Creating Cluster Assignment Variables in the Data Source

Once the analysis is complete, it is often desirable to assign respondents to the identified clusters to utilize the segmentations in further analysis. In our example, we’ve identified three clusters which can be used in further analysis to understand and classify the respondents.

To create the cluster assignments on the data source:

1. Click on the Save Cluster Assignment button in the top dialog.

The Save Cluster Assignments overlay opens .



Figure 1 - The Save Cluster Assignments overlay

2. Select the appropriate calculation run that you wish to use for the basis of the variable creation, and then input the variable name and the name that you wish to label each cluster.

To complete the update of the data source, a BitStream update needs to be performed. This can be selected when creating the cluster assignment variable.

Once the generation task has been completed, this will create a new single categorical variable available in the Data Source toolbox. Under the Cluster assignments folder, a new sub-folder will be created for each cluster assignment created. This folder will include the name of the cluster analysis object, and the calculation run name. The cluster variable will be present in sub-folders as a single categorical variable, with each answer category representing a defined cluster .



Figure 2 - Example of Cluster assignment variables in the Data Source toolbox

Using Cluster Assignments in Subsequent Analysis

As mentioned earlier, these cluster categories can then be used for additional analysis purposes, for example in aggregate tables table and charts . in parameters and filter/segment expressions.

Note: Cluster analysis variables are stored in the BitStream files only, so cannot be used in hit lists or verbatim tables.



Figure 3 - Example of using Cluster assignments in an aggregate table