本研究從監控酒後駕車肇事趨勢變化著手,檢討我國改善酒後駕車肇事措施介入時效之適宜性。本研究以台北市為研究對象,蒐集民國85年3月至民國89年12月共五十八個月的酒後駕車肇事與防治措施相關資料。本研究分別應用統計程序控制法(statistical process control method; SPC)、成對t檢定以及卜瓦松迴歸模式(Poisson regression models)分析兩項防治措施對減少酒後駕車肇事次數之實施成效。研究結果顯示警察執法與將酒後駕車列入刑法公共危險罪具顯著降低酒後駕車肇事之成效,但酒後駕車列入刑法公共危險罪政策之實施卻隨時問對安全成效產生先升後降之週期性現象。
This study is undertaken to evaluate the effects of the new rehabilitation program for drunk drivers (NRPFDD) and the policy of drunk driving is criminal (DDIC) to reduce fatal accidents involving drunk drivers in Taipei City. The control chart method, pair-t tests, as Well as Poisson regression analysis were used in order to gain deep insight into the accident occurrence involving drunk drivers. The results of study show that police enforcement and the implementation of DDIC will significantly reduce the fatal accidents involving drunk drivers. The implementation of DDIC is expected to reduce the fatal drunk driving accidents significantly. The effect of the implementation of DDIC to reduce accident occurrence of drunk driving will become a quadric form as time passes by.