A Novel Visualization Approach for Multiple-Case Analysis Using Sankey Diagrams

Multiple-case studies are widely applied in biomedical and psychological sciences to investigate complex or rare conditions, yet their findings are often difficult to communicate using tables or conventional statistics. We propose the Sankey diagram as a novel visualization method for multiple-case research, highlighting replication logic across and within cases. A secondary analysis was conducted with patients diagnosed with borderline personality disorder (BPD) and their relatives, assessing attributional biases- personalization and intentionality- alongside BPD symptom severity. Data were categorized and visualized in a three-step Sankey diagram linking attributional biases, individual cases, and symptom severity, with flow widths representing numerical scores. Patients attributed higher intentionality to ambiguous social situations than relatives, and personalization significantly predicted symptom severity. The Sankey diagram made these associations visually accessible and highlighted one case that contrasted with our theoretical expectations, prompting refinement of the interpretation. Compared to traditional tabular formats, the Sankey approach offered a more intuitive and informative representation of both patterns and exceptions. This study demonstrates the potential of innovative visualization tools to enhance theory-building and the interpretation of complex case study data, particularly in psychiatry and psychology, and underscores their relevance for systematic application across multiple case studies..