Establishing Methodological Coherence when Using Observations, Stimulated Recall Interviews and Narrative Analysis
Qualitative research is multimodal in focus (Flick, 2014), often to triangulate data findings and to add rigor, complexity, richness and depth to the data analysis. However, critics often question the trustworthiness of data findings. Pre-observation interviews, videoed observations of clinical sessions followed by stimulated-recall interviews were chosen as the data methods for the eight participants in the study. Narrative analysis was chosen as the analytical strategy as it served as a data management mechanism to manage the variety of data sources and the volume of data in order to produce a coherent whole to facilitate the first level of analysis. It is well documented that videoed observations and stimulated recall interviews are compatible in their pairing as research methods however (Rowe, 2009), the compatibility between these research methods and narrative analysis needed to be established. The pre-observation interviews, video observations and the stimulated recall interviews were transcribed and coded. However, a further layering on the methodological and data analysis process was needed. This gap was filled by the use of narrative analysis in the construction of narratives (stories) using the data from the pre-observation interviews, video observations and the stimulated recall interviews.
Techniques from creative short story writing were used in developing the narratives which underwent several revisions. Verbatim extracts from transcripts were used to further authenticate the narratives and to enhance the internal validity of the analysis, nevertheless, the construction of the narrative was an interpretive act which meant that different levels of interpretation were used for different data sources. The constructs of trustworthiness (Guba, 1981), were used to allude to the coherence between stimulated recall interviews and narrative analysis. Credibility was achieved through triangulation of the data whereby multiple sources were gathered, i.e. pre-observation interviews, Video recorded observation data and stimulated recall interviews. Transferability was achieved through the narrativising of the data warranted thick descriptions of the empirical data that were presented in the narratives to allude the reader to the context. Dependability of the study was achieved through detailed data production planning and transparency of process. Confirmability was established through triangulation. Triangulation of the data was conducted in an effort to reduce the effect of researcher bias (Shenton, 2004). Although qualitative research is inherently multimodal in focus, the trustworthiness and the coherence of the methods used in a single study needs to be established so that the findings can be dependable and generative. This paper argues that videoed observations, stimulated recall interviews and narrative analysis form a triangulated process to establish methodological coherence and trustworthiness in qualitative research.