Document Detail

Title: RS131-1 - Qualititative Prediction of Estimate Accuracy
Publication Date: 3/1/1999
Product Type: Research Summary
Status: Archived Supporting Product
Pages: 27
This publication has been archived, but is available for download for informational purposes only.

This summary focuses on the development of the Estimate Score Program (ESP) software.
Order Now  


For engineering and construction projects, accurate early cost estimates are extremely important to the sponsoring organization and the engineering team. For the sponsoring organization, early cost estimates are often a basis for business unit decisions.

The accuracy of an early estimate depends on four determinants: who was involved in preparing the estimate, how was the estimate prepared, what was known about the project, and other factors considered while preparing the estimate. To measure the impact of these four determinants with respect to estimate accuracy, the CII Improving Early Estimates Research Team developed the Estimate Score Program (ESP), a software tool that evaluates 45 elements, organized in four divisions. The ESP software was developed to automate the scoring procedure and to assess accuracy and predict contingency based on historical cost data. The software and user’s manual are published by CII on a compact disc and publication (Implementation Resource 131-3, Estimate Score Program Guide [archived]). The research team also produced Implementation Resource 131-2, Improving Early Estimates: Best Practices Guide.

The ESP tool enables the business unit and the team to make better decisions based on statistically valid information. It improves team alignment by providing a better understanding of the estimate and the factors that can affect the accuracy of an early estimate.

This summary focuses on the research team’s development of the ESP tool (CII Implementation Resource 131-3). The tool itself is a credible, readily usable software that has the potential to improve not only early estimates, but project team alignment as well.