Network Analysis of Case Study Petra S.: A Mixed-methods Approach

Three waves of Petra’s interviews (her case study appears in Chapter 11) have been coded with the recently developed content coding scheme that applies over 150 prominent codes to describe each interview. These content codes are subject to quantitative analysis and visualization using the mathematical tools provided by network analysis. For each interview, content codes form a directed network of adjacent connections among the codes, and the edge weights reflect the frequency of each connection. In the analysis, node and network level statistics of centrality, connectivity, spread, subgroups, and homophily are offered to illustrate the node importance and various network structures. Inferential statistics use random graph modeling to test whether any of the network structures significantly differs from randomness such that they convey important information about how different codes connect with each other. Visualization of the trimmed networks provide further aid to understand the empirical connections of content codes. These quantitative analyses are enriched with interpretations of Petra’s life stories, thus providing both a panoramic view of the structure of her interviews, and a high-resolution view of some of their details. The use of network analysis to understand the structure of a qualitative interview opens doors to an array of mixed-methods research possibilities with the Faith Development Interview data and qualitative data alike.