大專生機車偏差駕駛行為的影響因素 -個人與學校的多層次結構方程模式之應用

Influential Factors of College Motorcyclists Aberrant Driving Behavior- Application of Individual and School Multilevel Structural Equation Modeling

張玉旻(Yu-min Chang)

指導教授 ： 陳惠國

國立中央大學/工學院/土木工程學系/碩士(2014年)

摘要

近年來，國內經濟發展促成機動車輛之快速成長，交通運輸需求日益漸增，進而延伸出許多道路交通安全的問題。根據交通部統計資料顯示，台灣地區登記之機車數量近十年內有逐漸增加之趨勢，且依據警政署統計通報可知，A1類交通事故肇事車種歷年主要以機踏車為主，約佔了將近50%，其可視為具有高肇事風險的車種類別。然而，又對於交通事故原因歸類，主要以汽(機、慢)車駕駛人過失為主。換言之，交通肇事的發生多來自於駕駛者本身因素所產生的駕駛偏差行為而導致。
本研究以台灣地區共159所大專院校，以集群抽樣之方式，分為北中南東部挑選其中80所學校，共有2,639位大專生參與調查，旨以多層次結構方程模式(MSEM) 探討學校與學生間之駕駛者偏差駕駛行為的影響因素，包含：風險感認與駕駛態度，此方法論為目前較新穎之統計分析，其融合了HLM其樣本具有階層之關係，延伸自線性迴歸概念，以及SEM將具有共同性之觀察變數背後的因素潛在變數萃取而來。
其結果顯示，偏差駕駛行為、風險感認、駕駛態度以及校園機車安全風氣，彼此間具有顯著的相關性；組內(學生間)層級的風險感認及駕駛態度對機車偏差駕駛行為有極顯著的負向影響，即學生駕駛者對於風險感認的程度有大與對交通安全的態度越高，表示其產生偏差駕駛行為上的頻率越小。而組間(學校間)層級所產生的校園機車安全風氣對偏差駕駛行為有極顯著的負向影響，即各學校所產生對於機車安全的不同風氣中之對於騎乘機車所形成的偏差行為的程度越大。針對上述結果，亦提出相關安全政策的討論與未來研究議題的建構。

關鍵字

多層次結構方程模式 ； 偏差駕駛行為 ； 風險感認 ； 駕駛態度 ； 校園機車安全風氣

並列摘要

In recent years, due to the rapid growth of the demand for transportation, making numbers of cars and motorcycles growing, and resulting into many traffic safety problem. According to Minstry of Transportation and Communcations statistics shows that the number of motorcycles registered in Taiwan has gradually increased the trend the last decade. As stated by Ministry of the Interior, the motorcycles in A1 accident the number of events is the largest and accounts for nearly 50%. Therefore, the motorcycles can be regarded as the types of vehicles having a high accident risk. However, traffic accident was mainly due to driver faults. In other words the traffic accident occurred from the driver factors about aberrant driving behavior.
This study conducted a survey on 2,639 students from a total of 159 universities from 80 universities in Taiwan. The sample of clusters divided into North, Middle, South, Eastern, and used multilevel structural equation modeling (MSEM) to analysis data. Explore the influencing factors of drivers’ aberrant driving behavior (ADB) between schools and students including risk perceptions and attitudes. MSEM is the currently newest statistical analysis, which combines hierarchical liner model (HLM) has related hierarchical sample data and extended from the concept of linear regression, and structural equation modeling (SEM) has extracted from the commonality of observed variable factors behind the latent variable.
The results show that aberrant driving behaviors, risk perceptions, and attitudes, are correlated with each other; in both individual and group level, risk perceptions and attitudes have a significant negative influence on ABD. It namely the extents of students’ drivers and each school ethos to have a higher risk perceptions and attitudes of a large traffic safety, which means that frequency of aberrant driving behavior to become smaller. According the results, this study proposes some discussions about safety policy and recommendations of the research issues for the future.

參考文獻

[11] 林芝嶸，機車駕駛經驗對小客車駕駛行為之關聯研究，國立交通大學運輸科技與管理學系，碩士論文，民國99年。

[12] 林鉦棽、彭台光，「多層次管理研究：分析層次的概念、理論和方法」，管理學報：23卷，6期，649-675頁，民國95年。

[15] 吳以雯，乘客協助駕駛人安全行車傾向之研究，國立交通大學運輸科技與管理學系，碩士論文，民國102年。

[16] 吳敏惠，國民小學校長領導風格、學校組織氣氛與教師組織承諾之研究，中原大學教育研究所，碩士論文，民國98年。

[18] 李仁豪、余民寧，「二層次結構方程式模型的應用：以教育心理學為例」，師大學報：教育類，53卷，3期，95-123頁，民國97年。