The objective of K-means clustering is ultimately to assign each observation (row) of the input data to its “closest” cluster. There are multiple ways that distance can be calculated, but regardless of the metric chosen, the result of K-means clustering is that each observation (row) is assigned to a single cluster, and the center of that cluster can be used to define its location.

This tab of results reports the scaled variables (columns) used in the analysis. Appended to these columns is an additional column for each different number of clusters specified to use when grouping the data. For example, if the analysis attempted to group your data into 3, 4, 5, or six clusters separately, there would be four additional variables appended to this table. The values of these columns represent the distance from the observation (row) to its assigned cluster center. Cluster assignments can be found on the Clustered data tab of the results, while the location of individual cluster centers can be found on the Clusters details tab of the results.

Note that the within cluster sum of squares (WCSS) is a measure of variance that uses these distance values in its calculation. For a given cluster, the WCSS is equal to the sum of the squared distances of all points assigned to that cluster from the cluster center. Clusters with lower WCSS are more compact (assigned points are closer to the cluster center) while clusters with larger WCSS are more spread out.