Nonmetric Multidimensional Scaling and Hierarchical Clustering: Procedures for the Investigation of the Perception of Sports
David Michael Levine · Research Quarterly for Exercise and Sport · 1977
Nonmetric multidimensional scaling and cluster analysis are described as procedures that can be used to study the perception of sports. Measures of the proximity of objects (sports) are discussed along with alternate data collection methods. Nonmetric multidimensional scaling develops a map of the objects in a varying number of dimensions and enables the researcher to interpret the factors that differentiate the various objects. Hierarchical cluster analysis develops a tree-like structure in subdividing the objects into homogeneous groups. These techniques are illustrated in a study of 13 sports. Two factors that differentiated the sports are revealed by nonmetric multidimensional scaling: team vs individual and degree of action. This result, along with the clusters developed through hierarchical clustering, leads to the conclusion that these sports have been primarily distinguished by how team oriented they were perceived to be.