The Scottish Government

« Previous | Contents | Next »

Listen

CHAPTER TWO METHODOLOGY

Sample design

2.1 A purposive approach to sampling was adopted, wherein study participants were selected on the basis of specific criteria of interest. There were a number of characteristics which might have informed the development of the sample framework, including the Social Inclusion Partnership ( SIP) ² status or urban/rural classification of the survey area. However, since the sample size was constrained by resource issues, there was a limit to the number of factors which could be incorporated into the framework. The findings of the first wave of qualitative interviews suggested that the impact of moving was mediated by age, household composition and distance of move. The intention was therefore to interview approximately equal numbers of respondents from each household type (i.e. Adult, Family and Older ³) who had either moved to a new area or stayed in the same area. Since it would not have been possible to make meaningful comparisons between a further sub-group of urban/rural respondents, it was initially decided that only respondents resident in urban areas should be recruited. Thus the sample framework set out below was devised:

Table 2.1: Qualitative sample target framework

2.2 As it transpired, the issue of selecting participants who had undergone area change was not as straightforward as is represented by the sample framework. The apparently straightforward distinction between moving to a new area and staying in the same area proved more complex in some cases. A number of respondents (6) who reported in the survey that they had moved to a new neighbourhood had in fact returned to an area they knew well from previous times in their life, often having close family and social contacts nearby. One survey respondent who was recorded as moving neighbourhood did not perceive herself as having done so. Thus, only 4 respondents in the qualitative sample had in fact moved to an entirely new area. On this basis, we would not expect the qualitative research to reveal such disruptive effects on social and community outcomes as might have been expected.

Interview schedule

2.3 The interview schedule aimed to capture changes in the broad areas of housing and area change and attendant impacts on health, well-being, social and community outcomes. The full interview schedule can be viewed in Annex 1.

Recruitment

2.4 After receiving ethical approval from the University of Glasgow's Research Committee, recruitment commenced in August 2007. In order to reflect the variety of new housing types to which study participants had been moved, a range of new developments across Glasgow were targeted for recruitment. Batches of 5 to 10 respondents (over 60 in total) were initially contacted by letter, requesting permission to conduct an interview. If the enclosed reply slip was not returned within 2 weeks, and there was an active phone number available, they were then contacted by telephone to request permission. Many respondents did not return the reply slip and did not have an active telephone number and were dropped from the pool of potential participants. A number also refused to participate further, suggesting that perhaps after their relatively lengthy involvement in the study, a degree of 'research fatigue' had set in. However, the majority of the people with whom successful contact was made by phone agreed to participate. Given the initial difficulties in recruitment, a decision was taken to scale the original sample framework downward slightly, from 24 to 22. Since the numbers of older respondents in the survey sample are very small, this downward revision was achieved by decreasing the number of older respondents in the qualitative sample. It was also necessary to expand the geographical area initially targeted, to include non-urban areas and locations outwith Glasgow. The achieved sample framework is set out in the table below.

Table 2.2: Achieved sample

Interviewing

2.5 One-to-one in-depth interviews were conducted between August 2007 and January 2008. The interviews were conducted in respondents' homes, except one which was conducted at the respondent's place of work. Prior to commencing the interview, informed consent was obtained from the respondent, at which time participants were given an opportunity to ask questions about the interview process. A shopping voucher to the value of £10 was provided to thank respondents for their time.

2.6 In the majority of cases, only the respondent was present during the interview. However, on 3 occasions the respondent's partner was present and contributed to the interview. Subject to the participant's permission, the interviews were recorded using digital recording equipment. Permission was granted by all participants. The interviews were then transcribed by a professional transcription company in preparation for analysis. Appropriate data protection procedures were followed at all times.

Analysis

2.7 The software package NVivo 7 was used to conduct a thematic analysis of the data. This involved identifying the overarching themes of interest in the research context, then coding text in the interview transcripts which corresponded to these themes. The coded text was then examined in detail to identify recurring themes and patterns within the data, which were then sub-coded in a further round of analysis. When this process was completed, connections between emergent themes and respondent characteristics were investigated in some depth. Although area change was one of the original sample criteria, as discussed above, it proved to be less relevant than expected. Therefore other emergent factors, such as such as reason for moving and dwelling type transition, were pursued in more depth.

Sample characteristics

2.8 Of the 22 respondents, 19 were female and 3 were male. The majority were resident in the Greater Glasgow area, including Faifley, Paisley and Cumbernauld. One respondent was resident in Whitburn, and one in Blantyre. The respondents had lived in their current accommodation for between 3_ and 5 years. In keeping with the sample framework, 8 respondents belonged to adult households with no children, 8 to family households and 8 to older households. Two of these were single parent families, and 2 were resident with adult children. In terms of employment status, 5 respondents were retired, 5 were sick or disabled, 9 were in full or part-time work, and one was a full-time housewife. The socio-demographic characteristics of the sample are presented below.

Table 2.3: Socio-demographic characteristics of the sample

Limitations

2.9 As with all qualitative research, the primary limitation of this element of the SHARP study is that the findings cannot be generalised to the wider population. A strength of the SHARP study however, is that the mixed method approach allows triangulation between the quantitative and qualitative findings. There are also a number of limitations specific to this study. The method of recruiting respondents from an existing survey sample may lead to selection bias, in that individuals with particular characteristics may be more likely to agree to participate in a further wave of the research. Further, respondents with certain characteristics form the majority of the sample. For instance, those who moved from a flat to a house and those who live in a regeneration area predominate within the sample. Although this is not dissimilar to the quantitative sample distribution, it means that the absolute numbers with certain characteristics in the qualitative sample are so small that conclusions based on the findings are somewhat tentative. The potential implications of these limitations are discussed further in Chapter 6.

« Previous | Contents | Next »