The development of early-life coastal recreation experience scale: Competing model and cross-validation

• Main Authors: Cheng Shih-Hao, 鄭士豪
• Other Authors: Chang Liang-Han
• Format: Others
• Language: zh-TW
• Published: 2011
• Online Access: http://ndltd.ncl.edu.tw/handle/25695952917051836629

碩士 === 國立澎湖科技大學 === 觀光休閒事業管理研究所 === 99 === The purpose of this study was to develop a suitable early-life coastal recreation experience scale for Penghu residents. According to flow path of the scale development, the study intended to build a robust and standardized measurement tool for current or future researchers. In the first, an intensive literature review was completed to classify domains of the scale. Based on that, four domains, personal outdoor recreation, organizational outdoor recreation, media and education, and perception of environmental behavior, were extracted. The research population included one city and 5 rural townships of Penghu County. Questionnaires were distributed to collect data by purposive sampling method. There were 761(94.1%) usable questionnaires were obtained. Exploratory actor analysis was used to identify the constructs underlying the 16 items of early-life coastal recreation experience scale. Moreover, a competing model, from confirmatory factor analysis (CFA) of structural equation modeling (SEM), was applied to pinpoint the optimum model. The results revealed that four factors oblique model and second order with one factor model passed the model fit criteria and model specification. However, regarding to parsimony, the hierarchical model with one factor model was adopted and the second order was named as “early-life coastal recreation experience”, the first order factors were named as “personal outdoor recreation”, “organizational outdoor recreation”, “media and education”, and “perception of environmental behavior”. Finally, cross-validation was used to examine predictive validity of the model. The results showed that the model had appropriate reliability, construct validity, and predictive validity, and can be applied for future research.