Description
This technical paper summarises the exploratory quantitative analysis undertaken to investigate how vocational education and training (VET) students cluster and segment in the Australian VET market. The paper outlines the clustering algorithms used and provides insights into the identified market segments, with case studies used to explore key segments (students in targeted English programs; students in social inclusion programs; and migrant students) in more detail.
Summary
About the research
This paper summarises the exploratory quantitative analysis undertaken to investigate how vocational education and training (VET) students cluster and segment in the Australian VET market. This analysis is outlined in three sections:
- The first section focuses on ‘clustering’ as a technique for grouping data and the three clustering algorithms used. These are then discussed in more detail to provide some insights into how they operate. Their specific data requirements, along with their strengths and weaknesses, are also considered.
- In the next section the outputs of the clustering approaches are considered. The resultant clusters are examined to better understand them, and meaningfully label and group them into segments.
- With the insights gained from the clustering process, the final section of this paper returns to the raw data. This step was necessary to further explore (in this case, only some of) the identified market segments. Here three key market segments are explored: students in targeted English programs; students in social inclusion programs; and migrant students.
Key messages
- Two of the three clustering algorithms (k-means and agglomerative) were applied to the total VET activity (TVA) data.
- After considering the output across these two clustering algorithms, several segments within the Australian VET market were identified:
- targeted English programs/students
- overseas students (studying in Australia)
- younger students (includes VET in Schools programs)
- migrants
- social inclusion programs/students
- jurisdictional priorities
- program enrolments not elsewhere identified (NEI)
- subject only enrolments NEI.
- The VET system collects largely categorical variables — with different levels of consistency and completeness — for millions of students, programs and subjects. As a result, it is not well suited to the application of clustering algorithms. Despite this, two clustering algorithms (k-means and agglomerative) were applied to the data, with a third (DBSCAN) unable to be applied successfully.
- While clustering algorithms can carve a dataset into clusters, identifying something that is meaningful to practitioners in a way that explains the clusters is not always guaranteed. Sometimes it can be challenging to bring a useful human perspective or narrative to the clustered outputs. The approach taken in this paper was to look at the features in each cluster that were overly represented compared with all students.
- The algorithms applied assumed single cluster membership to the exclusion of all others. This is an analytically useful (but unrealistic) simplification. In real life, the identified market segments are not mutually exclusive, and students may belong to more than one segment.
- The research approach was unable to conclusively use the clustering outputs to determine whether the identified clusters align with, or bring insights to, the other typologies for segmenting the Australian VET market that can be found in academic literature.