Assessing the generalizability of study results: an outdoor recreation application

Abstract

Onsite surveys of visitors have become a common method of providing information regarding recreation area visitors' norms, attitudes, motivations, management preferences, etc. Study results are often implied to be representative of a future time period for management planning purposes. One enduring method is segmenting visitors based on motivations and examining preferences for management actions/settings by group. When such a study is directly intended to guide management actions, the 'groups' must offer some degree of insight into future visitors, i.e., generalize across time. Data gathered during 2004 at the Kennecott National Historic Landmark within Wrangell- St. Elias National Park and Preserve were used to establish a hypothesized relationship between motivations and preferences for management actions. The study was replicated in 2005 to assess the generalizability of the results. A K-means cluster analysis on visitor motivations revealed five unique groups of visitors (n=206). Each group was linked with preferences for six hypothetical management actions. Utilizing generalizability theory and cluster profiling, results suggest the same five visitor types were present at Kennecott in 2005 (n=198). Furthermore, five management options exhibited evidence of generalizability across time. Although the degree of preferences varied slightly over time, opinions towards the management options did not change.