An examination of indexes for determining the number of clusters in binary data sets

Evgenia Dimitriadou, Sara Dolničar, Andreas Weingessel
  • Psychometrika, March 2002, Springer Science + Business Media
  • DOI: 10.1007/bf02294713

What is it about?

Market segmentation (targeting product offers on carefully chosen segments of consumers or target markets) is the key to organisational success. Good market segmentation decisions can only result from well implemented market segmentation analysis. But being an exploratory analysis, there are many pitfalls that can reduce the validity of the analysis. This study offers guidance on how to avoid one of these pitfalls: selecting the wrong number of clusters or market segments.

Why is it important?

Market segmentation is widely used in academia to create new knowledge and by industry to identify a promising target segment. High quality market segmentation analysis is the key to achieving these two aims.

Perspectives

Professor Sara Dolnicar
University of Queensland

Improving market segmentation methodology has been one of my key research interests since my PhD which I commenced in 1994. In collaboration with two world-class computational statisticians (Friedrich Leisch and Bettina Grun) and our students we have made a few contributions which we hope will reduce the risk of suboptimal market segments resulting from future analyses.

Read Publication

http://dx.doi.org/10.1007/bf02294713

The following have contributed to this page: Professor Sara Dolnicar

In partnership with: