The federal government currently collects information for the Information Technology (IT) Projects in which it invests, but it does not apply this information to more accurately predict the performance of these projects. As a result, project managers within the federal government are not effectively planning for and managing projects in an optimal manner. This produces low project performance and ineffective resource utilization. A tool that accurately predicts the performance of IT projects would allow project managers to more closely monitor at-risk projects and plan accordingly to optimize project performance. This research project originally attempted to develop a model that more effectively predicted cost and schedule performance of IT projects by examining both continuous and categorical predictors, such as planned project schedule duration, planned project lifecycle cost, project manager experience level, software development lifecycle methodologies, and product release frequency. Ultimately, the research was merely able to achieve better results for schedule performance with a misclassification rate of 31.5% relative to 45% for a comparable prediction model. Understanding the potential impact of predictors on schedule performance of IT projects before they begin enables organizations to strategically reallocate resources before and during projects to minimize downtime and the number of potentially unused resources.
|In Administrative Set: