本研究在群聚分析（cluster analysis）的架構下研究國際觀光旅館的選址策略。有別於前人的研究多著重於分析地理位置、價格、以及規模大小如何影響飯店的選址決策，本研究的目標為了解在飯店在選址興建之後，上述三個變數如何影響國際觀光旅館的營運獲利表現。此研究以台灣本島68 家國際觀光旅館中在1998 年至2008 年間連續營運的37 家飯店為研究對象。考慮到研究中價格變數可能產生的內生性問題（endogeneity），我們以土地價格指數，以及員工薪資作為工具變數（instrumental variables），並應用二階段最小平方法（G2SLS）來修正內生性所造成的誤差。研究結果顯示，在不考慮內生性問題的前提下以最小平方法估計（OLS）所得的結果，與二階段最小平方法（G2SLS）的分析結果呈現相當大的差異。此比較結果顯示了修正內生性問題的重要性。二階段最小平方法（G2SLS）的估計結果顯示，新加入市場之國際級觀光旅館應在定價上採差異化策略，拉開與區域競爭者的價格區間，同時縮短與競爭者間，地理位置以及飯店規模差異，來追求獲利的極大化。

This study aims to allocate the positioning strategies of international tourist hotels (ITHs) in Taiwan based on the framework of cluster analysis. Unlike previous studies which focused on the founding decisions of hoteliers in terms of geographic location and the associated strategies regarding price and scale, this study focuses on how these three product dimensions influence ITHs’ operational performance. Our data consists of 37 ITHs among 68 ITHs in Taiwan from 1998 to 2008. We use land price index and wage to instrument for an endogenous price difference variable and estimate a General Two-Stage Least Square (G2SLS) model to correct for potential bias. The results differ dramatically between the G2SLS and OLS models, reflecting the importance of correction. The results from G2SLS suggest that new ITHs entering the market should differentiate themselves by expanding the price differences from their competitors, and play an agglomeration strategy by minimizing the differences in both geographic distance and scale to improve their financial performance.