Almost Confessional: Managing Emotions When Research Breaks Your Heart

Social scientists have increasingly shown how qualitative research can be an emotional experience for researchers. Literature on this subject has tended to focus on the emotionally upsetting impact of data collection, often framing this as a form of emotional labour which can be managed by researchers adopting confessional style narratives throughout the research process. But what about the potentially life-affirming impacts of emotions in research? And what happens when confessional style narratives create, rather than dilute, emotional trauma? We use our experiences of conducting qualitative research on two very emotive topics – baby loss and sibling bereavement – to explore the role of emotions in research. We go beyond the predominant focus on doing research to shed light on emotions in the wider research process (from recruitment to impact). We will highlight the dual-edged nature of emotions in research, emphasising some of the more beneficial impacts. Drawing on the Weberian concept of Verstehen which focuses on the importance of understanding, we will also develop a more nuanced form of emotion management in this context. In doing so, we offer an original contribution to methodological discussions in this field, as well as to more conceptual debates on emotional labour.


Introduction
The role of researcher emotions within the research process has, up until recently, been neglected by certain disciplines in the social sciences. In the past, positivist approaches were prioritised in certain fields such as criminology (Scheirs and Nuytiens, 2013). Other disciplines such as psychology and interdisciplinary fields such as death and dying studies have shown an ontological preference for objectivity and neutrality, rather than reflection and emotion (Campbell, 2002;Jewkes, 2011;Visser, 2017). According to Holland (2009: 11), the 'dead hand' of Cartesian dualism (which set reason against emotion) served to keep emotions outside key sociological concerns until the late 1970s. In recent years, increasing emphasis has been placed on the co-productive nature of research and the roles that both researchers and participants play in the production of knowledge (Sinha and Back, 2014;Bell and Pahl, 2018). Researchers are part of the social world under investigation and must recognise that they both effect and are affected by the shared experience of research (Gilbert, 2002;Valentine, 2007). It is perhaps no surprise, therefore, that the role of researcher emotions in the research process has become a more pressing concern in disciplines across the social sciences (Holland, 2009;Watts, 2008).
Existing literature on this issue has identified a range of factors (from the sensitive nature of the subject matter under study through to prior personal experience), as invoking a range of emotions in researchers. These include: frustration, loneliness, sadness, boredom, apprehension, guilt, exhaustion, fear, humour, and repulsion (Dickson-Swift et al., 2009). Some researchers even discuss physical pain and distress experienced during and after fieldwork (emerald and Carpenter, 2015). Unsurprisingly therefore, literature in this area has tended to frame research as a form of 'emotional labour' and has sought to advocate a range of researcher coping techniques -including the adoption of reflexivity and confessional style narratives. While literature does acknowledge that emotions are an important and central part of knowledge production (Holland, 2009;Rager, 2005), the more life-affirming impact of emotions in research remains neglected.
Qualitative approaches are often identified as methodological approaches that present emotional challenges for the researcher (Jewkes, 2011;Pearce, 2010;Sikes and Hall, 2019). More recently, however, emotional concerns have been raised by researchers in the context of other types of research such as secondary data analysis (Jackson et al. 2013). While certain methods are perceived as having a greater emotional impact on the researcher, certain aspects of the research process -namely data collection -tend to be the focus of discussions. Research has begun to acknowledge the emotional nature of data analysis and transcription (Butler et al., 2019). However, although literature has highlighted the need for a dialogic approach to dissemination and impact (Sinha and Back, 2014), the role of researcher emotions in this process remains underexplored. This is despite there being some clear areas of concern (Jackson et al., 2013).
In this article, we will draw on our own experience of conducting research on two very sensitive research topics (baby loss and post-mortem & sibling bereavement) to explore the role of emotions within a wider research context. One of the authors (Kate Reed) is a senior academic who has conducted research on various sensitive research topics relating to reproductive health, while the other author (Laura Towers) is an early career researcher (ECR) whose reflections are based on her PhD research. We will draw on our diverse experiences and career stages to explore the role of researcher emotions at various stages of the research process -from recruitment through to public engagement, impact, and research-led teaching. By considering emotions in this broader research context, this article aims to offer an original contribution to methodological debates in the field.
We begin by detailing background literature and the article's conceptual focus as well as outlining the methods used in the two studies. The main part of the article is concerned with a discussion of three different aspects of the research process: from starting out, to managing the analytical process, through to dissemination and impact. Throughout all three sections, we will highlight the dual-edged nature of emotions in research, emphasising some of the more beneficial impacts on the researcher in the process. We will also examine some of the emotion management strategies advocated as ways of dealing with difficult emotions during research (such as autobiographical and confessional style approaches). We acknowledge the value of these approaches. In fact, we draw on our own reflexive research accounts to substantiate our points throughout the article. There are times, however, when such confessional type approaches may not be appropriate. Drawing on the classical sociological concept of Verstehen, our article concludes therefore by suggesting a more nuanced way of dealing with emotions in social research. The article will champion an approach which recognises the importance of reflexivity but stops short of full emotional disclosure. By offering a more productive framing of emotional 'research work', and developing new ways of dealing with emotional challenges, we seek to contribute to wider sociological debates on emotional labour.

The 'emotive' turn in sociological research
Feminists have, for some time, raised questions around epistemology and the creation of knowledge, arguing for the need to place emotion at the centre of knowledge production (Jaggar, 1989). The sociology of emotions (a sub-field of the discipline which emerged in the 1970s and 1980s) is situated in this context. This body of sociology advocates the need to move beyond the ghost of Cartesian dualism and place emotion at the centre of sociological analysis (Holland, 2009). Sociologists have subsequently sought to explore how emotions are triggered, interpreted, and expressed through an individual's participation in social groups (Hochschild, 1983(Hochschild, , 2009Kemper, 1991), exploring the social conditions behind emotions, and their role in individual, community, and organisational contexts (Pawlowska, 2020). With the emergence of a wider 'affective turn' which has taken place across the humanities and social sciences in recent years (Hardt, 2007: ix), sociologists have become increasingly preoccupied with both researching emotions and exploring the role of emotion in research (Brownlie, 2011;Burkitt, 2012).
While literature on researcher emotions is growing in sociology (and in other disciplines), existing research on this issue has tended to be concerned with certain topics such as emotive subjects (e.g. sexuality, death, or terminal illness), or those involving danger (e.g. illegal behaviours), or on the presence of a political threat or social conflict (Lee and Renzetti, 1990;McCosker et al., 2001). Scholars working on race and ethnicity have, however, also drawn attention to the role of researcher emotions in the research process (Caballero, 2009;Widdance Twine and Warren, 2000). Certain subject areas -such death and dying -have been identified as being particularly sensitive and emotive areas of research for both researcher and participant (Borgstrom and Ellis, 2017;Valentine, 2007). For example, although studies on bereavement consistently show that participation in research can be an 'empowering experience' for bereaved individuals (Moss and Moss, 2012: 464), social researchers continue to express concern about the potential for causing harm when researching this subject (Dyregrov, 2004).
Qualitative research tends to be the focus of discussions on emotions in research (Jewkes, 2011;Pearce, 2010;Sikes and Hall, 2019). Researchers conducting qualitative research may be more emotionally vulnerable due to the traumatic or risk-taking nature of the research and close contact with participants (Bloor et al., 2007;Hubbard et al., 2001). Less attention has been given, perhaps, to those who work with secondary or quantitative data (Moran and Asquith, 2020). However, this form of data collection can also leave researchers feeling traumatised and lacking in agency (Jackson et al., 2013). While qualitative methods have dominated discussions on emotions in research, so too has fieldwork. Less is known about the role of emotions in stages prior to, and proceeding, data collection. Other aspects of the research process, however, can be very emotional. For example, in her study of juvenile prostitution, Melrose (2002) found managing her feelings during analysis particularly challenging, something she did not anticipate. Furthermore, while the role of emotions in data analysis is acknowledged, less is known about emotions in dissemination and impact, although there are some clear areas for consideration. Jackson et al. (2013) argue, for example, that making decisions about what to include in publications and presentations can be emotionally challenging for the researcher due to ethical concerns over giving voice to participants. As we seek to show in this article, there are also a range of emotional issues raised during other aspects of the research process -from recruitment to impact activities -all of which require further attention.

Managing emotions 'productively' in research
Existing literature has tended to categorise emotions experienced during research as a form of emotional labour (Hubbard et al., 2001;Dickson-Swift et al., 2006). Although initially developed by Hochschild (1983) in the context of the commercial services, the concept of emotional labour has long been used to explore the work involved in managing emotions in a range of different professional settings (Bolton and Boyd, 2003;Reed and Ellis, 2020). As several authors have pointed out, professionals often employ different types of emotion management strategies in various settings and according to whether they take place 'frontstage' within the public realm or 'backstage' in the private sphere (Bolton, 2001;Boyle, 2005;Reed and Ellis, 2020). In the context of social research, authors have shown that emotional labour can be difficult and sometimes gendered work, a form of labour that often leads to detrimental physical and mental symptoms (Bloor et al., 2007;Dickson-Swift et al., 2006;Sampson et al., 2008;Watts, 2008). While it is widely recognised that emotions can enhance the researchers understanding of the subject under study (emerald and Carpenter, 2015;Holland, 2009;Rager, 2005), literature continues to focus specifically on the negative impact of emotions on the researcher. Attention is rarely drawn to some of the more life-affirming aspects of emotionally challenging research.
Researcher reflexivity and a range of self-care practises are often recommended by the literature to help researchers navigate emotional challenges in research -from keeping a diary to peer debriefing (Borgstrom and Ellis, 2017;Rager, 2005). Reflexivity and the acknowledgement of researcher positionality have long been central to feminist and qualitative research (Bondi, 2009;Denscombe, 2014). Increasing emphasis has been placed by the literature, however, on researcher vulnerability and more public disclosures of personal experience (Behar, 1996;Valentine, 2007;Visser, 2017). Acknowledging our vulnerabilities as researchers and being open about personal experience throughout the entire research process is perceived to facilitate greater empathy towards participants as well as being therapeutic for the researcher (Behar, 1996). This emphasis has led to a proliferation of autobiographical and confessional style accounts, especially in emotive areas such as death and dying (Borgstrom and Ellis, 2017). These approaches are useful for drawing out unanticipated emotional responses in research. There is concern, however, that the voice of the respondent can be eclipsed through confessional style approaches in favour of that of the researcher, potentially reinforcing, rather than overcoming power relations in the research process (Faria and Mollett, 2016;Finlay, 2002). Furthermore, while such approaches can serve to heighten a researcher's emotional awareness, they do not always provide a proper outlet for researcher emotionality (Borgstrom and Ellis, 2021).
This article seeks to contribute to and extend this existing literature on emotional labour in two respects. First, it aims to show that while articulating and managing emotions in research is undoubtedly hard work, it can also be a productive and life-affirming experience for researchers (Reed and Ellis, 2020;Wouters, 1989). It often acts as a reminder as to why we do social science research in the first place, something that could be better accounted for in literature on emotional labour in research. Second, while exploring and highlighting the value of existing reflexive and self-care practises, the article also concludes by offering a more nuanced approach to emotion management. We will draw on the sociological concept of Verstehen, first advanced by Dilthey and then Weber (Brewer, 2011;Sumner, 2011), to suggest a more nuanced approach. Verstehen sociology emphasises the necessity of understanding the meaning of human action (Sumner, 2011;Turco and Zuckerman, 2017). It is not, however, about researchers plunging headfirst into the experiences or others, nor is it about researchers revealing their own personal information to research participants or to the wider public. Rather, it is about searching and analysing (through words, images, behaviours, and institutions) the way 'people actually represented themselves to themselves and to one another' (Geertz, 1984: 126).
Using the concept of Verstehen, we argue, could help us to develop an approach to research that is almost confessional. This approach would focus on achieving maximum 'understandability' of participant experience (Turco andZuckerman, 2017: 1280) while stopping short of full emotional disclosure by the researcher. It recognises the importance of situated knowledges and the need to reflect on researcher positionality (Haraway, 1988). Rather than drawing attention to the ways in which social identity informs knowledge production, however, our verstehen approach seeks to tread the boundaries between understanding and emotion. Developing such an approach, we argue, could be beneficial in two respects. It may preserve the centrality of participants' voices while also helping to protect some researchers from further emotional discomfort. This could, we argue, also facilitate the development of new frameworks for research that are both 'rigorous yet not disinterested' (Behar, 1996: 175) and which have the potential to transcend the boundary between emotion and rationality.

Overview of the research projects
This article is based on the authors' experiences of conducting two sensitive sociological research projects. Study A (2015-2018) conducted by Kate focused on the exploring the role of magnetic resonance imaging in perinatal post-mortem. Study B conducted by Laura focused on investigating people's experiences of sibling bereavement (Towers, 2019). We acknowledge that these are particularly emotive projects. The experiences garnered through researching them are, however, more commonplace than is often assumed.

Study A: exploring experiences of baby loss and post-mortem
Kate has conducted multiple sensitive research projects around reproductive health. She was the Principal Investigator (PI) on this project which also involved collaboration with staff in the NHS. The project was funded by the Economic and Social Research Council (ESRC) and ethical approval was received from the UK National Research Ethics Service. Based on go-along ethnography, the project included mobile observations and in-depth interviews with different types of professionals whose work informed post-mortem practice -from pathologists through to hospital chaplains, coroners, and the police Ellis, 2019, 2020). Parent experiences of different types of loss, including miscarriage, stillbirth, neonatal death, and sudden infant death (SIDS 1 ), were also sought. A total of 22 in-depth interviews with bereaved parents and other family members were conducted. Parents were recruited via local online forums, memorial services, established support organisations, and mortuary MRI post-mortem consent forms. Interviews were conducted on a face-to-face basis or via the telephone, with one parent providing a written statement. Parents were invited to bring memory items to interviews to help them to talk through their experiences of loss. Notes were taken during the observations and interviews were digitally recorded. Data were analysed thematically using an inductive and reflexive approach (Braun and Clarke, 2019).

Study B: relational experiences of sibling bereavement
Study B was an ESRC-funded PhD project which applied a relational lens to explore how the death of a sibling continued to shape the lives of surviving siblings over the life course. Following university ethical approval, a single semi-structured object elicitation interview was carried out with 36 individuals at various locations across England, lasting approximately 2-3 hours each. Participants were recruited via national bereavement charities (the Compassionate Friends and Child Bereavement UK) as well as social media advertising and the university research volunteer list. Varying durations of time had passed between the sibling death and time of interview, with the shortest gap being 5 years and longest being 41 years. Although not specifically narrative interviews, an awareness of personal, social, and cultural narrative was maintained throughout data collection. Interview transcripts were coded and analysed using a narrative approach to thematic analysis (Riessman, 2008).
In what follows, we will draw on various sources, including fieldnotes, personal diary entries, anonymised emails, and other forms of research feedback to explore the role of emotions in research. We begin at the start of the research journey with a focus on recruitment.

Starting out: the power of (un)anticipated emotions
The role of researcher emotions during or after fieldwork has tended to form the central focus of existing literature. Attention is beginning to be paid, however, to emotive aspects of earlier stages of research (Callabero, 2009). In this section, we draw on examples from Laura's study to explore some of the unanticipated emotional challenges that researchers can face during the recruitment process. As she started her PhD research, Laura felt that her main challenge would be finding respondents willing to take part in a study on bereavement (a concern shared by Currie et al., 2016). She did not anticipate, however, that recruitment would pose significant emotional challenges for her as a researcher. She posted a recruitment notice on twitter and multiple people responded, many of whom did not fit the research criteria as the following fieldwork diary extract reveals: Within minutes of advertising the research on Twitter I received an email from someone fitting neither the age nor time lapsed requirements who is really keen to participate. I was really struck in the email by their need to convince me that neither of these things should be an issue. Clearly some people have a very strong desire to take part, which I wasn't expecting.
This presented an unanticipated dilemma for Laura. Ethical approval for the study was granted on the basis that participants would be 18 and above and bereaved more than 5 years. It was also essential that participants were of an age at the time of death that they would be able to recall their sibling and the relationship/s they shared. Laura did, however, aim to make the research as inclusive as possible. For example, she did not want to restrict participants according to sibling cause of death and sought to use a self-defining approach to recruitment to encapsulate a range of sibling relationships (Valentine, 2007). These decisions were made at the start of the research process to avoid prioritising one sibling's experience over another. Laura soon realised, however, that it was not possible to say yes to everyone while also maintaining the focus and integrity of the project. She felt significant anxiety and guilt about saying 'no' to potential participants who had experienced other forms of loss. The following fieldwork diary entry reveals the emotional distress this caused her: Today I had to do what I vowed I wouldn't and turn away 2 participants. I felt awful doing it and avoided sending the email all day. Those are the moments that it becomes starkly real that this is your research, based on your decisions. I explained my reasons as best I could but it didn't make me feel any better. Ultimately you're not denying them the chance to take part in your research, you're denying them the opportunity to speak about their sibling in a free and open environment for as long as they want. You're denying them the opportunity to be the sole narrator of their experience. For some, you're denying them the chance to be a sibling again, if only for a couple of hours.
Although an emotional and potentially painful experience, research consistently shows that bereaved people often do want to participate in social research, gaining enormously from talking about their experiences in depth (Buckle et al., 2010). Laura understood that in rejecting potential participants, she was therefore denying people the chance to talk about their experiences and felt the heavy weight of responsibility. In total, she had to turn down nine requests to participate for reasons such as the time of death breached the ethical requirement of 5 years lapsed; the participant's situation did not fit the research criteria; data saturation had been reached; and the data collection period had passed. The sadness she felt when responding to these individuals was often overwhelming. She sought to convey her sincere regret in her response to inquiries, as illustrated in the email correspondence below: Thank you for getting in touch. I absolutely agree that the loss of a sibling through stillbirth is just as valid and heartbreaking as any other sibling bereavement, and I'm really sorry that you've experienced that. I'm sorry to say, however, I'm looking to speak with people who lost their sibling as a young adult. This is so that people had time to form a living relationship and establish memories of a time before and after the sibling's death. It is in no way a statement of prioritisation. I'm genuinely really sorry but I wish you well.
Recruitment can throw up a range of unanticipated emotional challenges for researchers. This reinforces the need, therefore, for existing literature to consider emotional challenges at various stages of research, including those occurring prior to data collection. As illustrated here, embarking on her PhD research Laura felt responsible for the feelings of her non-participants. Developing sensitive communication with them about why they could not participate clearly entailed a certain amount of emotion work (Dickson-Swift et al., 2006;Hubbard et al., 2001). The responses to Laura's 'rejection' message, however, when they came, were gracious and understanding, helping to alleviate some of her guilt and anxiety. Such an overwhelming interest in participating in her project also offered her some reassurance, further reinforcing the underlying value of her research project (Dyregrov, 2004). What remains clear, therefore, is that while articulating and managing emotions in research is undoubtedly hard work, it can also be a productive experience for researchers (Reed and Ellis, 2020;Wouters, 1989).

Embodied emotion: managing the analytical process
There has been a growing awareness that conducting bereavement-focused fieldwork can be very emotionally challenging (Valentine, 2007). Researchers working in this field have subsequently sought to advocate various self-care strategies to deal with emotional fallout from peer support to writing reflective diaries (Valentine, 2007;Visser, 2017). Both Kate and Laura knew their research would be emotionally challenging from the outset. However, they both assumed that face-to-face interviews with bereaved parents and siblings would be the most emotionally intense aspect of the research, something which is reinforced by the wider literature on researcher emotions (Bloor et al., 2007;Hubbard et al., 2001). While the fieldwork was undoubtedly intense, what was often more challenging was the process of transcription and analysis which took place afterwards. In this section, we explore both authors' emotional experiences of this part of the research process, along with two of the self-care strategies they adopted to manage emotions.
While researchers often get emotional during interviews and observations, the presence of other factors in the field can provide some distraction. For example, when listening to participants during interviews, the researcher is often simultaneously concerned with making a good impression and acting in a way deemed appropriate for a sensitive researcher (Komaromy, 2019). However, the process of listening back to participants' experiences through the solitary process of transcription can be an emotionally intense experience. As Dwyer and Buckle (2009: 61) state, during transcription, researchers 'carry the individuals with us', ensuring that their 'words, representing experiences, are clear and lasting'. Laura felt that the individual grief of her participants was amplified through the process of listening, thereby increasing the emotional distress felt during transcription, as articulated in the fieldwork diary extract below: I feel so emotionally drained after hearing that interview played back. There is so much pain and hurt in her voice. It keeps echoing round my head like it's trapped in there.
Similar experiences are often felt by researchers when reading and analysing data. Some of the transcripts in Kate's study, for example, offered detailed and graphic accounts of individual experiences of baby loss. Reading through and annotating the transcripts as part of the analytical process, the researcher could physically feel and sense what the parents had been through. This process, although necessary, was also very sad. This is illustrated in the fieldnotes below: Sometimes, after spending the day in the data, I find it really hard to pick myself up off the floor and function normally, the data make me so sad.
Feeling emotional during transcription and analysis can assist with the interpretation of data and lead to the production of emotionally sensed knowledge (Evans et al., 2017;Holland, 2007). As Melrose (2002) argues, however, managing emotions at this intense stage of research can be particularly challenging. Kate and Laura sought to draw on a range of reflexive self-care strategies outlined in existing literature to manage their emotions during analysis, including sharing experiences with peers. Kate did not wish to 'confess all' to other social researchers. What was helpful, however, was discussing the data with her NHS collaborator -Elena (a Radiologist) -who had facilitated respondent access but had not participated directly in data collection. As a clinician, Elena dealt with the experiences highlighted by parent respondents regularly. Although sad, sharing experiences with Elena was hugely beneficial: it facilitated a more in-depth insight into the data and reaffirmed the value of the research. It also helped Kate to put work-life balance issues into perspective as illustrated by the diary entry below: Elena and I talked through some of the difficult experiences I had been reading in the interview transcripts. I had been feeling really bogged down with university administration all week and was really fed-up. Trying to gain a better understanding of parent & professional experience with Elena, although sad, helped me to put everything else in perspective.
Personal experience is often placed at the centre of discussions on self-care (Behar, 1996;Valentine, 2007;Visser, 2017). This has led to a set of recommendations centring on autobiographical and confessional style reflexive practice, both during the research and in dissemination (Borgstrom and Ellis, 2017). Existing literature emphasises the value of keeping a research diary to facilitate this process (Rager, 2005;Valentine, 2007). According to Punch (2012: 87), although researchers will never be 'able to fully understand the impacts of their emotional and personal struggles of conducting fieldwork', a field journal 'may encourage a more systematic and critical engagement with such issues'. Laura kept a diary throughout the research process. She found this process valuable but emotionally draining: I feel exhausted. I know I'm supposed to make notes but it's really hard when you feel so emotionally drained and actually all you want to do is sit and cry. I don't want to reflect on my feelings right now, I just want to let them all go and sit here, enjoying the silence.
As shown in this section, transcription and data analysis can be just as emotionally challenging as data collection. More attention must be given, therefore, to this issue in discussions on emotional labour in research. Reflexive self-care techniques such as sharing experiences with peers and keeping a diary can both provide useful tools in dealing with researcher emotions. Certain caveats, however, must be noted. For example, diary keeping, while therapeutic, can be tiring and the usefulness of sharing experiences is often contingent on the availability of appropriate collaborators and networks. Reflexive self-care practices, therefore, as Borgstrom and Ellis (2021) note, while heightening a researcher's emotional awareness, cannot always provide an outlet for emotionality. Researchers may start to internalise the sensitivities they are researching, carrying difficult emotions with them long after the project has ended. This suggests a need, perhaps, to develop more nuanced forms of emotion management which maximise participant understanding while stopping short of full emotional disclosure by the researcher. Before turning to consider this issue further, however, we move on in the articles penultimate section to explore the role of emotions in research dissemination.

Knowing your audience: disseminating emotive findings
As the literature has begun to show, emotions in research are not just restricted to the actual research process -to recruitment, data collection, and analysis -but also extend beyond this to dissemination (Evans et al., 2017). When preparing papers for publication, for example, researchers must attempt to strike a balance between doing justice to participants' stories while being mindful of reader experience and trauma (Jackson et al., 2013). Researchers can also feel guilty about turning sensitive participant stories into data (Mallon and Elliott, 2019). Emotional issues arose for both Kate and Laura in various forms of research dissemination as will be explored in this section.
Presenting papers at academic conferences was often challenging for Kate. Baby loss is a common experience and she was mindful that members of the audience may have experienced this form of loss. Presentations using data from this project, therefore, were always prefaced with a trigger warning about content. There was one incidence, however, where Kate presented a paper at a university some distance from home. The paper focused on exploring the issue of emotional labour in post-mortem work using film clips and images from the project. Once the paper had been given, the conference organiser invited questions from the audience, a request that was met with stony silence. The diary extract below details Kate's feelings: I was mortified, there was only stunned silence. I was worried that maybe the audience thought the paper was academically poor, or worse that I had upset people due to the paper's content. As I waited for my taxi to the train station two participants (former nurses) came outside and told me what a great paper it had been. I still worried all the way home on the train and didn't feel better until I started to receive emails a few days later from attendees who had been deeply moved by my paper. That is when I knew the value of what we were doing, & when the penny finally dropped, that silence on this subject is normal, and that this is precisely what we are trying to challenge with our research.
Existing literature focuses closely on doctoral student and ECR welfare in sensitive research (Mallon and Elliott, 2019). Laura was sometimes asked following presentations how she had found the process of doing sensitive research. While she appreciated this interest, Laura often felt uncomfortable discussing her feelings in public. Kate, by contrast, was asked less frequently about how she managed her emotions during her research. One issue that she did worry about being asked about during presentations, however, was whether she had experienced baby loss herself. This was personal information that she did not want to disclose to anyone. Fortunately, the issue arose only after the project had ended and when she was using her research findings to teach MA Social Research students about sensitive research. During a discussion about rapport building, a young male student asked this question directly. This took Kate by complete surprise and she had to quickly find ways of deflecting the question. She did not feel comfortable adopting a confessional style approach in any given context, but especially not in a teaching setting. This is reflected by the diary entry below: During the research process I only disclosed my own experience to participants when it felt appropriate (I only did this on a couple of occasions). I wanted the interviews to be about participants not about me. The issue of whether my own experience should form the basis of reflexive methods sections in academic papers, teaching or presentation never arose. It was too private and really no one else's business.
While disseminating data to academic audiences could be difficult, feeding back the findings to participants and other stakeholder groups could be particularly daunting. For example, Laura sent respondents a written summary of findings. Her greatest concern was that individuals would not feel well represented by the research. Disseminating the findings to participants turned out, however, to be an extremely rewarding experience for two reasons. First, it felt restorative to give something back to participants; to show them that they had been listened to and something had been created from their time and words (Rager, 2005). Second, it provided a timely reminder that the research was valuable and meaningful for those who took part, something that other researchers studying bereavement have often sought to stress (Buckle et al., 2010;Moss and Moss, 2012). Knowing that her respondents had gained something from research participation provided Laura with further reassurance that this was the case in her research on sibling bereavement. The following extract reflects the sentiments shared in many participant responses: Thank you so much for sharing this. It made me realise, I have only ever really read individual siblings' experiences. Every now and again there are some similarities to my feelings, but it's not the same as hearing a range of lived experiences. Of course we will all be different, but to see them together as themes is incredibly validating. I just wanted to share that I have found it such a valuable thing to read and feel really grateful that you did this piece of research.
Although seldom discussed by existing literature, there are also various issues to be considered when disseminating findings to various stakeholder groups through public engagement and impact. For example, researchers often need to make sure that forms of engagement meet the needs of different audiences (from stakeholder groups to members of the public). Working with sound and visual artists, and a graphic designer, Kate and her research team curated a touring exhibition 'Remembering Baby' 2 based on their research. It was an interactive exhibition featuring visual images, film, memory quilt, physical objects, sound installation, and parent/sibling artwork. Some of the installations depicted hospital processes and medical images and wooden memory boxes, including heart-breaking items (tiny baby grows). The research team worried about how this exhibition would be perceived and therefore consulted extensively with bereaved parents through charities to ensure the exhibition was curated sensitively. Although a significant amount of emotion work went into ensuring the exhibition was appropriate, it was such a rewarding experience when the team received positive responses such as these anonymised comments below: A truly powerful and emotive exhibition expertly put together with sensitivity and professionalism. Something the exhibitors should be proud of.
Thank you. This came at just the right time. Our due date is coming up and I really needed something to help me work through this.
As indicated here, when considering the impact of emotions on the researcher, further attention must be given to a range of dissemination activities, including teaching and impact. The emotional challenges faced by the researcher during dissemination can be sudden and unexpected, requiring researchers to think 'emotionally' on their feet (Woodthorpe, 2007). While some of these emotions can be challenging, they can also be extremely lifeaffirming, as illustrated by Kate's exhibition experience, boosting researcher morale and further reinforcing the value of social research. To cope with the unpredictability of emotions in dissemination, however, researchers often adopt the role of 'emotional juggler' (Bolton, 2001) as they try to protect their own and their audience's emotions simultaneously. Furthermore, researcher self-disclosure during dissemination can create rather than alleviate researcher feelings of emotional discomfort. It may even challenge the boundaries of professionalism (as indicated by Kate's teaching experience). We need to think, perhaps, of developing additional modes of emotion management that can strike an adequate balance between emotionality, analytical rigour, and professionalism. We will move on to explore such an approach more fully in the conclusion.

Conclusions
This article reinforces the importance of the role of emotions in research. Research is a dialogic and co-productive endeavour between researchers, participants, and other potential stakeholders (Sinha and Back, 2014;Bell and Pahl, 2018). It is essential, therefore, that researchers acknowledge the ways in which they both effect and are affected by their research. As Woodthorpe (2007: 9) argues, 'it is an arrogant researcher who will dismiss their emotions and feelings and render them invisible in their analysis'. Although existing literature does recognise the role of emotions across different areas of the research process (Jackson et al., 2013), however, there has been a tendency to concentrate on emotions during and immediately after data collection. We have sought to go beyond this by considering the role of researcher emotions across other parts of the research journey from recruitment through to knowledge exchange and impact. We have also highlighted the dual-edged nature of emotions in research. Not only can emotions lead to the production of emotionally sensed knowledge -as argued by existing literature (emerald and Carpenter, 2015;Evans et al., 2017;Holland, 2009) -but they can also reinforce the value and importance of social research (Reed and Ellis, 2020;Wouters, 1989).
Both the authors of this article are sociologists at differing career stages. Many of the emotional challenges and benefits they have faced throughout their research, however, have been similar. Differences between them have tended to relate not to the researchers own experiences, but rather to the perceptions and reactions of others. This was highlighted for both authors during dissemination -particularly when presenting their research findings at academic conferences or when using their research to teach students. Existing literature has tended to highlight the importance of attending to the needs of PhD students and ECRs who may be conducting emotionally challenging research for the first time (Mallon and Elliott, 2019). While important, ECRs (as Laura's experiences show) may not always welcome that level of public attention to their feelings. By contrast (as shown by some of Kates' experiences), emotional challenges can be perceived as less of an issue for more experienced researchers who may be viewed as better equip to deal with them. While vital support mechanisms must be in place for researchers at the start of their careers, therefore, it is important to ensure that these are not patronising in tone. Furthermore, we must also acknowledge that all researchers -no matter what their career stage -can be emotionally affected by their research.
Recommendations for emotion management have increasingly centred on autobiographical and confessional style research narratives whereby the researcher is encouraged to 'bare all' to participants, to peers, and to wider members of the public throughout the research process (Behar, 1996). While both authors support the value of reflexivity, as we have sought to show in this article, confessional style narratives may not always be desired or appropriate. We draw on the sociological concept of Verstehen (Brewer, 2011;Sumner, 2011) to suggest a more nuanced approach. Using the concept of Verstehen, we argue, could help us to develop an approach to research that is almost confessional. Such an approach, we argue, would require the researcher to tread the boundaries between the engaged and the analytic. It would encourage researchers to build empathy with potential and consented participants through reflexive practice (e.g. writing memos and notations) before, during, and after research. Full emotional disclosure by the researcher, however, to either participants or publics is neither the aim nor an expectation of this approach.
Rather, the focus remains on developing a systematic account of participant words and stories. Such an approach, we argue, acknowledges the role of researcher emotions privately while ensuring the voice of the participants remain centre stage in public accounts. This approach, we argue, feeds into existing feminist debates about positionality but draws on connections of emotional experience rather than social identity.
There are, of course, limitations to taking an almost confessional approach. It cannot, for example, prevent the need for researchers to think emotionally on their feet (Woodthorpe, 2007) when they are put on the spot about their emotions either during or after research. Researchers may need additional support mechanisms -from counselling to personal support networks -to help them navigate these continued challenges (Rager, 2005). It could, however, provide an emotional buffer for researchers who find full and/or public disclosures of personal information challenging. Furthermore, by not being completely emotionally available to either research participants or other stakeholders, researchers may also be able to maintain a clearer set of professional boundaries both during and after research. We argue therefore, that the almost confessional approach which occupies a space 'between passion and intellect, analysis and subjectivity, ethnography and autobiography, art and life' (Behar, 1996: 174) could provide researchers with an additional tool with which to navigate some of the most difficult emotional challenges in social research.

Authors' note
The ideas for this article were first presented as part of an opening panel on Wellbeing at the White Rose Doctoral Training Partnership (WR DTP) annual conference in 2020.