Descriptive data analysis.

This study aimed to develop predictive models that could be used to estimate the cost and schedule performance of reconstruction of transportation infrastructure damaged by hurricanes and to determine the predictors that are robustly connected to the developed models. Stepwise multiple linear regression and extreme bound analysis (EBA) were used to develop the models and determine the robust and fragile predictors, respectively. The results demonstrated that seven cost performance predictors and nine schedule performance predictors accounted for Adjusted R-Squared of 92.4% and 99.2%, respectively. The results of the EBA revealed that four cost and seven performance predictors were robustly connected to the developed cost and schedule performance predictive models. It was concluded that increases in laborers’ wages, the number of inspections, information and data management, and addressing safety and environmental issues prior to a project’s execution were predictors of both the cost and schedule performance of reconstruction projects. The outcomes of this study provide knowledge and information that will be helpful to decision-makers who are responsible for mitigating delays and cost overruns, and effectively allocating their limited resources available following a disaster.本研究旨在构建可用于估算飓风损毁交通基础设施重建的成本与进度绩效，并确定与所构建模型强相关预测因子的预测模型。研究分别采用逐步多元线性回归与极值边界分析（Extreme Bound Analysis, EBA）完成模型构建与稳健、非稳健预测因子的识别。结果显示，7个成本绩效预测因子与9个进度绩效预测因子分别可解释调整后决定系数（Adjusted R-Squared）的92.4%与99.2%。极值边界分析结果表明，4个成本相关预测因子与7个进度绩效相关预测因子分别与所构建的成本、进度绩效预测模型强相关。研究结论指出，在项目执行前提升劳工薪资、增加巡检频次、优化信息与数据管理，以及落实事前安全与环境管控要求，是同时影响重建项目成本与进度绩效的核心预测因子。本研究成果可为负责缓解灾后延误与成本超支、高效调配灾后有限资源的决策者提供极具价值的知识与信息参考。