Early identification of distressed capital projects: a longitudinal approach

Abstract

PurposePrevious studies investigate factors affecting project outcomes. Yet, it has not been fully explored regarding which factors differentiate healthy projects from distressed projects in the early stage of the project delivery process. The purpose of this study is to investigate the links between project-planning factors and project outcomes in the closing phase.Design/methodology/approachThe authors use a longitudinal survey method to examine the predictability of project-planning factors. Subsequently, the authos employ confirmatory factor analysis and hierarchical logit regression to develop project-distress classification models.FindingsAnalysis of 90 capital projects shows that performance variation in the project planning phase explains a substantial portion of project distress at completion. Subsequent univariate logit analysis shows that S5 (quality of scope control system) and Tn1 (new practices and technologies) variables have the strongest predictive abilities. Hierarchical logit analysis further shows that a combination of 15 metrics in the project-distress measurement model produces strong and stable predictive power.Research limitations/implicationsThis study assesses how well performance variation in the project-planning phase predicts project distress before construction phase. It does not assume the reported results apply to all types of projects. Nonetheless, future studies could generalize our findings by incorporating more types of projects.Originality/valueThis study takes a systematic approach, combining longitudinal survey, measurement theory and hierarchical logit analysis to identify distressed projects early, offering managers an opportunity to take early corrective actions. Practitioners may use this approach to investigate other types of projects and further refine the project-distress classification model into a project-specific model, thereby reflecting projects' unique characteristics.