本研究以問卷方式對墾丁南灣的遊客進行調查，調查結果並應用「非市場估價法」之「旅遊成本法」(Travel Cost Method, TCM)來建立墾丁南灣遊憩需求實證模型，並以計數模型中的TPOIS、TNB以及On-Site Poisson模型來修正現場樣本的截斷與內生分層偏誤，據此推估墾丁南灣的海灘遊憩效益。
初步研究結果顯示On-Site Poisson模型結果較好，較適合用來評估墾丁南灣的遊憩效益。如以On-Site Poisson模型估算墾丁南灣的遊憩彈性，其價格彈性為-0.2385、所得彈性為0.1193。至於在墾丁南灣的遊憩效益評估部分，以TPOIS模型估計的消費者剩餘為32,154(元/年/人)，而TNB模型估計的消費者剩餘為32,258 (元/年/人)，另外On-Site Poisson模型估計的消費者剩餘則為27,397 (元/年/人)。三個模型中，TNB模型的消費者剩餘最高，其次是TPOIS模型，On-Site Poisson模型最低。至於在同一需求函數之估計模型下，三種遊憩效益(CS、CV、EV)之結果則極為接近，其原因為本研究實證資料所得對旅遊需求並無顯著影響，即所得效果不明顯，因此CS、CV及EV無顯著差異。 The recreation benefit of Kenting Nanwann beach was estimated by travel cost method. The trip demand function was constructed by different count data models, namely, Truncated Poisson (TPOIS), Truncated Negative Binomial (TNB) and On-Site Poisson.
This study has found that the most appropriate model was the On-Site Poisson to estimate recreation benefit of Kenting Nanwann beach. The price elasticity of demand and income elasticity of demand were -0.2385 and 0.1193. The average consumer surplus (CS) under TPOIS model, TNB model, and On-Site Poisson model were NT$32154, 32258, 27397, respectively. The results of this research indicated that CS under TNB model was the highest, and under On-Site Poisson model was the lowest. The recreation benefits estimated by the different indices (CS, CV, EV) in the same count data model were not significantly different. The reason was that the effect of income was not significant in the trip demand function.
