Welcome to the UCREL Clustertool

Clustertool allows you to perform Hierarchical Agglomerative Cluster Analysis on your own data via this webpage.

Enter the data to be clustered (click for instructions):

The following options allow you to tweak the clustering procedure. If you are not familiar with the theory of cluster analsyis, it is probably OK to go with the default options.

Select a preprocessing technique to standardise your data:

Select a proximity statistic:

Select a linkage method (for clustering):

Happy with the options above? then

The format of the data that you paste in should be as follows:

The data should be in the form of a table.
On each row, the columns must be separated by TAB characters.
Every row MUST have the exact same number of columns.
Each column represents a single variable in your data.
The data in the columns should be numeric.
Each row represents a single "data object", i.e. something in your data that has values for variables.

Often, it is easy to create a table like this in Excel and then copy-paste it across. Here is an example table:

In this table, there are two variables (number of first person pronouns and second person pronouns). The data objects are separate texts, indicated by three-letter labels.

The rules for column and row labels are as follows:

You MUST have a header row with labels for the columns.
You MUST ALSO put labels for the data objects in the first column (they will appear in the resulting tree).
Column and row labels must be a single word with no spaces or punctuation marks inside them.