CHAPTER VI 



MARKET SEGMENTATION: 

 CLUSTER ANALYSIS OF HUNTER TYPES 



Marlcet Segmentation 



The mean net economic values presented thus far are useful in 

 gaining an understanding of how the average deer hunter values 

 his/her recreational experiences. Further, the values which 

 hunter subgroups place on their trips illuminate differences 

 across such things as residency and guided, nonguided status. 

 Useful as these groupings are, they nevertheless mask very real 

 differences and similarities between hunters and their 

 motivations for and expectations about hunting. There is no 

 truly average deer hunter, and even though such subgroups as 

 guided hunters share many of the same motivations, the term 

 "average guided hunter" remains a statistical construct of 

 questionable meaning. 



Individuals hunt for personal reasons, but this does not preclude 

 many individuals from sharing similar reasons for hunting. As 

 pointed out by Allen (1987), research suggests that there are 

 several methods available for identifying different "types" of 

 hunters within a sample. By identifying what "type" a hunter is 

 it becomes possible to attach dollar values not just to deer 

 hunting but to various types of deer hunting experiences. 



Cluster Analysis Design 



The cluster analysis used in this study was meant to isolate 

 subgroups, or "types", of deer hunters who defined their hunting 

 experience similarly. By understanding their collective 

 motivations for hunting we are better able to understand the 

 basis upon which they value their experiences. Cluster analysis 

 attempts to define subgroups of hunters which have significantly 

 different, yet conceptually meaningful characteristics. The 

 application of clustering used in this study follows closely that 

 suggested by Allen (1988) in his cluster analysis of Montana elk 

 hunters. 



The Montana deer hunting data had a large number of cases (in 

 excess of 2500) and therefore the SPSSx Quick Cluster program, 

 which efficiently clusters large files, was used. This program 

 sorts cases based on their Euclidean distance from cluster 

 centers which have been chosen from well distanced cases. The 

 clustering was performed on eight questions which asked hunters 

 to rate in importance reasons for hunting and factors which 

 influenced where they decided to hunt. These questions were 



41 



