The means-end approach in market segmentation – clustering of laddering data

The objective of this paper was to present results of research on two methodological issues related to the clustering of laddering data. The first was the methods of aggregation of information from ladders generated by one respondent and the second was the measurement of ladders’ dissimilarity. Two methods of aggregation of information from ladders were proposed, three sequence dissimilarity measures were presented and all combinations of them were used in the analysis. The clustered data originate from a research on Polish adolescents’ online consumer behaviour wherein the means-end approach was used. 1004 high school students participated in the research. Data were clustered from 2 to 10 groups, six modes of analysis were used, thus 54 solutions were built. Solutions with the same number of groups were compared with the adjusted Rand index. Analysis indicates the influence of sequences dissimilarity measures and methods of aggregation of information from ladders on clustering results.