Decision Theory for Performance Evaluation of New Technologies Incorporating Institutional Issues: Application to Traffic Control Implementation

This dissertation develops a new framework for transportation evaluations. Most evaluation techniques fail to adequately assess all factors involved in transportation projects, with qualitative and institutional issues typically receiving less attention than easily quantifiable technical factors. This dissertation uses quantitative decision-theory techniques to develop a flexible approach that allows an analyst to look at all of the myriad issues involved in the evaluation of transportation projects.The research approach focuses on identifying an overall worth, which provides decision-makers with a quantitative measure to compare different system components. The innovative technique developed here integrates the multiple-attribute value function (MAVF) technique with the analytic hierarchy process (AHP). The overall worth of a project may be a combination of its worth under various operational conditions, with subjective relative weights, depending on the decision-makers. A hierarchy of such combinations are possible where the values for individual attributes themselves can be derived from the decision-makers using MAVF schemes. Certain complications arise in the technique, which require the development of a new scaling approach through the use of a universal scaling proxy. The research utilizes a hierarchical approach throughout the analysis while examining a total of four weighting schemes.The methodology is applied to the Anaheim Field Operational Test, a federally funded project, that implemented new traffic control technologies in Anaheim, California's special events area. The research's primary focus is on the city Traffic Engineer's values and preferences over the entire hierarchy. The development of six testing scenarios creates an opportunity to investigate the effects of many evaluation components as well as individual branches within the hierarchy.The evaluation looks at the percentage change in value between the system "before" and "after" implementation across scenarios. While the new system appears to decrease in value for most scenarios, one scenario, the alternate data set, actually shows an overall increase in value. The special event only operations scenario shows improvement over the base case, which indicates the system performs better under these conditions. The evaluation provides valuable insight into the behavior of the system under various conditions and provides guidance for future applications of this evaluation tool