Department of Psychology, School of Human Sciences University of Surrey, GuildfordSurrey GU2 7XH, UK
Dear Sir,
We welcome the letter from Schilling and Conrad concerning our qualitative study of identity among adults conceived by donor insemination (DI) (Turner and Coyle, 2000). We hope to encourage the use of qualitative research in various scientific domains. We are therefore glad to have the opportunity of addressing the concerns raised by Schilling and Conrad, elaborating the approach we took and thereby helping readers to develop a better understanding of some common principles in qualitative research. We also wish to draw attention to some important research questions arising from our work that have been perceptively summarized by Schilling and Conrad and which represent a useful agenda for future research on the topic.
Schilling and Conrad's greatest concerns seem to arise from the nature of our sample. We therefore need to state clearly that, in common with most qualitative studies, our research does not claim to have obtained data from a representative sample of people who were conceived through donor insemination (DI). Many qualitative approaches are concerned with the in-depth analysis of the accounts and experiences of a small number of participants and any conclusions are specific to that group. In the discussion section of our paper (p.2048), we pointed out that any conclusions drawn from our study must necessarily remain tentative until evidence is provided by other complementary studies to support their general applicability. Some of Schilling and Conrad's points (including their concern about the `unsubstantiated' nature of our claim that non-disclosure of DI can lead to `psychological damage') seem to arise from a failure to appreciate the tentativeness of these conclusions. We also explicitly described the process by which qualitative research advances knowledge through a series of detailed, small-scale, complementary studies rather than through broader, large-scale studies with representative samples. An additional consideration here is that it is impossible to determine what constitutes a representative sample of people who have been conceived by DI because we do not know the parameters of this population. In such cases, it is usual to aim for a sample that is heterogeneous along key dimensions relevant to the research questions [a strategy used by others (Davies et al., 1993), in sampling another population whose parameters were unknown].
One point that Schilling and Conrad raise concerns our inability to indicate what proportion of those who saw the letter advertising the research subsequently took part. We cannot do this as we omitted to ask the convenors of the donor offspring networks that we contacted to keep a record of how many network members received the letter. However, we remain unsure about why it is important to know the response rate. The important thing in qualitative research is to provide a full description of those who took part so that future research can obtain data from individuals with different perspectives, with the aim of progressively building up a complete picture of the research area. We provided this information in our article (p.2044). We therefore feel confident that we have treated the sampling issue appropriately and in accordance with the criteria for good qualitative research (Elliott et al., 1999; Yardley, 2000
). Concerns about `the selectivity of the data' thus seem misplaced.
Schilling and Conrad express the wish that we had provided more information about the disclosure scenarios encountered by our participants. As we noted in our article, there were no commonalities about the disclosure scenarios and so we opted not to report this data as we did not feel that even tentative conclusions could have been drawn from them. However, we concede that these data might have helped to contextualize more fully our participants' responses to disclosure.
Schilling and Conrad raise a frequently presented concern about qualitative research when, pointing to the emphasis on identity problems in our analysis, they ask if we only found what we were looking for in our data. There are two points to be made here. Firstly, the data were analysed by two researchers, the aim being that one researcher would identify and compensate for the analytical `blind spots' and idiosyncratic interpretations of the other. The analysis was also checked by anonymous reviewers during the peer review process prior to publication. Secondly, in some qualitative approaches, researchers are required actively to seek out data which will contradict or at least modify any emerging interpretation. We consistently used this strategy in our analysis. Nor was it the case that our interview schedule focused only on potential problems. We took care to ensure that it permitted the reporting of a wide range of experiences, including the possibility that participants may have found that their experiences acted as a stimulus or resource for personal development and insight (something which is reported in the analysis on pp.20456). If our analysis appeared overly problem-focused, this was only because our participants' accounts were mostly problem-focused.
Regarding Schilling and Conrad's point about the need to differentiate between adopted children and children who were conceived through donor insemination, we did acknowledge this on pp.20412, as they concede. Nevertheless, in the absence of previous research on identity among people conceived through DI, we considered that the degree of overlap on some issues was sufficient to allow us to use the adoption literature to give some substantive shape to our interview schedule. It is worth noting that, although Schilling and Conrad say that people who were adopted often report a sense of having been abandoned by their birth-parents while people conceived through DI were obviously wanted by their parents, some DI offspring in our study felt that they had been abandoned by their donor fathers and by the medical profession.
Schilling and Conrad's comments about the social fathers of DI children are interesting. Our research focus was on the identity-related phenomenologies of individuals conceived through DI but there is much further research to be conducted on these people's broader family contextsa project to which Schilling and Conrad have already made a valuable contribution. Their closing sentencetogether with their earlier questions and hypotheses about disclosure scenariosrepresents a useful crystallisation of some key research questions that flow from our study. Our work represented an initial, general foray into a largely uncharted field. We hope that other researchers will now attempt to gather data (including, perhaps, quantitative and psychometric data) which help to answer the important questions that Schilling and Conrad pose and which can be used to inform best practice in this field.
Notes
1 To whom correspondence should be addressed
References
Davies, P.M., Hickson, F.C.I., Weatherburn, P. and Hunt, A.J. (1993) Sex, Gay Men and AIDS. Falmer, London.
Elliott, R., Fischer, C.T. and Rennie, D.L. (1999) Evolving guidelines for publication of qualitative research studies in psychology and related fields. Br. J. Clin. Psychol., 38, 215229.[ISI][Medline]
Turner, A.J. and Coyle, A. (2000) What does it mean to be a donor offspring? The identity experiences of adults conceived by donor insemination and the implications for counselling and therapy. Hum. Reprod., 15, 20412051.
Yardley, L. (2000) Dilemmas in qualitative health research. Psychology and Health, 15, 215228.[ISI][Medline]