Received date: 14 May 2012 Accepted date: 10 July 2012
A growing literature highlights the significance of relationships within research, and the ways in which social characteristics such as gender and ethnicity shape these interactions. This paper explores these relationships within a study of the experience of infertility among British South Asians, in which the researcher was a woman of white British ethnic origin. Age, gender, language, ethnicity, educational status and reproductive history all played a role in shaping relationships in this study and demonstrate the relational re-production of identity through research interactions. Differently configured ethnicidentities between the researcher and the participants played a role in shaping the data collection, but did not necessarily appear to have a negative impact on the research. It is argued that the need for ‘matching’ of researchers and participants by reference to an essentialised understanding of aspects of social difference is flawed, and a more nuanced consideration of this relationship is required among social science researchers.
ethnic matching, methodology, reflexivity, research encounters, researcher identity
Recognition of the importance of identity within the research setting has led to a call for increased reflexivity by researchers, that is, a need to question their own positionality and the power dynamics within the research process (Riley et al, 2003; Ali, 2006). This paper addresses these relationships within the context of a study that explored British South Asians’ experience of infertility. The researcher was white British. The specific social characteristics of the researcher and the participants shaped the production of understanding in the study at the micro-level. Following Bourdieu, who suggests that social scientists should step back and scrutinise the ‘view from above’ that their epistemological gaze imposes (Jenkins, 2002, p. 49), this paper explores the need for researchers to acknowledge their position as social agents and the impact that this has on the research process. Bourdieu refers to this process as ‘objectified objectification’ (Bourdieu, 1990, p. 30), in which social scientists must apply the epistemological project as much to their own actions and frames of reference as to those of their participants. This position is aligned philosophically with that of a reflexive research practice in which a ‘conscious attempt’ is made to understand what social understandings have been (re)produced in the research process (Riley et al, 2003). This paper therefore has two aims. First, it aims to explore how the matching of researchers and participants according to particular social signifiers, including but not limited to ethnicity, shapes research interactions. Secondly, in attempting to objectify the processes of objectification in this study, it aims to demonstrate a reflexive research practice.
Considerations of identity and belonging or ‘insider/ outsider’ debates have a long tradition in social science research (Morris et al, 1999). The emic/etic dilemma represents a long-standing desire to understand the impact of a researcher’s position or standpoint in relation to participants and their social worlds. With historical origins in social anthropology (e.g.Malinowski, 1922), these debates have included a range of disciplines and fields of enquiry (see Kanuha, 2000). Of note, the significance of the identities of the researcher and the participant for how research is carried out in practice has been highlighted by feminist writers such as Ann Oakley (1981), Stanley and Wise (1993) and Janet Finch (1984). These authors highlighted the importance of woman-centred research, in which the emphasis was on women interviewing and empathising with other women in a way that men could not. The assumption in this standpoint feminism was that only women could fully understand other women and their shared inequality in a patriarchal context (Harding, 1991). Recognition of the power dynamics in research by feminists was crucial to the development of debates in this area: it was no longer taken for granted that the research interaction itself was a neutral space for data collection. However, these assumptions were later questioned by postmodern writers and post-colonial feminists, who argued that by prioritising the significance of shared gender, white mainstream feminism had failed to recognise other forms of difference, and thus argued that white women are not and cannot be experts on all women (Mohanty et al, 1991). This highlighted the need to recognise that race, ethnicity, culture and, most importantly, experiences of power inequalities and racism in a postcolonial context could be as significant as, or even more significant than, gender for a shared understanding between women.
In parallel with feminist debates about the importance of shared gender in research encounters, scholars in the field of race and ethnicity research in the 1970s and 1980s argued that ethnic origin was the most significant identity when understanding relationships between researcher and researched. For these authors, the impact of the researcher’s ethnicity was crucial to the success of the research interaction. They maintained that white interviewers conducting research with black respondents would have a particular (negative) biased outcome, especially if the research topic itself related to racial issues (Schuman and Converse, 1971; Campbell, 1981). The proposal was that respondents would be less truthful if the interviewer was of a different ethnicity to themselves. It has been suggested, by those for whom the matching of individuals is a desirablemethodological tool, that there are benefits to be gained in terms of the validity or authenticity of the research accounts produced via such a device (Gunaratnam, 2003). It is argued that similarities between those engaged in research may engender a deeper understanding and may also displace traditional power relationships between interviewer and interviewee, creating an alternative methodological and epistemological position (Collins, 1990; Bhopal, 1995).
In much the same way that a prioritisation of gender as the defining feature of research relationships was questioned, several authors have also questioned the notion of matching researcher and participants according to ethnicity (Edwards, 1990; Phoenix, 2001; Gunaratnam, 2003; Culley et al, 2007). For example, Rhodes (1994, p. 556) has argued that excluding white researchers from research with black people suggests that blackness connotes homogeneity or an ‘artificial harmony.’ To suggest that research can only take place between individuals and groups matched by ethnicity is to ignore other formations of difference that may be present and may be differently perceived and differently valued by individuals and groups. Furthermore, claims about authenticity provided by this approach suggest an essential notion of truth that is based in a realist ontological position, and which can be uncovered by rigorous social research. It also maintains that certain accounts are more genuine than others, since they are congruent with the truth (Rhodes, 1994). The suggestion that this or indeed any other signifier transcends all other formations of difference is problematic. Rather, it is important to see how ethnicity is continually adapted, contested, resisted or reinforced within the research arena (Nayak, 2006). Therefore it should not be assumed that white scholars cannot engage in research with black minority ethnic communities, or that they cannot access authentic accounts, or indeed that black researchers cannot conduct research with white participants. Instead, white researchers should be aware of white privilege, should seek to challenge its effect on research, and should subject research to processes of reflexivity (Andersen, 1993).
The work of these writers, and the suggestion that race/ethnicity and other social signifiers as they relate to the research process are complex and shifting, can be aligned philosophically with the work of postmodern writers who have contributed to the idea of a de-centred subject (Foucault, 1980; Hall, 1992; Anthias, 2002). Here ethnicity is not the only, or even the most significant, category to which people relate themselves in any social situation. For example, gender can at times bemore significant (Culley et al, 2007; Egharevba, 2001; Bhopal, 2010). It is therefore simplistic to imagine that a person can ever be wholly an insider or an outsider in any given social setting or on the basis of any unitary facet of identity, since the boundaries and markers of membership are constantly shifting, and are more flexible at some times than at others (Coleman, 2007; Ryan et al, 2011; Bhopal, 2010). Membership of a particular group or community is also mediated by a number of other social signifiers, including gender, religion, language, health status, educational and occupational status and age.
This paper draws on experiences of data collection within a study of the experience of infertility in British South Asian communities to illustrate the ways in which research interactions are carried out amidst a range of complex social signifiers (Gunaratnam, 2003; Culley et al, 2007). In doing so the paper aims to demonstrate reflexive practice and calls for a more sophisticated understanding of the intersection of a number of social signifiers in operation within any research setting or interaction.
Infertility is estimated to affect around 9% of women aged 20–44 years, which means that over 72 million women worldwide are currently affected and around 40 million are seeking fertility treatment (Boivin et al, 2007). In the UK it is estimated that approximately one in seven couples will experience difficulty in conceiving a child (Human Fertilisation and Embryology Authority, 2008). The social, psychological and material implications of infertility for individuals, and especially women, have been well documented, particularly in the American and British contexts (Greil, 1991; Franklin, 1997; Inhorn and van Balen, 2002; Throsby, 2004). However, the dominant construction of infertility in western countries is that of a medical problem affecting white middle-class couples (Culley et al, 2009). There has been a corresponding absence of research on how infertility is experienced within minority ethnic communities in western countries. This study was part of a body of work (Culley et al, 2004, 2006; Culley and Hudson, 2009a,b) devised to address this theoretical and empirical lacuna in the knowledge of infertility experiences. Using a qualitative, interpretive approach, it employed a multiplemethod design in order to understand the experience of infertility within British South Asian communities (Hudson, 2008). Comparable evidence from anthropological work in South Asia suggests that within highly pronatalist cultures, involuntary childlessness is heavily stigmatised (Bhatti et al, 1999; Papreen et al, 2000; Reissman, 2000; Winkvist and Akhtar, 2000). This studywas therefore designed to explore public perceptions of infertility and its treatment, as well as individual experiences of involuntary childlessness.
A favourable ethical review was obtained from the Health and Life Sciences Research Ethics Committee at De Montfort University, Leicester. Participants were given both verbal and written information about the study, and gave their consent either verbally (focus groups) or in written form (interviews).
The first phase of the study, which explored these public or community perceptions, consisted of 13 single-gender focus groups with a total of 87 participants of Indian, Pakistani and Bangladeshi ethnic origin (see Table 1). These focus groups were not designed to include those with specific experiences of infertility, but included ordinary members of the public, recruited via community contacts and key individuals in local community organisations. The groups were homogeneous in terms of ethnicity, religion and age, although there was some diversity within the groups.
In a number of cases, focus groups for the study were convened and facilitated by one of a team of bilingual community researchers working with the researchers, as opposed to having live translation of the discussion (see Culley et al, 2007). Where this was the case, and I was not directly involved in facilitating the group, instead I attended the group, took notes, observed and listened to discussions.
The second phase of the study included in-depth interviews with 15 individuals (see Table 2) of South Asian ethnic origin, who had experience of infertility. Interviews have been used by a number of authors working in the field of infertility (Greil, 1991; Inhorn, 1994; Franklin, 1997; Letherby, 2000; Becker, 2000; Reissman, 2000; Daniluk, 2001; Throsby, 2004; Thompson, 2005), demonstrating their value in accessing the stories of involuntary childless individuals in a range of social settings. Participants were recruited via community contacts and word of mouth (n = 7), Internet advertising (n = 5), private infertility clinics (n = 1) and patient support organisations (n = 2). All of the interviews with infertile individuals were conducted face to face in English by the author either in the participant’s home (n = 12) or in anothermutually convenient location (n = 3). In total, 13 of the 14 interviews were conducted one to one with women, and one interview was conducted face to face with a couple.
Although at the outset of the research I had been concerned that my perceived ethnic difference might inhibit research interactions in some way, what became apparent throughout the course of the data collection was that a number of social signifiers were of importance in shaping the relationship between researcher and participants, a process which could not easily be attributed to some kind of simplistic and unitary ethnic dissimilarity. In the discussion that follows, examples from the fieldwork are used to highlight the ways in which ethnicity is mediated through, and at times superseded by, age, gender, educational status and reproductive identity.
Focus Group 10 took place in a women’s information and support centre with four younger Indian and Pakistani women, most of whom were in their twenties. The contact person who had convened the group, and who was also taking part in the focus group discussion, requested that the facilitator should be another woman of South Asian heritage. This was arranged, and a woman of Pakistani heritage led the group. I attended the discussion in my role as co-facilitator, to observe, take notes and assist where necessary. As the discussion, which took place in English, got under way therewere several occasions when the younger women directed their comments and responses towards me, effectively excluding the older South Asian facilitator. This was particularly apparent when the discussion turned to issues of accessing higher education and juggling expectations about education, work and family life.
(Participant C) The older generation had what, six, seven, eight kids and nowadays you know, if you asked us we’d say ‘Oh no, we don’t want seven, eight kids, two or three’s just enough! (laughter)
(Participant B) It’s only because nowadays careers take over rather than in those [days] they didn’t and they were just at home and looking after the children and making more children. Nowadays it’s more career-minded women that are sort of taking over.
(Focus Group 10, younger Pakistani and Indian women)
It appeared that, in these instances, the fact that I was of a similar age to them was more important in enabling the discussion of shared moments of experience than the ethnic origin and potential shared culture of the older facilitator. Therefore to presume that I, as a white person, was classed solely as an outsider based on my perceived ethnicity would be highly problematic in this case, as the relationships that I fostered with the research participants were not contingent upon this categorisation alone.
Characteristics such as age and educational experience also appeared to facilitate the building of rapport with a number of interview participants later in the study. For example, studying in higher education (doing a PhD) emerged as a shared experience, as did age. Conversations before and after the interviews often turned to work life, home life and studies. It was these interactions and shared experiences that were important in developing rapport and establishing relationships. These interactions were also potentially facilitated by a shared gender identity, since the experience of juggling home and family, education and employment is often associated with a feminine gender identity, particularly among women of childbearing age (Emslie and Hunt, 2009).
The sharing of gender was also significant in interactions within the focus groups. Focus Group 4 was carried out in another community centre with a group of older (over 45 years of age) Gujarati-speaking Indian women, and was led by two Gujarati-speaking facilitators. In a number of instances the women in the group stopped the discussion and translated what was being said into English, so that I and another Englishspeaking researcher who was also present could be included in jokes that were being made about men during the group discussion. This sharing of gender worked to unsettle notions of difference, which operated at the linguistic level (since this group was conducted mainly in Gujarati) and at the ethnic level, as the participants and facilitators were of Asian Indian descent and the researchers observing the group were white. This finding again suggests that ethnicity is perhaps not the only, nor the most significant, social identifier at play in constructing the research interaction, nor is it necessarily stable within research encounters. In this case, shared gender was more central to the shaping of the encounter than shared ethnicity, language or age. Most importantly, the diversity of social and cultural difference within the groups in this study demonstrates the unpredictability and complexity of dynamicswithin research interactions (Archer, 2002).
Given the sensitive nature of the topic, it was important to develop a rapport with the participants. In most cases this was fostered through the numerous phone calls and emails that I exchanged with the interview participants when organising the interview, or afterwards when some participants made requests for information. Developing a relationship with the participants was also an integral part of the interview process, particularly since, in most cases, these took place in people’s homes. Ultimately, the research interaction is a human one (Lalor et al, 2006; Bolognani, 2007; Bhopal, 2010), so listening with empathy, developing trust and sensitively facilitating the sharing of experience are skills required by the researcher irrespective of who they are.
This is not to say that ethnic identity has no impact on the research process. Instead, I suggest that it has an impact on how the research is shaped but does not necessarily add authenticity if matching takes place (Rhodes, 1994; Nayak, 2006). My ethnicity and status as an outsider to South Asian culture facilitated recruitment to the study, adding an extra layer of separation and therefore confidentiality between myself and the participants. This may have been particularly pertinent given the topic of infertility. Had I been a member of a South Asian community, as suggested by some authors (Bhopal, 1995; Berg, 1999), this might have actually limited the number of people who would have been willing to speak to me. In this respect, being perceived as an outsider could be construed in a positive light. In a further example of this, a number of the participants described at length the significance of religious or cultural traditions, such as those which might take place when a male child is born, which they might not have described for someone who was considered to be a cultural insider. Expectations about what existing knowledge researchers already hold of the social context and what kinds of expertise they have may therefore inhibit or facilitate discussion. In the case of this study, there was a great deal of elaboration on the part of the participants and a willingness to speak to me, an apparent outsider.
... in our sort of culture, if you don’t have a child, you don’t, you can’t do certain prayers or, you know, like, if it’s a baby’s, sort of what you call christening, if this lady [didn’t] have a baby, she’s not involved in these kind of things, it’s you know, all those issues.
(Aisha, individual interview)
Thus participants’ willingness to speak appeared to be linked to my status as a cultural stranger (Bhopal, 2010). Whether this was so cannot, of course, ever be known for certain, but what is clear is that the dichotomy is disturbed by the findings from this study. Recognising the importance of shifting and multiple identities is therefore crucial to understanding the research relationship (Gunaratnam, 2003). Also important here is the recognition that participants could be described, as Ryan et al (2011, p. 57) also suggest, as ‘performing’ culture for the benefit of the researcher. This is particularly pertinent where the author is constructed as a cultural stranger and this positioning is used by the participants to demonstrate expertise about South Asian culture. Given these complexities, simple notions of matching, based on essentialised understandings of ethnic identity, culture and belonging, are therefore flawed and do little to assist the reflexive process (Adamson and Donovan, 2002).
A final reflection on the significance of identity, access and status concerns the issue of reproductive history. Many writers working in the field of infertility have described their own experiences of infertility and assisted reproductive technologies (ARTs). In many cases, a personal history of involuntary childlessness has been the catalyst for their work in this field (Letherby, 2000; Becker, 2000; Thompson, 2005). This is not an experience I have shared with those authors or with the participants in this study, although I have no children. There were instances in which my reproductive identity became significant in the course of the data collection. A number of participants were interested to know more about me, and, in the course of the discussions that we had before and after the interviews, asked whether I had children. I offered the information that I did not, and also that I was not married, which for some seemed to make sense; if I was not married, I would not yet have chosen to have children. Because of my age at the time of the interviews (29 to 30 years), this did not appear to be problematic. Had I been notably older, the fact that I did not have children might perhaps have led people to ask, or at least possibly to assume, that this was because I could not have them. A discussion with one participant alerted me to this possibility when she asked what had led me to be interested in this area of research. The notes that I made after this interview read:
I thought, perhaps she had wanted to know if I had personal experience of infertility, so although she didn’t ask directly, I offered this information. I don’t recall any of the others asking me this.
In this example,my perception of this interaction was that she wanted to know whether I had experience of infertility, although of course I could have misinterpreted this question. If I had had this experience, sharing this information would no doubt have had an impact on the interviews, and this facet of our identities could have become the most salient in shaping the research, although this is of course hypothetical. Since few people asked, it was difficult to know howIwas perceived in this respect. I felt that had I myself been a parent, this might have been a difficult piece of information to convey, given the topic of research and the participants’ involuntary childlessness.
This paper aimed to demonstrate reflexive practice and to call for a more sophisticated understanding of the intersection of a number of social signifiers in operation within any research setting or interaction. It has attempted to highlight some of the intricacies of negotiating research interactions, and how individual identities and performances, whether as researcher or participant, come to have a bearing on how research narratives are made. Following Bourdieu (1990), these reflections allow a distancing from the research process and a questioning of the particular frame of reference that I, as a researcher, bring to this study.
The findings of this study highlight the importance of a number of social signifiers for shaping the research interaction. Differently configured ethnic identities between the researcher and the participants clearly influenced data collection, but did not appear to have a negative impact as reported in other studies (Campbell, 1981). Instead, being seen as an outsider could, in some respects, have a positive role in facilitating access and engendering positive feelings about confidentiality. This finding further problematises the tendency for ethnicity to be regarded as the central axis of difference in research (Flintoff et al, 2008). Instead, this study demonstrates that a number of subjectivities were of importance at any one time.
For example, gender was clearly a salient characteristic, which at times superseded and mediated other identifiers, such as age, language and ethnicity. It could be suggested that this is linked to the perception of infertility and its treatment as ‘women’s work’ (Franklin, 1997, p. 113). Thus the topic of this research may have been seen in gendered terms, especially in highly pro-natalist communities such as those originating from South Asia, where childlessness is considered to be a serious transgression of normative practice and associated expectations about childbearing for married women (Culley et al, 2004; Hudson, 2008). It was clear from the participants’ accounts, both in the focus groups and in the individual interviews, that the experience of infertility was gendered, with women being much more likely than their partners to be affected by an inability to conceive (Hudson, 2008). The cultural ideal of transition to motherhood for adult women is clearly a pervasive one, both in western contexts (Maher and Saugeres, 2007) and in other socio-cultural settings (Ram and Jolly, 1998). The sharing of a gendered identity within the research setting and the associated rapport building and reflections on reproductive identity might have been differently configured if men had been more central to the research and/or if the researcher had been male. Indeed it is more than possible that my identity as a woman was what ultimately permitted me any access at all to narratives about (in)fertility and reproduction.
An analysis of subjectivities in research allows us to account for and challenge ideas about matching unitary facets of subjectivity. Age, gender, language, ethnicity, education and reproductive history all played a role in shaping the relationship with the participants in this study, and demonstrate the relational reproduction of identity through social interactions. No one signifier was ever-present as the defining shared characteristic, but rather it was found that different elements of identity became important at different points in the discussions. This is a particularly important finding when working with ethnicity and culture. It represents a rejection of the culturalist approach to research in which findings are interpreted through the lens of cultural difference, and encourages researchers to explore instead the familiar as well as the exotic in research interactions (Kelleher and Hillier, 1996).
The need for flexibility and reflexivity in research is significant, and is perhaps hidden when the research that is being carried out does not cut across perceived racial/ethnic groupings, or gender (Oakley, 1981). The need for the researcher to comment on the impact that their identity has on the interaction will often not be seen, especially when the researcher and participants are matched (e.g. with women interviewing women, or white researchers interviewingwhite participants), and is therefore left unmarked and unproblematised. Exploring and deconstructing our perceptions about the nature of social signifiers in research allows us to better understand what it is that is being studied and why.
This paper also highlights the need for researchers to explore and problematise not just the traditionally universalising categories of ethnicity and gender, but also those perhaps less frequently discussed or interrogated in research, such as language, age, education and shared experience. Through an analysis of these multiple points of difference and connection, researchers can further explore the dynamics of research interactions and the ways in which they affect the production of findings. It is not always the most obvious or expected differences that have a prominent role in shaping relationships; those which have previously been hidden or given less consideration can become defining features. In this regard it would therefore be impossible to match researchers and participants, even if it were desirable to do so. The possible list of criteria against which researchers and participants could be matched would be endless.
What is clear from this paper is that the ongoing dilemmas facing researchers in this context are com plex, and that they present ambiguities and contradictions which are not easily overcome, but instead need to be continually worked on and reflected on in order to continually question the epistemological assumptions that underpin social science research. This speaks directly to wider debates about the quality of qualitative research, particularly the proposal that researchers should engage in a ‘methodologically aware’ practice in order to enhance the plausibility of their findings (Seale, 1999). Researchers should therefore continue to contribute to open debate and discussion about the practical and theoretical decision making which informs their work, and the impact that their own subjectivities have on the research process.