機車為台灣地區重要的交通運輸工具之一，且發生事故時死亡及受傷人數都相對高於其他車種；而臺灣地區各縣市中以高雄市和台北市為人口密度最高的兩大城市，高雄市之機車密度更高居全台各縣市之冠，甚至超過排名第二的台北市1.95 倍，因此造成機車事故傷亡嚴重程度的原因對於平時以機車做為交通運輸工具的民眾是十分重要且不可輕忽的議題。
　　過去有關交通事故資料的研究方法多以傳統統計方法如卜瓦松迴歸模式、多項羅吉特模式等為主，近年來也有研究是應用非參數型模式如分類迴歸樹、貝氏神經網絡等來分析，但這些方法皆忽略了交通事故資料中存在的層級結構。多層次模式(Multilevel model)可以處理具有階層結構的資料，考慮到個體層級及總體層級之間的影響，並能對誤差項進行多層次的分割估計，此模式常被應用在社會與行為科學領域，卻很少被應用於道路安全的領域上。
　　本研究使用台灣內政部警政署民國97 年道路交通事故資料庫，應用多層次模式及二項羅吉特模式，針對高雄市和台北市號誌路口之機車事故當事者來探討機車事故嚴重程度之因素，並對兩模式進行比較。本研究將資料分為事故當事者層級以及交通事故層級來進行多層次模式分析，並檢驗高雄市及台北市之事故資料是否具有層級結構。最後分析結果發現，在模式適合度部份，雖然是以二項羅吉特模式適合度較佳，但在顯著性方面是以多層次模式有顯著影響之變數較多，且驗證出兩城市之事故資料是具有階層特性的，這也代表了若忽略層級資料結構，會導致不良的估計值和標準差。而在多層次模式下顯著的變數可以發現，事故當事者層級之變數是以年齡越高、男性、無駕照資料和酒後騎車的事故當事者發生較嚴重傷亡的機率比較高；交通事故層級則是以事故發生在夜間、速限大於50km/h的路段和四岔路口的車禍傷亡機率較高。

Motorcycles have been one of the most important transportation modes and the number of injury and fatality in motorcycle accidents was relatively higher than other vehicle type in Taiwan. Kaohsiung city and Taipei city of aiwan have both the highest population density, the motorcycle density of Kaohsiung city is the highest in Taiwan even more than Taipei city which ranked the second city for 1.95 times. Thus, the cause of severity of motorcycle accident is very important for the motorcyclists.
　　Many methods had been employed to study road accident data before. These methods include the statistical models, such as Poisson, Negative binomial regression, Multinomial logit model and so on. And Non-parametric models also had been employed to analyze road accident data in recent years such as Classification and regression tree, Bayesian neural network and so on. No matter statistical models or non-parametric models, the possible existence of hierarchical structures within accident data is commonly ignored. Multilevel model can deal with the data which have a hierarchical structure. This model’s feature except consider that hierarchical structure, another important feature is it can reduce standard error and improve the statistical power. A good number of applications of multilevel model have been found in sociological research, but rarely used in transportation research.
　　The traffic accidents data from Taiwan National Police Agency of Ministry of the Interior for 2008. This study developed a Multilevel model and binary logit model to identify the significant factors affecting the severity level in motorcycle traffic crashes at signalized intersections in Kaohsiung city and Taipei city, and compare these two models. The traffic data consist of two different levels: level 1 consisting of individual-level characteristics and level 2 consisting of crash-level characteristics. The fitness assessment of binary logit model was better than multilevel model, but the significant covariates of multilevel model are more than binary logit model, and model assessment ensured the suitability of introducing the crash-level random effects. It means that disregarding hierarchies can lead to the production of models giving unreliable estimates of prevision and incorrect standard errors. Results indicated that severity level was found to be associated with individual’s age, gender, license and drink condition, as well as light condition and speed of crash.

1. 內政部警政署：http://www.npa.gov.tw/NPAGip/wSite/mp?mp=1
2. 王濟川、郭志剛(2004)，Logistic 迴歸模型-方法及應用，五南圖書出版股份有限公司。
3. 交通部統計資料: http://www.motc.gov.tw/
4. 行政院衛生署：http://www.doh.gov.tw/cht2006/index_populace.aspx
5. 吳宗霖 (2010)，台灣地區汽車傷亡事故肇因之研究，中央大學統計研究所碩士論文。
6. 林正基 (2004)，高雄市大學生機車意外事故流行病學調查，高雄醫學大學健公共衛生學研究所碩士論文。
7. 林立 (2007)，機車事故之受傷部位及傷亡程度之探討，嘉義大學運輪與物流工程研究所碩士論文。
8. 邱皓政 (2008)，量化研究法(一)：研究設計與資料處理(修訂版)，台北：雙葉。
9. 員旭成 (2007)，臺灣地區交通事故傷害肇因之研究，銘傳大學應用統計資訊系碩士班碩士論文。
10. 孫璋英 (2004)，汽機車單一車輛事故駕駛人死亡勝算模式之研究，台北大學統計學系碩士論文。
11. 陳柏華 (2010)，以多層次模式探討組織安全氣候強度對貨運司機危險駕駛及健康之影響，成功大學交通管理科學系碩士班碩士論文。
12. 黃韻芝 (2006)，臺閩地區道路交通事故之研究，中央大學統計研究所碩士論文。
13. 溫福星、邱皓政(2009)，多層次模型方法論：階層線性模式的關鍵議題與試解，臺大管理論叢，第19 卷第2 期，頁263-294。
14. 溫福星(2006)，階層線性模式：原理、方法與應用，台北：雙葉。
15. 詹子儀 (2004)，花東地區交通事故特性與嚴重程度模式建立之研究-以小型汽車為例，屏東科技大學機械工程系碩士論文。
16. 簡瑞增 (2008)，應用非參數型模式分析大卡車交通事故傷亡嚴重程度，嘉義大學運輸與物流工程研究所碩士論文。
17. Al-Ghamdi, A.S. (2002). Using logistic regression to estimate the influence of accident factors on accident severity, Accident Analysis and Prevention, Vol.34,pp.729–741.
18. Boufous, S., Finch, C., Hayen, A., Williamson, A. (2008). The impact of environmental, vehicle and driver characteristics on injury severity in older drivers hospitalized as a result of a traffic crash. Journal of Safety Research,Vol.39, pp.65–72.
19. Bustamante, M.X., Zhang, G., O’Connell, E., Rodriguez, D., Borroto-Ponce, R.(2007). Motor vehicle crashes and injury among high school and college aged drivers. Association of Emergency Physicians, Vol.17, No. 9, pp. 742.
20. Chang, L.Y. (2005). Analysis of freeway accident frequencies : negative binomial regression versus artificial neural network. Safety Science, Vol.43, No.8, pp.541–557.
21. Chang, L.Y. (2005).Empirical analysis of the effectiveness of mandated motorcycle helmet use in Taiwan. Journal of the Eastern Asia Society for Transportation Studies, Vol.6, pp.3629–3644.
22. Chang, L.Y., Wang, H.W. (2006). Analysis of traffic injury severity- an application of non parametric classification tree techniques. Accident Analysis and Prevention, Vol.38, pp.1019–1027.
23. Chiu, W.T., Kuo, C.Y., Hung, C.C., Chen, M. (2000). The Effect of the Taiwan Motorcycle Helmet Use Law on Head Injuries, American Journal of Public Health, Vol.90, No. 5, pp.793–796.
24. Delen, D., Sharda, R., Bessonov, M. (2006). Identifying Significant Predictors of Injury Severity in Traffic Accidents Using a Series of Artificial Neural Networks. Accident Analysis and Prevention, Vol.38, pp.434–444.
25. Evans, L. (1992). Car size or car mass : which has greater influence on fatality risk? American Journal of Public. Health, Vol.82, No. 8, pp.1105-1112.
26. Farmer, C., Braver, E., Mitter, E. (1997). Two-vehicle side impact crashes : the relationship of vehicle and crash characteristics to injury severity. Accident Analysis and Prevention, Vol. 29, No. 3, pp.399-406.
27. Gonzales, M.M., Dickinson, L.M., DiGuiseppi, C., Lowenstein, S.R. (2005). Student drivers : A study of fatal motor vehicle crashes involving 16-year-old drivers. Annals of Emergency Medicine, Vol.45, No. 2, pp.140-146.
28. Haddon, W., Jr. (1999). The changing approach to the epidemiology, prevention,and amelioration of trauma : the transition to approaches etiologically rather than descriptively based, Injury Prevention, Vol.5, pp.231-235.
29. Haque, M.M., Chin, H.C., Huang, H. (2009). Modeling fault among motorcyclists involved in crashes. Accident Analysis and Prevention, Vol.41, pp.327–335.
30. Helai, H., Abdel-Aty M. (2010). Multilevel data and Bayesian analysis in traffic safety. Accident Analysis and Prevention, Vol.42, pp.1556–1565.
31. Helai, H., Chor, C.H., Haque, M.M. (2008). Severity of driver injury and vehicle damage in traffic crashes at intersections A Bayesian hierarchical analysis. Accident Analysis and Prevention, Vol.40, pp.45–54.
32. Jones, A.P., Jørgensen, S.H. (2003). The use of multilevel models for the prediction of road accident outcomes. Accident Analysis and Prevention, Vol.35,pp.59–69.
33. Keall, M.D., Frith, W.J., Patterson, T.L. (2004). The Influence of Alcohol, Age and Number of Passengers on the Night-time Risk of Driver Fatal Injury in New Zealand. Accident Analysis and Prevention, Vol.36, pp.49–61.
34. Khorashadi, A., Niemeier, D., Shankar, V., Mannering, F. (2005).Differences in rural and urban driver-injury severities in accidents involving large-trucks An exploratory analysis. Accident Analysis and Prevention, Vol.37, pp.910–921.
35. Kim, D.G., Lee, Y., Washington, S., Choi, K. (2007). Modeling crash outcome probabilities at rural intersections Application of hierarchical binomial logistic models. Accident Analysis and Prevention, Vol.39, pp.125–134.
36. Kockelman, K.M., Kweon, Y. J. (2002). Driver injury severity : an application of ordered probit models. Accident Analysis and Prevention, Vol.34, pp.313-321.
37. Lenguerrand, E., Martin, J.L., Laumon, B. (2006). Modelling the hierarchical structure of road crash data Application to severity analysis. Accident Analysis
and Prevention, Vol.38, pp.43–53.
38. Li, X., Lord, D., Zhang, Y., Xie, Y. (2008). Predicting motor vehicle crashes using support vector machine models. Accident Analysis and Prevention. Vol. 40, pp.1531-1542.
39. Lin, M.R., Chang, S.H., Huang, W., Hwang, H.F., Pai, L. (2003). Factors Associated With Severity of Motorcycle. Annals of Emergency Medicine, Vol.41,No. 6.
40. Lin, M.R., Chang, S.H., Pai, L. and Keyl, P.M. (2003). A longitudinal study of risk factors for motorcycle crashes among junior college students in Taiwan. Accident Analysis and Prevention, Vol.35, pp.243–252.
41. Lord, D. (2006). Modeling motor vehicle crashes using Poisson-gamma models : examining the effects of low sample mean values and small sample size on the Estimation of the fixed dispersion parameter, Accident Analysis and Prevention,Vol.38, No. 4, pp.751-766.
42. Lord, D., Mannering, F. (2010). The statistical analysis of crash-frequency data : A review and assessment of methodological alternatives. Transportation Research Part A, Vol.44, pp.291–305.
43. Neyens, D.M., Boyle, L.N. (2007). The effect of distractions on the crash types of teenage drivers. Accident Analysis and Prevention, Vol.39, pp. 206–212.
44. O’Donnell, C., Connor, D. (1996). Predicting the severity of motor vehicle accident injuries using models of ordered multiple choice. Accident Analysis and Prevention, Vol. 28, No. 6, pp.739-753.
45. Ouyang, Y. et al. (2002). Modeling the Simultaneity in Injury Causation in Multivehicle Collisions. Transportation Research Record 178(02-3349),pp.143-152.
46. Preusser, D.F., Williams, A.F., and Ulmer R.G. (1995). Analysis of fatal motorcycle crashes : crash typing. Accident Analysis and Prevention, Vol. 27, No.6, pp.845-851.
47. Riviere, C., Lauret, P., Ramsamy, J.F.M., Page, Y. (2006). A Bayesian neural network approach to estimating the energy equivalent speed. Accident Analysis and Prevention, Vol.38, pp.248-259.
48. Riviere, C., Lauret, P., Ramsamya, J.F.M., Page, Y. (2006). A Bayesian Neural Network approach to estimating the Energy Equivalent Speed. Accident Analysis and Prevention, Vol.38, pp.248–259.
49. Savolainen, P., Mannering, F. (2007). Probabilistic models of motorcyclists’ injury severities in single- and multi-vehicle crashes. Accident Analysis and Prevention, Vol.39, pp.955–963.
50. Shankar, V., Mannering, F. (1996). An exploratory multinomial logit analysis of single-vehicle motorcycle accident severity. Journal of Safety Research, Vol.27, No. 3, pp.183-194.
51. Ulfarsson, G., Mannering, F. (2004). Differences in male and female injury severities in sport-utility vehicle, minivan, pickup and passenger car accidents. Accident Analysis and Prevention, Vol.36, No. 2, pp.135-147.
52. Wong, J.T., Chung, Y.S. (2007). Rough set approach for accident chains exploration. Accident Analysis and Prevention, Vol.39, pp.629–637.
53. Wong, J.T., Chung, Y.S. (2008). Analyzing hetrogeneous accident data from the perspective of accident occurrencesis. Accident Analysis and Prevention, Vol.40, pp.357–367.
54. Xie, Y., Lord, D., Zhang, Y. (2007). Predicting moter vehicle collisions using Bayseian neural networks : an empirical analysis. Accident Analysis and Prevention, Vol.39, pp.922-933.
55. Yannis, G., Papadimitriou, E., Antoniou, C. (2007). Multilevel modelling for the regional effect of enforcement on road accidents. Accident Analysis and Prevention, Vol.39, pp.818–825.
56. Yau, K.K. (2004). Risk factors affecting the severity of single vehicle traffic accidents in Hong Kong. Accident Analysis and Prevention, Vol.36, No. 3, pp.333-40.
57. Zajac, S.S., Ivan, J.N. (2003). Factors Influencing Injury Severity of Motor Vehicle Crossing Pedestrian Crashes in Rural Connecticut. Accident Analysis and Prevention, Vol.35, pp.369–379.