車內導引資訊系統影響駕駛者路線移轉行為之分析

Route Switching Behavior under the Influence of In-Vehicle Information with Latent Variable Modeling

本研究探討駕駛者與車內導引資訊系統所提供資訊的互動關係。有別於以往相關研究中駕駛者特性通常以社經變數的外部（explicit）型式「因子化」顯現，本研究將無法量化的因素如認知、態度與偏好等真實駕駛者行為主觀因素轉化為隱藏變數處理，反映用路人個人行為內部（implicit）特性，由此了解車內導引資訊系統對駕駛者路線移轉的影響。本研究利用圖示範例配合敘述性偏好問卷，探討旅運者在熟悉及不熟悉路網環境以及不同的質量資訊型態組合下的決策行為機制、影響其路線決策之因素及駕駛者之態度和認知問題的反應。再應用線性結構關係（LISREL）之統計方法，歸納出駕駛者態度及認知之隱藏變數，最後，將隱藏變數納入解釋變數中構建離散性選擇模式（羅吉特模式）。校估結果顯示，即時交通資訊的提供對駕駛者有正面的影響且會因此增加其路線移轉之傾向，駕駛者態度及認知特性也會因路網環境的熟悉程度與決策點區段的不同而有所變化。

The aim of this study is to explore the personal variation in response to various information strategies supplied by in-vehicle information devices. A three-stage methodology was implemented in this study. The first stage was a stated preference style of survey by which key elements of individual attributes as well as the responses to specially designed information scenarios in simulated displays were observed and recorded. This study then employed the technique of structural equation (LISREL) to extract each driver's individual characteristics into two latent variables identified as "attitude" and "cognition" towards in-vehicle information. Finally, logit-based discrete choice models were calibrated to describe individual's route switching intension under the corresponding information strategy. The emprical results demonstrate the usefulness of the inclusion of such latent variables.