A fuzzy clustering approach to evaluate individual competencies from REFLEX data

ABSTRACT

We empirically illustrate how concepts and methods involved in a grade of membership (GoM) analysis can be used to sort individuals by competence. Our study relies on a data set compiled from the international survey on higher education graduates called REFLEX. We focus on the subset of data related to the perception of own competencies. It is first decomposed into fuzzy clusters that form a hierarchical fuzzy partition. Then, we calculate a scalar measure of competencies for each fuzzy cluster, and subsequently use the individual GoM scores to combine cluster-based competencies to position individuals on a scale from 0 to 1.

KEYWORDS:

• Fuzzy clustering
• grade of membership model
• classification
• competence
• higher education
• REFLEX project

AMS SUBJECT CLASSIFICATION:

• 62-07
• 62P25

Acknowledgements

The author would like to thank Timo Huijgen, from the Research Centre for Education and the Labour Market, Maastricht University, ROA, for providing REFLEX project data. He also express his gratitute to two anonymous referees for their suggestions and careful reading of the earlier version of this manuscript.

Disclosure statement

No potential conflict of interest was reported by the author.

Notes

1 The data file is available for scientific research purposes (see [48]).

2 We used $\ln \left(\right)$ instead of ${\log }_2\left(\right)$, and adapted the formula for the fuzzy case.

3 Except UK.