Term: R^{ 2}
Last Updated: 2011-02-22
In traditional statistics, R^{ 2} values summarize the percent of variation shared between two measures. Similarly, in cluster analysis, they indicate how much of the variation in predictor variables exists within versus between cluster analysis groups. In cluster analysis, R^{ 2} values are greatest when the variation between groups far exceeds that within groups (i.e., people within each group are very similar but collectively the groups are very different).