多數的社會經濟統計資料係利用行政單元 (如縣市別、鄉鎮市區別、村里別)發布,但受到加總方式或空間單元大小、形狀之影響,經常將資料概括化,在空間分析上可能導致不同之結論 ,產生「可調整地區單元問題」(Modifiable areal unit problem, MAUP)(Openshaw 1984)。加總的統計資料無法呈現實際點資料的空間趨勢及呈現區域內的資料變異,可能產生偏頗的空間資訊被決策者使用,故國土資訊系統社會經濟資料庫分組近年來建置「統計區分類系統」之統計區圖資,做為蒐集社會經濟資料共用的空間統計單元。本文目的是將點資料彙整至統計區分類系統的空間單元 ,依序呈現鄉鎮市區及村里發布區及統計區 (包括最小統計區、一級發布區、二級發布區),研討不同空間單元保有空間趨勢的效能,利用空間自相關分析 (Spatial Autocorrelation Analysis, SAA)相關理論,如探索式分析 (explore analysis)、全域分析 (global analysis)與區域分析 (local analysis)等,探討登革熱病例在空間上是否有呈現聚集現象,且是否有熱源區域產生 ,以窺統計區空間單元之奧秘及效用 ,進而擴大推廣運用。

Most of socioeconomic statistical data are published with administrative district level. It causes the modifiable areal unit problem (MAUP) as a result of aggregating data and size or shape of spatial unit (Openshaw 1984). Aggregated data cannot demonstrate the spatial pattern and heterogeneity of original dataset. It maybe provide the bias or wrong spatial information to decision maker. From those results, A Spatial Statistical Area and Classification System (SSACS) have been implemented by National Geographic Information System in Taiwan. SSACS is supposed to collect socioeconomic of sharing spatial units. The purpose of this paper is aggregating point dataset into different spatial levels including township, village, 2nd level dissemination area, 1st level dissemination area and statistical area to exam the performance in persevering the spatial pattern and protecting the privacy. The material is dengue case in Kaohsiung. This paper uses those approaches such as spatial autocorrelation analysis (SAA), explore analysis, global analysis and local analysis to detect the hotspot area.