Group discussions

Th e group discussion data consisted of four group discussions with six participants each, including myself in the capacity of moderator. Th e research participants were recruited by convenience and purposive sampling. First, I recruited participants to whom I had easy access (members of an art club that I attended) and asked them to recommend other suitable participants.

On the basis of my experiences with the fi rst group, I contacted the manager

of a local service facility and with her help recruited more participants.

Th e discussions were held in the city of Tampere in southern Finland in 2000. Th e participants ranged in age from 71 to 86 years. Each group had fi ve discussants (11 women, 9 men). Lasting from 60 to 90 minutes, the discussions were audio-recorded with the consent of each participant and transcribed word for word (~130 pages).

Transcription is not simply a technical procedure that takes place before the analysis proper, but it is always part of the analysis itself (Silverman 1993, 117, 2005, 163, Nikander 2008b). Also, judgement and interpretation is involved in decisions about the style and detail of transcription. Th erefore transcripts can never be perfect or fi nal, but the style of transcription needs to be adapted according to the research task at hand (Kvale 1996, 170–172, Silverman 2005, 163). Based on my analyses of the biographical data and my growing interest in the interactive nature of the focus group data, I felt it was important to have a more nuanced transcription of the interactive elements of the group discussion data, i.e. to notate such elements as emphasis, overlapping talk, pauses, laughter or particularly quiet talk. In this work I followed Silverman’s (1993) recommendations. Background information (age in years, education and occupation at the time of retirement, marital status, housing, health status) was collected after the discussions by self-report questionnaires.

Th e discussion groups consisted of a mixture of female and male discussants who diff ered in terms of their health status, marital status, education and housing situation. Group 1 had 5 male participants, all of whom attended the same art group, and the discussion was held at the place where that group used to meet. Th e participants of the three other groups were resident and non-resident clients of a service facility, and these groups met on the premises of the facility. Th is facility is a municipal centre that provides services for retired people aged 60 or over. Service users pay a small service fee, but most of the funding comes from the local council. Services provided include meals, physiotherapy, barbering and hairdressing, library and internet access, as well as various leisure and recreational activities. Th e centre provides housing services for older people who are unable to cope with everyday activities in their own homes, but who do not require care in a nursing home or hospital. Group 2 consisted of 2 men and 3 women (two married couples and one single woman, non-residents), all of whom belonged to the service centre’s literature circle. Group 3 consisted of 5 women (two residents and three non-residents). Group 4 consisted of 2 men and 3 women (residents). Groups 1 and 2 represent pre-existing groups, but in groups 3 and 4, too, the participants knew one another at least by sight, and some of them were friends. Th e groups were thus rather heterogeneous, which fi tted in well with the original goals of reaching a wide variety of

people and tracing diff erent perspectives and experiences. Th e discussions produced rich data with varying perspectives, but similar aspects and patterns of talk did emerge as the presentation of data analysis and the extracts seek to disclose.

To initiate the discussions, I introduced myself and described the research project. I said I was interested in the participants’ own views about health in general and their own health in particular, and also in what health means to them in their everyday lives. Th e discussions then proceeded according to a set agenda, although largely on each group’s own terms. Th is procedure provided a common basis for the discussions, but allowed diff erent views to emerge within and between the groups.

Th e discussions focused on the following topics:

defi nitions of health, personal and general

•

the individual’s own responsibility for health

•

to what extent can health be infl uenced by individual action

•

the most important things in one’s own health, and possible future

•

concerns about one’s own health

the relationship between age and health

•

Th e relationship between age and health was on the topic list, but it was also raised in the discussions without prompting. As a result, the discussions in each group were framed by ageing and old age. Th ese topics were covered in all groups, although the weight they received diff ered from group to group.

Although the main concern was with issues of health, other topics were also raised as the discussions unfolded. Th e exact wordings of the questions varied slightly between the diff erent groups since the questions were fi ne-tuned according to the interaction and atmosphere in each group. Questions were not addressed to individual discussants, but to the group.

My approach to the data and my own role within that data was guided by the idea that the interviewer or moderator is never an impartial or neutral participant, but interviews and group discussions are situations of active interaction (Gubrium 1993a, Holstein and Gubrium 1995). It is important therefore that the interviewer is conscious of his or her role in interaction rather than trying to stand back and “blot out” his or her infl uence. My position in these group discussions was a dual one. I acted as a moderator with the task of making sure that all the participants were given the chance to express their views. On the other hand, I was an active participant in the discussions, aiming to encourage discussion. Some of the questions were designed to address potentially sensitive topics such as responsibility for health. After presenting these questions I made sure to give the participants

ample time to develop their views and not to curtail or control the direction that the discussion took.

Th e group discussions in this research could be called focus groups in that they were “focused” on health issues, and I as moderator steered the talk in these groups by my questions. Th e reason I chose to collect this data in the form of group discussions was that, as Barbour and Kitzinger (1999) say, they can provide a way of exploring the processes and formation of shared beliefs and values, people’s experiences, opinions, wishes and concerns as well as ambiguities and uncertainties (ibid., see also Bloor, Frankland, Th omas and Robson 2001, Waterton and Wynne 1999). Focus groups can illuminate “the normative understandings” (Bloor et al. 2001) that are part of people’s shared stock of knowledge, which is often taken to be so self-evident that it may go completely unnoticed (ibid., 4). Focus groups diff er from group interviews in that questions and prompts are addressed to the whole group instead of asking questions of each person in turn (Barbour and Kitzinger 1999, 4–5). One of the benefi ts of having group members is that group interaction can off er an opportunity to study how diff erent views are brought forward, formulated and reformulated and debated with other group members (Wilkinson 1999, 67, see also Myers and Macnaghten 1999). In addition, it has often been argued that group discussions off er a more equal situation for discussion since the situation reduces the researcher’s infl uence and may give more room to the participants’ views (Wilkinson 1999, 70).

Th e transcribed text was loaded into Atlas.ti, a qualitative data software package used in the preliminary stage of analysis to help organize the data and to gain a systematic overview of the content. Th e data were coded by constructing coding categories such as defi nitions of age and old age, defi nitions of I (as a particular kind of person, e.g. “writer”), defi nitions of other people, defi nitions of oneself or other people as old, health talk (concerning one’s own or others’ health), talk about diff erent activities and interests (other than health) and relations with other people. Also, codes were created to list linguistic and rhetorical tools such as extreme case formulations, emphasized talk, humour and jokes, use of detailed lists, stories, use of outside witnesses and the context in which they were used. In order to explore how diff erent categories, defi nitions and activities were used to position oneself or other people as a particular kind of person or persons, talk was analysed in more detail by means of discourse analysis as discussed in Chapter 4.