水平型式大型購物中心業種多樣性區位配置策略之 空間型構分析A Space Syntax Analysis of Tenant Placement within Large-scale Horizontally-designed Shopping Centres

本文以空間型構指標，檢測水平型式大型購物中心內業種多樣性與區位特徵的關聯性，據以提供經營管理者進行配置策略上的建議依據。過去的研究建議水平型式的購物中心以同業種分散策略為基礎，但因缺乏空間區位特徵值的衡量方法，故並未提供確切的業種區位配置策略。本研究認為：過去的研究多以基本業種分類方式，直接作區位檢測，缺乏消費者行為層面的考量，故多無法在區位分析上得到顯著的結果；因此，本研究將基本業種，更進一步的以消費者購物主次要目的的核心周邊業種的萃取，以及根據消費者行為趨勢進行歸類，藉以提出業種多樣性區位配置的策略建議。本文廣泛蒐集世界各地共37場83個樓層平面的業種配置資料，並融入地理資訊系統及空間型構法為基礎的空間特徵資料表，進行四個層次的檢定，包含：(1)空間型構指標應用於購物中心內部空間複雜度的適用性檢定；(2) 28個業種的區位Logistic迴歸模型檢定；(3)以基本業種進行因素分析之核心與周邊業種之萃取；以及(4)以消費者行為的分類為基礎，進行核心型、周邊型、目的型、衝動型、比較型、便利型、娛樂型、以及服務型的業種類型區位Logistic迴歸模型檢定。結果顯示：本文所採用的空間形構指標，對大型購物中心內部空間的深度層次是有鑑別度的；而以基本業種分類的區位檢定，確實較無法產生承租組合的區位配置意涵，但若以消費者行為下的業種分類進行更深入的分析，則可以看到在總深度與便捷值上，多有依循著相關區位配置概念進行分布的顯著趨勢。

This paper aims to examine the location and placement strategies for various shoppers’ behavioral retail categories within large-scale horizontally-designed shopping centres. Although previous research has suggested that within a horizontally-designed shopping centre, retail stores of the same category should be dispersed to generate higher spillover effects, no significant placement strategies for tenant variety have been revealed. In this sense, we believe this is because previous research that used basic retail categories failed to incorporate shoppers’ behavioral fundamentals. Accordingly, this paper proposes the use of consumer-based coreperiphery retail categories and behavioral-based retail categories, which should more fully explain the data. By using Geographic Information System (GIS) and space syntax techniques, this paper gathers data on over 37 cases across different countries and builds up 83 floor plans with data on spatial features and tenant characteristics. The empirical study consists of 4 steps: (1) examining the effectiveness of selected space syntax indexes applied to 3 groups of floorplans based on complexity; (2) performing a locational logistic regression test of the 28 basic retail categories; (3) using factor analysis to extract the core and periphery retail categories; and (4) conducting a locational logistic regression test of the consumer’s behavioral-based classifications: core, periphery, impulsive, purposive, comparative, convenient, leisure and services. The results show that the selected space syntax indexes are suitable for showing the differences in spatial features within shopping centres. On top of that, only a few basic retail categories have relatively weakly significant results for the locational tests, and the behaviorbased classifications give rise to stronger explanations.