Correspondence and canonical analysis of relational data

摘要: Correspondence analysis, a data analytic technique used to study two‐way cross‐classifications, is applied social relational data. Such are frequently termed “sociometric” or “network” The method allows one model forms of and types empirical relationships not easily analyzed using either standard network methods common scaling clustering techniques. In particular, correspondence analysis model: —two‐mode networks (rows columns sociomatrix refer different objects) —valued relations (e.g. counts, ratings, frequencies). general, the provides scale values for row column units, visual presentation among rows columns, criteria assessing “dimensionality” graphical complexity goodness‐of‐fit particular models. has recently been subject research by Goodman, Haberman, Gilula, who have their approach problem “canonical analy...