Show or hides a table of the ten terms with the highest scores in each cluster. The score St,c for term t in cluster c is calculated as follows:
where mean(pt) is the mean of the term probabilities by cluster for term t and pt,c is the term probability by cluster for term t in cluster c.
Shows or hides a multidimensional scaling plot, which is a two-dimensional representation of the proximity of the clusters. For more information about MDS plots, see Multidimensional Scaling in the Multivariate Methods book. The Show Text button opens a window that contains the text of the selected documents.
The score formula that is saved uses the Text Score() JSL function with the weighting argument set to “LCA”.
For more information about latent class analysis, see Latent Class Analysis in the Multivariate Methods book.