道路交通事故之重建，常因為「採認何項證據推論事實」與「採認證據與事實之關聯程度」兩類問題，致使當事人不能認同專家意見，而質疑鑑定過程中的主觀心證為黑箱，甚至不接受法院的裁判結果。本論文將事前機率與事後機率，利用概似定理等統計方法，引入高速公路交通事故重建個案分析，再將各項事件、證據分類後，納入正反兩方意見，由事故重建專家進行德菲法，對構成事件的每項證據核予關聯分數，分別獲得每個事件正反兩方的概似比，最終獲得兩方個別的事後機率，再進行分數比較，做為證據量化之客觀方法。

The issues of "what evidence reasons the fact" and "how the evidence correlates with the fact" make people who are involved in a road traffic accident not to trust the reconstruction expertise. The identification process of subjective evaluation are often considered a black box. Even the Court’s decisions were refused. This study presents a mathematical approach to quantify the opinions of reconstruction expertise. This approach integrates the prior probability, posterior probability, and likelihood ratio theorem. Freeway traffic accident reconstruction cases were introduced. All evidence and facts on arguments were classified, and provided to the reconstruction experts for Delphi technique. These experts scaled the relationship between constitute event and each evidence items respectively in order to obtain the likelihood of each event. Finally, the score comparison was made to demonstrate the objective method of evidence quantification.