T20 cluster analysis
- 1. Cluster Analysis By Rama Krishna Kompella
- 2. Introduction • In order to go from data to information, to knowledge and to wisdom, we need to reduce the complexity of the data. • Complexity can be reduced on – case level : cluster analysis – on variable level: factor analysis
- 3. Cluster Anlaysis • Cluster analysis is another interdependence multivariate method • As the name implies, the basic purpose of cluster analysis is to classify or segment objects (e.g., customers, products, market areas) into groups so that objects within each group are similar to one another on a variety of variables. • Cluster analysis seeks to classify segments or objects such that there will be as much likeness within segments and as much difference between segments as possible.
- 5. What cluster analysis does • Cluster analysis – generate groups which are similar – homogeneous within the group and as much as possible heterogeneous to other groups – data consists usually of objects or persons – segmentation based on more than two variables
- 6. • Examples for datasets used for cluster analysis: – socio-economic criteria: income, education, profession, age, number of children, size of city of residence .... – psychographic criteria: interest, life style, motivation, values, involvement – criteria linked to the buying behaviour: price range, type of media used, intensity of use, choice of retail outlet, fidelity, buyer/non-buyer, buying intensity
- 7. • Proximity Measures – Proximity measures are used to represent the nearness of two objects – relate objects with a high similarity to the same cluster and objects with low similarity to different clusters – differentiation of nominal-scaled and metric- scaled variables – the calculation of the distances measures is the basis of the cluster analysis
- 8. • Two phases: – Forming of clusters by the chosen data set – resulting in a new variable that identifies cluster members among the cases – Description of clusters by re-crossing with the data
- 9. Cluster Algorithm in agglomerative hierarchical clustering methods – seven steps to get clusters 2. each object is a independent cluster, n 3. two clusters with the lowest distance are merged to one cluster. reduce the number of clusters by 1 (n-1) 4. calculate the distance matrix between the new cluster and all remaining clusters 5. repeat step 2 and 3, (n-1) times until all objects form one reminding cluster
- 10. • Finally… 2.decide upon the number of clusters you want to keep (decision often based on the size of the clusters) 3.description of the clusters by means of the cluster forming variables 4.appellation of the clusters with catchy titles
- 11. • Consumers and Fair Trade Coffee • 214 interviews of consumers of fair trade coffee (personal and telephone interviews) • Cluster analysis in order to identify consumer typologies • Identification of 6 clusters • Description of these clusters by further analysis: comparison of means, crosstabs etc.
- 12. • Description of Cluster 1 (11,6%): “self-oriented fair trade buyer” : • Searches satisfaction by doing the good thing • Is not altruistic • Buys occasionally • Sticks to his conventional coffee brand • High level of formal education • Frequently religious (catholic or protestant)
- 13. • Description of Cluster 2 (13,6%): “less ready to take personal constraints” • States that “fair trade coffee is hard to find” • Feels responsible for fare development issues • Believes that fair trade is efficient for developing countries • Is less ready to go to special fair trade outlets • Buys conventional coffee • Likes the taste of fair trade coffee
- 14. • Description of clusters Cluster 3 (18,2%): “less engaged about fair trade” : • Feels no personal responsibility with regard to development questions • Doesn’t see the efficiency of the consumption of fair trade goods • The only thing that can make him change is the influence of friends • Is older then the average fair trade buyer and has less formal education
- 15. • Description of clusters: Cluster 4 (32,2%): “intensive buyer” • Has abandoned conventional coffee brands • Has started to buy fair trade quite a while ago (> 3 years) • Shops frequently in fair trade stores (and not in organic retail) • Is ready to act for fair development and talks to friends about it • Relatively young, with low incomes and high educational values
- 16. • Description of clusters: Cluster 5 (18,7%): “valueoriented” • Together with cluster 4 highly aware of development issues • Ready to act and to constraint consumption habits • Buys for altruistic reasons • Highly involved in social / political action • Most frequently women, highest household income among all clusters • Own security is the basis for solidary action
- 17. • Description of clusters: Cluster 6 (5,6%): “does not like the taste of fair trade coffee” • Lowest purchase intensity of all clusters • Not willing to accept constraints in consumption habits or higher prices • Most members of these group are attached to a conventional coffee brand • Relatively high incomes, age within the average of all groups, lower level of formal education • Less religious than other groups.
- 18. Cluster Analysis Applications in Marketing Research • New-product research: – Clustering brands can help a firm examine its product offerings relative to competition. Brands in the same cluster often compete more fiercely with each other than with brands in other clusters. • Test marketing: – Cluster analysis groups test cities into homogeneous clusters for test marketing purposes.
- 19. Cluster Analysis Applications in Marketing Research • Buyer behavior: – Cluster analysis can be employed to identify similar groups of buyers who have similar choice criteria. • Market segmentation: – Cluster analysis can develop distinct market segments on the basis of geographic, demographic, psychographic, and behavioristic variables.
- 20. Advantages • no special scales of measurement necessary • high persuasiveness and good assignment to realisable recommendations in practice Disadvantages • choice of cluster-forming variables often not based on theory but at random • determination of the right number of clusters often time consuming – often decided upon arbitrarily • high influence on the interpretation of the scientist, difficult to control (good documentation is needed)
- 21. Questions?