Interdependence Analysis
Exploratory methods that reduce variables or group cases without a prespecified outcome
Interdependence analysis looks for structure in the data without designating any variable as the outcome. Two canonical approaches are covered:
Method pages
- Exploratory factor analysis – reduces many observed variables to fewer latent factors. Includes a brief pointer to confirmatory factor analysis via
lavaan. - Cluster analysis – groups cases into homogeneous clusters. Covers hierarchical methods (Ward, single, complete, average linkage), k-means, two-step clustering, and validation via the elbow method and silhouette.
Choosing between them
- Reduce variables (many measurements per case that may reflect a smaller set of latent traits) 192 factor analysis.
- Group cases (each case belongs to an unknown subgroup) 192 cluster analysis. Within clustering, pick the algorithm based on scale level and sample size:
- mixed types / small \(n\) 192 hierarchical
- metric / large \(n\) 192 k-means
- mixed types / very large \(n\) 192 two-step
Structure inspired by the University of Zurich Methodenberatung (methodenberatung.uzh.ch). All text, examples, R code, and reporting sentences are independently authored in English.