Cluster analysis as an exploratory tool in tourism research

Abstract

The objective of this dissertation is to attempt to find homogenous groups of British tourists, which portray similar expenditure patterns. The data which was analysed in this study are the British respondents to the 2003 expenditure survey. This survey was developed and is carried out annually by the Research and Information Division of the Malta Tourism Authority. The data was first inserted into the SPSS database and the respondents identified as singles, couples, families and groups were analysed separately in an attempt to discover group structures which had not been previously considered. Cluster analysis was applied as an exploratory statistical tool. The main steps involved in this study were the choice of: cluster variables using principal component analysis to aid in variable reduction, proximity measures to be applied and the choice of clustering methods to use. A two-stage clustering approach was performed and the clusters formed were profiled.