國民生活安全情境指標之建構

Abstract

本文結合心理學、犯罪學、統計學、公共政策及警政學門之理論，研擬建構「國民生活安全情境指標」，以提供國人對各縣市生活風險的預測及對安全情境的覺知，作為選擇居住、旅遊、活動等之參考指標。本項指標可提供政府部門作為評估公共政策成本與政策效益之關聯程度，進而規劃施政策略及精進社會安全政策，達到應用社會指標以評鑑公共政策之功能。 國民生活安全情境指標係以主成分因素分析法將所蒐集之治安類、交通類、公共安全類及環境衛生類等34項指標群，層級聚斂其主成分之因素指標，共建構五個階層之指標體系。第五層計34項子指標，第四層計11項分指標，第三層計有治安、交安、公安、環衛4個構面中指標，第二層計有警政管理領域及公安環衛領域2面向次指標，第一層為「生活安全情境」總指標。 本文以因素負荷量的迴歸係數作為權數值，編製出各層級指標之指數。另為進行縣市之間橫向的比較及跨年度之縱向比較，將各年度之指標指數，以2000年之平均數及標準差進行標準化轉換。並將各縣市生活安全程度綜合指數予以排序。 本文將客觀性指數與各民調機構所做治安滿意度主觀性民調指數相互檢驗發現：治安構面指數及生活安全綜合指數均與治安滿意度、施政滿意度達顯著與高度相關，顯見本研究建構的指標體系具有高度之信度及效度。再與各縣市人文區位社經因素進行分析後發現：「人口密度」與治安滿意度及生活安全綜合指數也具有顯著的負向相關。 本研究獲致下列結論：（1）人口密集度與生活安全度具有顯著的反向關係；（2）都市化及工業化程度是影響生活安全度的重要因素；（3）都市化造成社會結構分化，削弱社會聯繫力及社區意識凝聚力，使社會規範力減弱。

This study, being truly interdisciplinary in its nature, utilized theories and practices of various disciplines such as psychology, criminology, statistics, public policy and police administration, in constructing an indicator system for national life safety in Taiwan. It is hoped that the indicators can offer citizens with tools for evaluating relative safety risks among different municipalities and counties. The indicators also provide realistic measures for governments in evaluating cost and benefit analysis of various public policies. The national life safety indicator system was made of 34 official statistics, the fifth or the bottom level, which cover four areas of safety related fields: personal safety, traffic safety, public safety and environmental and food safety. Factor analysis and regression analysis were performed and 34 statistics were integrated into 11 forth level indicators, then four third level indicators were reached using similar methods, then two second level indicators, and finally the total index for the top level. Typically, principal components were used in factor analysis for deciding the number of dimensions in each level, regression coefficients were used as weights while combining statistics into indicators. To compare indicators over years, the study chose 2000 official statistics as baseline and used standardization procedures for data 2001 through 2003. All indicators third level and above for each county were ranked so that comparisons can easily be made. The study also looked into the relationship between the indicators system and opinion polls on personal and public safety satisfaction issues. Results were remarkably consistent. In addition, the study also found that population densities, opinion polls on safety satisfaction issues, and the safety indicator system were very closely related while adding socio-economic indicators into the analysis. The study reached the following conclusions: population density and safety of life were negatively correlated, the extent of urbanization and industrialization were key factors of people’s life safety, and urbanization contributed to social disintegration, weakened social network connection and community coherence.