Juxtaposing clustering techniques in framing analysis

Abstract

We assess and map the difference between two frame extraction techniques in media research. Framing scholars argue that latent class analysis (LCA) is the “most advanced clustering technique” that gives a clear criteria for the number of clusters of frame elements in the analysis. In this work, however, we show that the community detection method utilizing modularity optimization algorithm in complex networks gives a more precise and reliable number of clusters in framing analysis.