複檻式決策過程模型之建構：條件評估法中抗議性答覆之處理

Model of Multiple-hurdle with Decision Process: A Possible Resolution of Protest Responses in Contingent Valuation Questions

本研究所設計的「複檻式決策過程模型」（multiple-hurdle with decision process model），是一個可以分析有經過明顯的選擇過程而後顯示最後支付金額的消費決策資料。由於此一模式對選擇決策過程的考量，同時對零支付樣本觀察值的豐富解釋能力，因而得以處理條件評估法中的抗議性答覆樣本。以支出差異詮釋利用開放雙界二元選擇誘導支付方式收集到的資料，同時以墾丁國家公園資源經濟價值的調查資料進行實證檢驗。結果顯示，複檻式決策過程模型因納入明顯選擇決策過程，因此，雙檻式模型（double-hurdle model）與標準Tobit模型具估計有效性，此外，亦更能詳細判別與解釋抗議性樣本觀察值的社會經濟特徵屬性，對不同決策過程之影響。同時，結果亦顯示雙檻式模型對零觀察值之解釋能力明顯優於標準Tobit模型，而本研究所建構的複檻式決策過程模型，對於分析具有大量零觀察值之抗議性答覆的樣本，則具有較雙檻式模型更佳之解釋能力，同時，也更能有效估計、分析開放雙界二元選擇條件評估方法收集來之資料。故而，以複檻式決策過程模型計算所得到平均每戶每年願意支付346元，及其對應的平均值95%信賴區間之上下限值，應該是可信度相對較高的數值。 The purpose of this paper is to construct a model, named as multiple-hurdle with decision process, which will be able to analyze data collected from a decision process with an obvious discrete choice and final continuous amount displaying. The model will extend to include not only real zero willingness to pay but also protest zero responses. The model is then applied to a set of data collected through double-bounded dichotomous choice with open-ended follow-up contingent valuation method for the benefit evaluation of Kengting National Park in Taiwan.The results have shown that model of multiple-hurdle with decision process not only has better fit than OLS, standard Tobit, and double-hurdle model but also has precise interpretation for the characteristics of protest responses at different stage of decision process. The mean willingness to pay for the natural resource of the site, estimated from multiple-hurdle with decision process, is 346 NT dollars per year for each household. And its corresponding 95% confidence interval of mean willingness to pay for the benefit is thus also more reliable than those computed from OLS, standard Tobit, and double-hurdle mode.

The purpose of this paper is to construct a model, named as multiple-hurdle with decision process, which will be able to analyze data collected from a decision process with an obvious discrete choice and final continuous amount displaying. The model will extend to include not only real zero willingness to pay but also protest zero responses. The model is then applied to a set of data collected through double-bounded dichotomous choice with open-ended follow-up contingent valuation method for the benefit evaluation of Kengting National Park in Taiwan.The results have shown that model of multiple-hurdle with decision process not only has better fit than OLS, standard Tobit, and double-hurdle model but also has precise interpretation for the characteristics of protest responses at different stage of decision process. The mean willingness to pay for the natural resource of the site, estimated from multiple-hurdle with decision process, is 346 NT dollars per year for each household. And its corresponding 95% confidence interval of mean willingness to pay for the benefit is thus also more reliable than those computed from OLS, standard Tobit, and double-hurdle mode.