Understanding delayed access to antenatal care: a qualitative study Rosalind Haddrill

3.2.2 Data collection

The choice of data collection methods reflects the aim of the study and the need to answer the research question in the most effective way. The question sought multiple perspectives amongst a diverse group of participants, including vulnerable women, with a complex range of experiences and attitudes, and from a range of backgrounds. This necessitated the application of methods which were flexible enough to respond to this diversity, whilst detailed and individual enough to generate useful data in an ethical way, and practically feasible.
Interviewing is one the primary methods of gaining multiple perspectives used by qualitative researchers, and probably the most widely employed (Marshall and Rossman 2011, Green and Thorogood 2014). A flexible and powerful tool that can open up many new areas for research, qualitative interviews can be used to investigate research questions which would be difficult to investigate quantitatively, for example through the use of structured interviews or questionnaires. Such tools, as considered in chapter two, risk oversimplifying a complex subject and further alienating a group of women who, statistically, are more likely to be marginalised anyway. In contrast, qualitative interviews reflect a more personal approach, with greater potential to examine the social and psychological processes leading up to or following an event, in this case late booking for antenatal care (Bryman 2008). Interviews also create a specific focus and allow participants to reconstruct events and experiences in a way that is not amenable to alternative methods such as observation. Observation has greater potential to reflect the researcher’s (etic) perspective rather than the participant’s. It has greater relevance for studies of social interaction and on-going experiences, rather than the consideration of individual feelings and actions around a fixed point in the past, for example relating to pregnancy discovery and late booking (Liamputtong and Ezzy 2005, Holloway and Wheeler 2010).
Interviews, whether individually or in groups, give a voice to participants and provide a varied source of data which allows health researchers to see the world through their eyes, exploring and understanding peoples’ experiences, meanings, health behaviours and views towards care (Ziebland and Wright 1997, Hall, McKenna and Griffiths 2012). The type of interviewing and the techniques used to record and present the interview data are highly influential on any findings (Fontana and Frey 1998). The interview setting is also significant. The participant’s choice of setting reflects the ‘naturalistic’ approach fundamental to qualitative research, allowing the exploration of social events as experienced by individuals in their natural context (Malterud 2001a). Overall the flexibility and personal nature of interviews reflects the inductive approach of the study, and has the potential to generate rich, detailed data about the everyday meanings associated with pregnancy discovery and care initiation.
Semi-structured interviews are probably the most common form of qualitative research conducted in the UK and are widely used in health research where they frequently provide the basis for exploring both practitioner and client perspectives (Fox et al 2001, Holloway and Wheeler 2010). It is appropriate for a project such as this, beginning with a fairly clear focus, providing some structure in order to compare findings, whilst allowing for flexibility in the interview process (Bryman 2008). The use of pre-determined questions reflects the overall study design. In particular, the application of prior knowledge of the topic, principally in relation to the socio-demographic characteristics of the women gained through an initial literature review, the multi-professional study design and the pragmatic health services research focus (Ziebland and Wright 1997).
These a priori areas for exploration were intended to reflect the potential complexity of meanings and experiences of the participants, that could not be examined with a positivist methodology. Rather, the interviews reflected the interpretive approach, with its aim to analyse how participants interpreted the ‘world’ of late booking (Liamputtong and Ezzy 2005, Bowling 2009). The resulting ‘guided conversations’ (Lofland and Lofland 1995) used a flexible interview guide based around the research question, an initial review of the literature and discussion with a group of academics and health and social care practitioners involved in the care of pregnant women. This provided a framework for the semi-structured interview process and ensured some consistency in the data collected, reflecting the influence of the pragmatic health services research approach and research governance requirements, for example ensuring a useful contribution could be made by the research (DoH 2005).
Any interviewing risks imposing a potentially inappropriate frame of reference on people, and limiting their responses, which in turn limits the researcher’s view of their world. The flexibility and minimum structure of the semi-structured method creates opportunities for spontaneity but also for meanings which can be clarified and followed up. It encourages participants to reveal their point of view and talk honestly about what is important to them, not the interviewer, but also to suggest alternative directions for and ways of thinking about the research (Bryman 2008). The aim of the interview is to go below the surface of the topic being discussed, exploring what participants say to produce rich and detailed data, whilst being sensitive to the language and concepts used, but also to uncover new areas or ideas that are not anticipated at the outset of the research (Britten 2006). As Leininger (1985) comments, interviews are designed to elicit both definitive and unexpected kinds of information from the interviewee.

3.2.2.1 Individual interviews

“As a researcher I sought to have people tell me about their lives from their perspectives rather than to force my preconceived interests and categories upon them. So I listened” (Charmaz 1991: 275)
Individual semi-structured interviews have the potential to provide a natural, informal setting for people to talk about their lives and describe their personal experiences in their own words (Britten 1995). The approach is based on the fundamental qualitative assumptions that a participant’s personal perspective is meaningful and knowable, and that it should unfold as they view it (emic), unlike structured interviews where the balance of control lies with the interviewer (Fontana and Frey 1998, Patton 2002, Marshall and Rossman 2011). This reflects a feminist criticism of structured approaches which may separate people from the context that influences their choices, and impose a hierarchical (and potentially exploitative) relationship upon participants. In depth semi-structured interviews have been called the archetypal ‘feminist method’ (Kelly, Burton and Regan 1994).
The use of individual interviews as a method reflects pragmatic, methodological and ethical considerations. These include the availability of participants, but also the richness of the data desired, which may be compromised in a group setting, as well as issues of privacy and confidentiality. During individual interviews people’s responses are less influenced by the presence of others, so they may be more prepared to discuss sensitive matters. Such interviews may provide more detail about an individual’s understandings and experiences than can be gained through group interviews, particularly where participants are unconfident or reticent (Bryman 2008). These were the primary reasons for using the individual interview method.
However the individual interview method does have its limitations. Focusing on individual rather than social interaction, it may neglect context, and its focus on verbal communication may favour more articulate participants (Kvale 1996). It is widely acknowledged that all interviews are ‘situated’ rather than natural encounters, unique social situations that occur between the interviewer and interviewee, and as such cannot be replicated between participants. They are inevitably constrained by the context in which they are collected and the relationship between interviewer and participant (Murphy et al 1998). As such they represent personal, contextual accounts rather than literal descriptions. However this reflects the qualitative approach, which values motivations, thoughts and personal ‘truths’ rather than factual accuracy (Holloway and Freshwater 2007).
At their most negative, individual interviews could be perceived as threatening, abnormal interactions, with a mistrust of the researcher or ‘official’ involvement, and this may be reflected in some women’s reticence to talk about their experience. Even at its most positive, fundamentally the encounter is still one of strangers meeting, with a risk that it might generate opportunities for ‘impression management’ (Murphy et al 1998). This might result in ‘public’ accounts of socially acceptable views, rather than the more difficult to ascertain ‘private’ views, as a result of the interviewee making assumptions of what the interviewer wants to hear or find out. Whilst multiple interviews would alleviate this unfamiliarity and the potential lack of trust that accompanies it, the nature of the interviews as soon after booking as possible and concerning women’s experiences and decisions around a fixed point (i.e. booking), and challenges with recruitment, mean that this was not appropriate or feasible.
These challenges meant that the qualitative interviews required considerable skill and sensitivity to the complexities of interpersonal interaction, and reflexivity on the part of the interviewer (Marshall and Rossman 2011). As Ezzy (2010) argues, all interviews are emotional and embodied performances, and good interviewing is facilitated by a reflexive awareness of this. The interviewer is also an active participant in the interview process and takes part in co-constructing meaning from the resulting social interaction. Interviewers need to be honest about this, and must reflexively account for the interplay between the social positions of themselves and the participants in this construction (Britten 2006, Bowling 2009, Holloway and Wheeler 2010). Oakley (2004) argues that interviews, particularly those with women, are most effective when the relationship between the parties is non-hierarchical but also when the interviewer invests their own personal identity in the relationship (‘no intimacy without reciprocity’ (p264)). This is discussed further in chapter four.