A Review of Qualitative Data Analysis Practices in Health Education and Health Behavior Research.

Abstract

Data analysis is one of the most important, yet least understood, stages of the qualitative research process. Through rigorous analysis, data can illuminate the complexity of human behavior, inform interventions, and give voice to people's lived experiences. While significant progress has been made in advancing the rigor of qualitative analysis, the process often remains nebulous. To better understand how our field conducts and reports qualitative analysis, we reviewed qualitative articles published in Health Education & Behavior between 2000 and 2015. Two independent reviewers abstracted information in the following categories: data management software, coding approach, analytic approach, indicators of trustworthiness, and reflexivity. Of the 48 ( n = 48) articles identified, the majority ( n = 31) reported using qualitative software to manage data. Double-coding transcripts was the most common coding method ( n = 23); however, nearly one third of articles did not clearly describe the coding approach. Although the terminology used to describe the analytic process varied widely, we identified four overarching trajectories common to most articles ( n = 37). Trajectories differed in their use of inductive and deductive coding approaches, formal coding templates, and rounds or levels of coding. Trajectories culminated in the iterative review of coded data to identify emergent themes. Few articles explicitly discussed trustworthiness or reflexivity. Member checks ( n = 9), triangulation of methods ( n = 8), and peer debriefing ( n = 7) were the most common procedures. Variation in the type and depth of information provided poses challenges to assessing quality and enabling replication. Greater transparency and more intentional application of diverse analytic methods can advance the rigor and impact of qualitative research in our field.