Doing Focus Groups. Rosaline Barbour
who looked at sexual functioning and cancer; and Klaeson et al. (2011) who talked with women about the impact of breast cancer on their sexuality. Rather than simply documenting attitudes or experiences such work has often provided detailed and nuanced accounts, such as did the study by Davis et al. (2014) into heterosexual men’s condom use and their explanations of their strategies of resistance. In contrast to one-to-one interviews, which excel at eliciting personal narratives, focus groups allow researchers to look at more subtle, but potentially highly relevant understandings and behaviours.
Limitations of focus groups
With regard to the limitations of focus groups for eliciting narratives, the issue is not so much that people will be reluctant to share their experiences in a group setting, as that having several participants competing to tell their individual and detailed stories is likely to produce ‘noise’; that is, data that it is hard to order and attribute to individual speakers. The nature of focus group discussions means that stories are unlikely to unfold sequentially, as they can do in a one-to-one interview, and hence the picture presented will be confusing and attempts to analyze data will be frustrated. Ong (2003) reports on a study of experiences of back pain, where the initial focus group allowed participants to tell their individual stories, with later groups focusing more explicitly on the research questions, suggesting that a series of focus groups may be more appropriate, where the intention is to build up a detailed picture of individuals’ experiences.
Accessing ‘attitudes’
Nor, as we have already seen, are focus groups an appropriate method if you want to measure attitudes. Puchta and Potter (2002) argue that attitudes are the end result of a series of analytic decisions, which suggests that we should be wary of thinking that there is any such thing as an ‘attitude’. They remind us that attitudes are ‘performed’ rather than being ‘pre-formed’ (Puchta and Potter, 2004, p. 27). The implications for the process of analysis and use to which focus group findings can be put are further discussed in Chapter 11.
While marketing researchers tend to focus on using focus group data to make inferences regarding the attitudinal stances or preferences of the wider consumer body, within social science research this is generally not the preferred end product. Nor are results generally required as speedily as with marketing research and there is a venerable survey tradition within the social sciences that serves this requirement much better. If you want to make statistical generalizations from your data, then focus groups are not the method of choice. ‘Focus group samples are usually both unrepresentative and dangerously small’ (Morgan and Krueger, 1993, p. 14).
This reservation also potentially applies to some of the usages made of focus groups by health services researchers – particularly those working alongside clinicians. Some difficult questions are raised by the use, for example, of focus groups in order to access patients viewed as ‘recalcitrant’ or reluctant to take up services, screening programmes, or health promotion advice. Research in this genre is very similar to that carried out under the auspices of organizational research and, in common with this approach, tends to view the research question through a predominantly ‘professional’ lens.
Researchers’ or participants’ concerns?
While the appearance of terms such as ‘non-adherence’ or ‘non-compliance’ can signal research carried out with this partial or partisan professional view, focus group research, again, spans a continuum. Even where the original concern is that of ‘non-adherence’, focus groups can provide valuable explanations, in terms, for example, of identifying competing priorities for patients. Bloor et al. (2001) argue that focus groups are particularly well suited to studying decision-making processes, for example, and the ways in which people weigh up competing priorities or the ways in which they qualify their views to take situational and circumstantial factors into account. Some research uses focus groups in this way in order to elucidate the decision-making processes involved – e.g. non-adherence to osteoporosis medications (Iversen et al., 2011); or adolescents’ use of inhaled asthma controller medication (Wamboldt et al., 2011).
Focus groups have also been used to provide a greater understanding of apparently illogical health-related behaviours, such as neglecting to take folic acid (Barbour et al., 2012) while pregnant. Such studies are characterized by a focus on the importance of lay understandings and take, as their starting point, the notion that apparently illogical beliefs and practices, once viewed from the perspectives of the people involved, are likely to display a coherent and possibly highly sophisticated logic. This, however, only becomes apparent when focus group participants are given scope to justify and expand on their views in a non-judgemental environment.
Some clinician-researchers have built on the findings from focus group studies in order to develop enhanced or more appropriate messages or treatment packages for specific groups of patients. Bryan et al. (2008) used focus groups to adapt ethnically appropriate, information-seeking and recruitment messages for a prostate cancer screening programme for men at high risk. A further example is provided by Bottorff et al. (2014) who designed tailored messages about the link between tobacco exposure and breast cancer for First Nations people and Métis in Canada. Such usages bring to mind the approaches used in marketing research with the aim of producing messages designed for particular target audiences and fall into the orbit of ‘social marketing’. A recent critique of social marketing (Crawshaw, 2012) points out that it consistently fails to take adequate account of the day-to-day lives, priorities and constraints of the individuals and groups being addressed.
The researcher working in this area treads a narrow line between encouragement, persuasion and empowerment, and much depends on the way in which focus groups are carried out, how they are embedded in studies and, importantly, who has commissioned the research and what is done with the findings. Certainly, much of the knowledge underpinning health promotion advice is essentially provisional in nature (with recommendations frequently being withdrawn or even reversed), and health professionals’ firmly held convictions should, perhaps, be subjected to a degree of scrutiny similar to that reserved for exploring the lack of conviction of their patients or clients. (This raises particularly difficult questions for social scientists engaged in health services research with practitioners/clinical colleagues.) Ultimately, it is possibly the extent and nature of consultation with patients in the course of a project that separates the potentially coercive from the potentially empowering in this field of research. Drawing on a model more akin to that espoused by design researchers, Stellefson et al. (2010) produced and proceeded to test and refine prototype materials for chronic obstructive pulmonary disease (COPD) self-management education DVDs for a rural population. While this approach admittedly accepts without question the value of self-management, there is evidence of a genuine attempt to engage with users and to seek to reflect their concerns.
The knowledge of clinicians and other professionals can undoubtedly be enhanced through focus group findings which provide an understanding of the processes involved and which can offer the potential to take account of the findings in future practice. A focus group study of the needs of released prisoners (Luther et al., 2011), for example, focused on identifying the implications for practice in order to limit return to high-risk behaviour in relation to HIV/AIDS.
Moran et al. (2012) provide an interesting example of an approach which successfully combined two orientations – that of focusing on specific practice issues and providing unanticipated insights. They were concerned to establish what patients (with ADHD and autistic spectrum disorders) and carers thought of routine outcome measures used in a Child and Adolescent Mental Health service. Although the emphasis was predominantly on the outcome measures and researchers sought to determine levels of agreement (using kappa coefficients), nevertheless, the discussion identified some broader and potentially useful concerns – interestingly, described as ‘irrelevant’ in terms of the original aims of the research:
One predominant fear was a fear of discharge and subsequent lack of access to support should problems recur in the future. Carers wanted rapid access to advice that they could dip in and out of, and were worried that outcome monitoring would be used to limit access to services. (2012, p. 74)
Similarly,