Finch and Mason's (1990: 26) Family Obligations Project was a study of 'patterns of support, aid and assistance... between adult kin' in Manchester. Initially, survey<br>research, using a structured interview, was conducted and yielded nearly 1,000 completed interviews. A sample of these interviewees was then approached to be inter viewed by semi-structured interview. The initial sample for this phase of the investigation was selected purposively, that is, with specific target subgroups in mind. These were divorced and/or remarried people and the youngest group at the time of the survey (18–24 years of age). Their rationale for this purposive selection is as follows: 'Since fieldwork was principally to be concerned with under standing the process of negotiation between relatives, we decided that it would be much more useful to focus upon individuals who might currently or recently have been involved in processes of negotiation and renegotiation of family relationships' (1990: 33).<br>Finch and Mason sampled five at a time from the total of each of these subgroups who were willing to be inter viewed again (112 in the divorced/remarried subgroup and 117 young adults). Individuals were sampled using random numbers. In addition, the authors wanted to interview the kin groups of individuals from the initial social survey as providing examples of 'negotiations between relatives over issues concerning financial or material support' (1990: 38). They decided to conduct two further interviews with the focal person in a negoti ation over family obligations and one interview with each of that person's relatives. However, the sampling strategy was based on the selection not of individuals as cases but of situations. In order to make the data com parable, they searched out individuals and their kin who had been identified in the survey-for example, as having moved back into their parents' home following a divorce. A further element in their sampling strategy was that the authors 'tried to keep an eye on the range of experiences that [they] were studying, and to identify any obvious gaps' (1990: 43). As a result of this ongoing 'stocktaking exercise', as they call it, they identified cer tain gaps in their data: men, because by and large they were the focus of interviews as part of kin networks rather than initial key informants in their own right; un employed people, particularly because of high levels at the time of the research; ethnic minorities, social classes<br>I, IV, and V; widows and widowers; and stepchildren and stepgrandparents. As Finch and Mason's experience shows, the process of theoretical sampling is not only one that gives priority to theoretical significance in sam pling decisions, but is also one that forces researchers to sharpen their reflections on their findings during the fieldwork process.
正在翻譯中..