When we conduct a research, there is a need to determine how far the research findings are believable, accurate and useful (Creswell 1998). Traditionally, particularly in quantitative research, the research reliability and validity is assessed to justify its quality. However, can assessing reliability and validity of qualitative research like narrative interview and narrative analysis be possible? Endeavouring to ascertain the rigour of the qualitative research, some scholars propose the criteria to evaluate its reliability and validity through trustworthiness of the procedures and data generated (Stiles 1993, Lincoln and Guba 1985; Wallendorf and Belk 1989). To establish research trustworthiness, they suggest several research techniques such as profound engagement, researcher’s reflexivity, respondent’s feedback, triangulation, or independent audit (Lincoln and Guba 1985; Wallendorf and Belk 1989).
In narrative interview and narrative analysis, its trustworthiness depends on the eminence of its interview and analysis procedure as well as its features such as the respondents, the researcher/interviewer, the interview questions, the transcriptions and the analysis. It is vital to choose the right respondents – the ones who embody narratives of experience which can answer the research questions. As the interview is obtained from the respondent’s own words, it is considered valid (Appleton 1995). However, distortion can arise through the interview and analysis process. The researcher and interviewer must have enough familiarity with the interview data to be able to see it from the respondent’s perspective. The researcher’s profound engagement with the respondent’s socio-cultural background (e.g., understanding the language) increases the likelihood of attaining the “experience-near (1)” (Geertz 1973) – the insider’s view or the respondent’s perspective of reality. Nonetheless, the researcher who is familiar with the respondent’s socio-cultural background may overlook certain nuances and ambiguities of data. Therefore, it is also important to maintain the “experience-far” (Geertz 1973) – the outsider’s view, so that we hold an ability to perceive the phenomena studied with the naïve ears (not the objective ears) that will not take things for granted (Wallendorf and Belk 1989).
While reliability and validity of quantitative research puts emphasis on the researcher’s objectivity, trustworthiness of qualitative research like narrative analysis ironically embraces subjectivity. Apparently, endeavouring to be ‘objective’ may limit our ability to grasp the complexity of human experience (Agar 1980). Willis (1980, p. 91) explains, “If we wish to represent the subjective meanings, feelings and cultures of others, it is not possible to extend to them less than we know of ourselves. The ‘object’ of our inquiry is in fact, of course, a subject and has to be understood and presented in the same mode as the researcher’s own subjectivity - this is the true meaning of ‘validity’ in the ‘qualitative’ zone.” To manifest this issue, the researcher should acknowledge his/her subjective judgement through the process of ‘reflexivity’ (Wallendorf and Belk 1989, Willis 1980).
Although high-quality research often aims to get beyond the respondents’ own understandings of their experiences, trustworthiness of narrative analysis can be enhanced by their feedback (Lincoln and Guba 1985, Wallendorf and Belk 1989). This refers to the practice of researcher sharing interpretations with the respondents. By this, the respondents can check, amend and provide feedback as to whether the interpretations are well-substantiated accounts consistent with their experience (Bryman 2001). Basically, the narrative analysis process should involve democratic interactions between the respondents and the researcher in order to attain some sorts of mutual dialectical creation of meanings (Hirschman and Holbrook 1992). Nevertheless, as we deem the narratives as socially constructed, we should allow multiple and contradictory interpretations to emerge (Atkinson 1990, Ricoeur 1976). In fact, in order to grasp the complexity and inconsistency of the texts, we are encouraged to strive for “conscious pluralism” in pursuing our research (Morgan 1983). That is, we should undertake the interpretive process that “does not assume any one answer to explaining consumer behaviour, on one single solution, but approaches consumer culture expecting to find multiple meanings and a rich construction of reality and illusion beyond the merely rational” (Elliott 1999, p. 121).
Triangulation is also another technique to enhance research trustworthiness (Lincoln and Guba 1985, Wallendorf and Belk 1989). Triangulation represents the combination of two or more data sources, methods or researchers. Wallendorf and Belk (1989) encourage triangulation since it not only enhance the research credibility but also generate a multiplicity of perspectives on the behaviours and contexts of the phenomena. Thus, adopting various modes of narrative analysis is recommended. Nietzsche (1967) asserts the more perspectives the narrative
The Journal of American Academy of Business, Cambridge * Vol. 18 * Num. 1 * September 2012 135
analysis, the richer and deeper the interpretations and knowledge. Triangulation across researchers would enrich the understandings of the phenomena being studied as each researcher brings his/her viewpoint to extend the interpretations. Holt (1991) also acknowledges that triangulation can improve the quality of the interpretation and thus attain increased trustworthiness in the eye of the reader. Indeed, the concept of ‘trustworthiness’ is socially constructed and hence its justification depends on the interpretation shared within a particular interpretive community.
Lastly, the narrative research is trustworthy when its interview and analysis procedure as well as its data can be audited and verified by other researchers (Lincoln and Guba 1985, Wallendorf and Belk 1989). Hence, keeping journals on decisions made throughout the research process provides an opportunity for the project’s assessment. Additionally, central to the research assessment is the accuracy in interview recording and transcribing. Qualitative data analysis software such as NUD●IST may be used to standardise interview transcripts. However, Burton (2000) cautions that an over-emphasis on standardisation can separate the data from its context so much that it almost becomes meaningless.