Unfortunately, I was not only a feminist researcher in this project. Indeed, as an applied social researcher under pressure to conform to what is seen as professional practice (Gillies & Alldred, 2012), I had striven to appear neutral as was my usual practice and did not disclose my pro-choice sympathies or activism in any overt way. It is possible that my efforts to appear neutral and non-judgemental and not disclose my own values on abortion enabled Cara to be honest, but equally she may have felt that this was an expected response. The recruitment process may also have played a role, as participants had been recruited (in the first instance) by abortion providers using materials that simply stated that we wanted to learn about their experiences and hear their voices, in order to improve services in the future. Whatever uncertainties remain about factors influencing Cara’s narrative, and it is important to acknowledge there are invariably such uncertainties, what is clear is that the data were unwelcome to the researchers.

Researchers have a role as interpreters and co-constructors (with the research participants) of knowledge (Hesse-Biber & Leavy, 2007). Even whilst accepting that researchers construct research narratives (Gill, 1998), and that multiple readings of the same data are entirely possible (Seu, 1998), the dilemma remained: how to represent the data? We had given participants the opportunity to relate their experiences and express their views, but how should we validate those views? Whilst researchers may acknowledge subjectivity and the role of interpretation what does this mean, and how far do you take it? To draw on interpretive ‘freedom’ and ignore what was a significant theme would surely be an abuse of interpretive privilege only made possible by the power position of the researcher.

Discussion

For this particular project, this research moment encapsulated a theme that appeared to challenge an assumption of our feminist standpoint methodology: that grounding knowledge in women’s abortion experiences could produce research that may be politically useful, in this case for pro-choice activism. This research might instead be politically useful to anti-abortion activists.

An immediate issue concerned interpretation and representation in the research report. This is not an uncommon issue for feminists undertaking applied social research, where policy-makers are expecting ‘objective’ evidence. (Willot, 1998). As reflexivity acknowledges viewing reality as a matter of competing interpretations, it is inevitably problematic for applied social researchers because the subjective knowledge may be seen to undermine the value of social research. This is confounded by the overriding tension between being a researcher and being a feminist. As a feminist, you want to listen and value women’s voices, but what if – as shown in this research – you fundamentally disagree with the direction of travel that those voices may be pushing the research. Additionally, as a feminist activist, it is not really enough to give voice; you would seek that voice to guide your political action. The issue that we faced as a research team was how to represent this participant’s experience and views in a situation in which the narrative could be co-opted by anti-choice activists.

Ultimately, we sought to retain our initial objective of staying close to the ‘reality’ of our participants and therefore began to explore the emergent theme of immorality (‘killing my baby’) in our analysis (Hoggart, 2012). We did not want to repress Cara’s views and feelings – her subjectivity – in favour of our own, and we were also aware that though this was not a dominant theme, neither was it an outlier. Our analysis recognised the importance of moral decision-making for the research participants, but we also acknowledged that this focus can be problematic for some women who do not feel comfortable about their abortion, as indeed it had been for Cara. This led to recommendations around challenging abortion stigma (a sense of abortion as morally wrong), and reframing abortion as a sexual health issue in Sex and Relationships Education (SRE) rather than an opportunity to discuss moral dilemmas in Religious Education (RE) (Hoggart, Phillips, Birch, Koffman, 2010)².

Some time later, following the completion of the project, and further reflection which was enabled by a shift in career away from working for an applied research institute, I was able to return to the data to explore this clash of emotions further. I had retained an anxiety that we had privileged our ‘truth’, and not really engaged with the emotions of the research participants. The value of analysing emotions in research has been increasingly acknowledged: ‘When strong emotional reactions emerge during the research process, these can be used to understand the researched phenomenon’ (Rodríguez-Dorans, 2018, p. 748). I was curious to explore how this might be done utilising feminist standpoint methodology.

Historically, a fundamental critique of the Standpoint claim that feminist research may be more objective than androcentric traditional research because it produces less distorted knowledge (Harding, 1987), was that this is misguided; and that feminists should privilege subjectivity over objectivity; emotionality over rationality; and experience over experiments (Stanley & Wise, 1983, 1993).¹ Feminist standpoint theory continues to challenge positivist claims to neutrality and objectivity in research, whilst continuing to privilege the idea of research subjects as ‘knowers’ (Intemann, 2010). It has, however, become more grounded. Whilst arguing that research involving marginalised groups should begin with the experiences of marginalised groups (Harding, 2008), feminist standpoint also increasingly foregrounds the influence of social and cultural contexts (Harding, 2004; Intemann, 2010). There is thus a close connection to understanding that qualitative research is embedded in complex historical and cultural contexts (Denzin, 2010), and that narratives are socially and historically situated (Jackson, 1998; Squire, 1998). Finally, feminist standpoint acknowledges that politics are intimately involved in research (Harding, 2004). Indeed, Hartsock (2004) argued that because feminists are involved in political activity, it is in their interests to strive towards truth in their research projects, as flawed research may precipitate ineffective political activity.

These insights have been crucial to my own work on abortion over the last few years. I have consciously developed an analytical approach based on an exploration of abortion-related stigma that is socially located and, crucially, includes analysis of the negative emotions – including distress and regret – that might accompany abortion experiences. Examples include theorising the ways in which young women’s moral anxieties about abortion may negatively affect their pregnancy decision-making (Hoggart, 2012); and arguing that academic research on abortion should engage with the issue of emotions, particularly post-abortion emotions (Hoggart, 2018). This shift in direction would most likely not have been achieved without dwelling on that emotional moment and engaging with feminist standpoint theory in a curious and flexible manner. Continuing with the reflexivity at the heart of this paper, however, leads me to think again about whose truth is being represented. For whilst Cara’s voice has been represented, and that the reflexive approach has affected the knowledge created, it is difficult to deny that the interpretation remains that of the researcher.

Disclosure statement

No potential conflict of interest, beyond that discussed in the paper, was reported by the author.

Additional information

Notes on contributors

Lesley Hoggart

Lesley Hoggart is Professor of Social Policy Research and Research Director of the School of Health Wellbeing and Social Care at the Open University. Her research projects are focused on reproductive health, abortion policy and politics, and sexual health, and she has published widely in these areas. Her recent outputs include the award-winning MyBodyMyLife abortion story-telling exhibition.

Notes

1. Feminist standpoint has also been criticised for its reification of a single, universal feminist standpoint, which allows for the continued marginalisation of a range of perspectives (including for example, black, lesbian, post-colonial, or working-class perspectives) (hooks, 1990; Stanley & Wise, 1993). Recently, such different perspectives have been brought together through the development of intersectionality theory which recognises the multiple intersections in people’s lives, including race, gender, skin tone, accent, education level, migration status, language and other life situations (Crenshaw, 2001), and feminist standpoint too has developed. Patricia Hill Collins first introduced the idea of Black feminist epistemology to standpoint feminism, arguing that it derives from the personal experience of Black women dealing with both racism and sexism. Collins (1990) went on to utilize the concept of intersectionality originally developed by Crenshaw (2001) to theorise simultaneous overlapping of multiple forms of oppression, such as race, class, gender, sexuality and nation as intersecting, mutually constructing systems of power.

2. This report repeated the title from the research commissioned. Since publishing the report we have understood that the term ‘repeat abortion’ is stigmatising of people who have more than one abortion, and we regret the use of it.