Investigating the number of clusters

To help identify different remote working groups, the data scientist in your team has calculated several dimension scores. These are questions from the remote working survey that were combined into one score.

If an employee has a positive score (e.g. + 1.5), they feel very positive about this aspect. On the other hand, if they have a negative score (e.g. -1.5), they feel very negative. The higher/lower the score, the more positive or negative they are, respectively.

• Social interaction: whether the employee can collaborate well and have enough interaction with colleagues while working remotely.
• Skills: whether the employee feels they have learned the right skills for remote working.
• Support: whether the employee feels management and the company support remote working.
• Well-being: whether the employee feels better while working remotely, physically and mentally.

Using the explorable, you can try out different combinations of dimensions and try out different numbers of clusters. Each point in the plot represents one of the employees and their score for each of the visualized dimensions.

The data scientist and HR expert have determined that three clusters is the optimal solution in this case.