With regard to ensuring rigour in qualitative inquiry, Spiers (2002:2) asserts that “[w]ithout rigour, research is worthless, becomes fiction, and loses its utility”. Central to assessing the rigour or quality of quantitative research are the criteria of reliability and validity. A number of frameworks have been put forward (and debated) in the qualitative inquiry literature; however, a widely adopted set of criteria has been that proposed by Lincoln and Guba (1985). They have replaced the concepts of reliability and validity with that of trustworthiness. Rather than criteria of internal validity, external validity, reliability and objectivity which are typically used to establish trustworthiness within a quantitative research paradigm, they originally proposed that research using qualitative methods, such as this study, should consider the criteria of credibility, transferability, dependability and confirmability. In response to their critics they later expanded these criteria to include that of authenticity (Guba & Lincoln, 1989; Lincoln & Guba, 2000).
With respect to discourse analysis, Potter and Wetherell (1987) propose four validity criteria:
coherence, participant‟s orientation, new problems and fruitfulness. By 1996, however, Potter had revised his view on the appropriate criteria and while retaining the criterion of coherence had replaced the latter three criteria with deviant case analysis, participant‟s understandings and reader‟s evaluations.
4.8.1 Trustworthiness
According to Lincoln and Guba (1985:290) with respect to trustworthiness the basic issue is simple: How can an inquirer persuade his or her audiences (including self) that the findings of an inquiry are worth paying attention to, worth taking account of? What arguments can be mounted, what criteria invoked, what questions asked, that would be persuasive on this issue?
Using the criteria outlined by Lincoln and Guba (1985) and Potter (1996a), the following discussion will establish the trustworthiness (or rigour) of the study by auditing the events and influences on the research process and my reactions to them. Koch (1994) notes that although the
100 readers may not share the interpretation presented by the researcher they should be able to follow the way in which it was derived. This is a result of the fact that each of us brings to the analysis our own preconceptions that influence the dialogue between researcher and text or the reader and the interpretation. My own prejudices and preconceptions are what will be discussed below.
During the period of this study these initial beliefs were challenged and rescripted; a process that continues as part of the constant dialogue that sustains and creates knowledge.
4.8.2 Credibility
Lincoln and Guba (1985) suggest a number of techniques that make it more likely that credible findings and interpretations will be produced: activities in the field that increase the probability of high credibility, peer debriefing, negative case analysis, referential adequacy and member checks. Of these techniques, four were adopted during this study: activities in the field, peer debriefing, member checks and negative case analysis. The latter criterion has been considered as being synonymous with Potter‟s (1996b) concept of deviant-case analysis.
Activities in the field
With respect to activities in the field that increase the probability of high credibility, Lincoln and Guba (1985) suggest three techniques: prolonged engagement, persistent observation and triangulation. During my eight years of experience as the only black staff member teaching white students in a predominantly white institution and working with all-white colleagues, part of me resisted allowing whiteness to penetrate my skin, however at the same time I felt compelled to prove that I could do equally well or better than my white colleagues. In contrast, for the last four years I have worked for a historically black institution with the majority of black colleagues and students. Within this context I continually find myself defending my black position from those who experience my contributions to knowledge and my accounts on how to do certain things as being assimilated into whiteness. This experience and continuous struggle around my own identity suggest that I am a “stranger in a strange land”. Of course, such prolonged engagement risks the introduction of “distortions based on a priori values and constructions” (Lincoln &
101 Guba, 1985:302). In an attempt to minimise this problem notes were written and kept for referral as new ideas and challenges emerged. These challenged my a priori beliefs such that they came to be perceived as too simplistic in light of the complexity of the issues being explored.
It is, no doubt, the nature of preconceptions to be simplistic; however, there is a danger in adhering to simple tenets in the face of complex and dynamic interacting factors as it may lead to ignoring the change processes that may be occurring and maintaining unwarranted commitment to ethnocentric beliefs. I look back on those early declarations and now perceive them as both right and wrong; each one could be (re)viewed through a different lens leading to agreement or disagreement.
The analysis that I have produced is, therefore, only one of many interpretations that could emerge from a study of black identity construction in contemporary South Africa and I accept that others may disagree with particular aspects of this interpretation. The exploration of the literature and the use of interviews were a form of triangulation, that of the use of different sources of data.
Peer debriefing
This involves exposing the work to a disinterested peer in order to illuminate aspects of the research that might otherwise remain implicit. Lincoln and Guba (1985) argue that this should not be undertaken by those in authority over the doctoral researcher such as members of the research committee. I would argue, however, that in this instance my supervisor and the professors in the School of Psychology were part of that process. During the PhD seminars they constantly spoke of me as the “expert” in this subject and their probing was for elucidation rather than dictates about what should and should not be included.
Peer debriefing was also ongoing through discussions with two other peers who were also engaged in similar PhD projects. My supervisor read the entire work as it progressed and provided written feedback, which gave an opportunity for reflection on the honesty and accuracy of what I was producing. This aspect of trustworthiness is consistent with Potter‟s (1996)
102 criterion of reader‟s evaluations, in which readers are able to make their own evaluations and suggest alternative interpretations.
4.8.3 Transferability
Lincoln and Guba (1985:316) argue that it is not the responsibility of the researcher to “provide an index of transferability”. The responsibility of the researcher lies in providing sufficient contextual data, or “thick description”, such that the reader can make a judgement of transferability. Thus, Sandelowski (1986:27) proposes the notion of fittingness: A study meets the criteria of fittingness when its findings can „fit‟ into contexts outside the study situation and when its audience views its findings as meaningful and applicable in terms of their own experiences. In the presentations held in PhD seminars I have received comments from other black people that they find this meaningful in the context of their own experience and provided an opportunity to view their experiences from the perspective of another. I have since the beginning of this project attended two conferences in South Africa and in Sweden. From the above it can be contended that this work meets the criterion of transferability.
4.8.4 Dependability
One way in which the study can be shown to be dependable is through an audit (Lincoln &
Guba, 1985). As Sandelowski (1995) notes, a study and its findings are auditable when another researcher is able to follow the decision trail used by the investigator in the study. This requires explicit discussion of the theoretical, methodological and analytic choices taken throughout the study. In the present study the findings from earlier writers in the field are contrasted with my findings in a critical way, while acknowledging multiple versions or readings of any text to consider the usefulness of one to the other. Finally, readers‟ evaluation is made possible as a result of the presentation of the transcripts alongside the interpretations being made. As skilful interactants, judgements can be made by the readers about cultural competencies and the wider claims made by the study.
103 4.8.5 Confirmability
According to Lincoln and Guba (1985), confirmability can be achieved as part of the audit to determine dependability; a process that is supported by keeping a reflexive journal. I kept a diary as a reflective tool throughout the journey of this study. I posted insights, questions, issues to explore, matters to return to and so forth. The diary contained (i) personal notes, on which questions, emerging insights and new directions were posted; and (ii) methodological issues which signalled areas for further exploration. It was this process that helped further explore one of the questions in my interview schedule: the question asked respondents about a moment when they felt despair and human degradation. From the first four interviews conducted, I observed that when responding to this question respondents became tense. I then explored this question further and rephrased it so as to ask what that event said about who they were.
4.8.6 Authenticity
Lincoln and Guba (1989) proposed five authenticity criteria: fairness, ontological authenticity, educative authenticity, catalytic authenticity and tactical authenticity. As the authors themselves note, these criteria have not been received without challenge (Lincoln & Guba, 2000). It is difficult to assess to what extent this study has met the hallmark of authenticity, as the last four criteria focus on the ability of the study to have social and political impact. Ontological and educative authenticity respectively relate to the raised awareness of both the research respondents and those who surround them and the extent to which the research helps those involved become aware of one another‟s perspectives. Authenticities are concerned with the ability of an inquiry to prompt action on the part of the research respondents and the involvement of the researcher in training respondents in social and political action if wanted by the respondents. Responses such as “I had not thought about that” or “I want to get back to you about that” or “I wish you had more time to discuss these issues” from some of the respondents during the interviews could be read as the first steps in raising the level of awareness with respect to some of the issues. While it is to be hoped that this study has generated new awareness and action amongst the respondents; this would appear to be a response that is better evaluated
104 from a future perspective. For example, some of the young adults appreciated the fact that they were interviewed and perceived it as an opportunity to think and talk about who they are for the first time. As Bryman (2001) notes, the authenticity criteria have generally not been as influential as the criteria of credibility, transferability, dependability and confirmability and it may be that the criteria to determine authenticity are better associated with a specific form of naturalistic inquiry such as action research. According to Lincoln and Guba (2000:180) the criterion of fairness is about balance, that “all stakeholders‟ views, perspectives, claims, concerns, and voices should be apparent in the text”. Thus, the evaluative focus is turned back on to the researcher as the key research instrument (Carpenter & Hammell, 2000). By my earlier declarations of my positioning, both biographical and philosophical, the reader has some measures with which to assess the extent to which my own positions have affected my engagement with the subject matter, data collection and analysis. The best I can argue is that through the reflexive process I have striven to ensure that the findings have emerged from the data and not from my own positioning.
Coherence
The final consideration with respect to rigour in this work is that of coherence. To what extent does this study draw upon previous work and provide a check of the adequacy of previous studies? Given the proviso that this work is only one of the possible readings or explanations of black identity construction in South Africa, I would argue that it demonstrates coherence by building on (and incorporating) the work of others in subject areas related to ambivalence and black identity. It has challenged the adequacy of previous explanations of the universality of black identity that do not account for the complexity of being black in South Africa today.