ABSTRACT
The use of design-build project procurement is rapidly increasing in the public sector of the United States construction industry. Documented design-build success and recent changes in federal procurement law are indicators of continued large-scale growth. The rapid growth, combined with a lack of experience among many public agencies, spawns a variety of issues that must be addressed. One issue is proper project selection. As new agencies experiment with design-build, appropriate project selection is a primary consideration affecting successful delivery. To date, there is no systematic or formalized method for selecting projects that are appropriate for design-build. This research presents a web-based automated predictive tool that can be used in the earliest stages of project selection to enhance the chances of design-build project success. The hypothesis proven true by this research is that there are public sector projects that are appropriate for design-build and success is predictable. Key project characteristics that are appropriate for design-build projects are identified and a statistically based model is presented. Using multi-attribute analysis, retrospective case study data collection methods, and multiple regression modeling, this research produces a model that predicts the success of the design-build process on future projects.
