Applying a mixed methods design to test saturation for qualitative data in health outcomes research

Abstract

Saturation, a core concept in qualitative research, suggests when data collection might end. It is reached when no new relevant information emerges with additional interviews. The aim of this research was to explore whether a mixed methods design could contribute to the demonstration of saturation. Firstly, saturation was conceptualized mathematically using set theory. Secondly, a conversion mixed design was conducted: a set of codes derived from qualitative interviews were quantitized and analyzed using partial least squares (PLS) regression to document whether saturation was reached. A qualitative study conducted by other researchers prior to this work (i.e. none of the present authors was involved in this study) was used to test saturation using PLS regression. This illustrative qualitative study aimed to investigate the impact of Clostridium difficile infection (CDI) on nurses' work in the hospital and the results were published elsewhere (Guillemin et al. 2015). Semi-structured interviews were conducted with 12 nurses. Saturation was characterized by the cumulative percentage of variability accounted for by PLS factors. After 12 interviews, this percentage was 51% which suggests that saturation was achieved at least on main themes. Two main themes identifying similarities in the experience of nurses caring for patients with CDI were identified: Organization/Coordination of the working day and Time-consuming work. Although dependent on the coding of qualitative data, PLS regression of quantitized data from qualitative interviews generated useful information for the determination of saturation.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 50 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.