Document Detail

Title: RR107-11 - Development of a Predictive Tool for Continuous Assessment of Project Performance
Publication Date: 1/1/1996
Product Type: Research Report
Status: Archived Reference
Pages: 196
This publication has been archived, but is available for download for informational purposes only.

Russell, Jaselskis, et al., Univ. of Wisconsin-Madison
Order Now  


The Construction Industry Institute (CII) convened a task force in 1992 to investigate predictive tools for the construction industry. The need for this research stems from the multitude of data generated by construction projects. Project managers need to use this information to assist them in managing projects. Few tools exist that enable them to use data from past projects to plan for completing an upcoming project on-time and at or under budget (defined as successful for purposes of this research). Likewise, limited tools are available that enable project managers to use this historical data during design, procurement, and construction of a project to ensure successful completion of a project with respect to schedule and budget. The research presented in this source document aims at developing a process whereby owner, engineer, and construction contractor organizations can collect, maintain, and analyze data that are provided throughout the life of a project to assess project performance.

The research was performed under the guidance of the CII Predictive Tools Task Force which consisted of 16 representatives from CII member companies. Data were collected on 54 projects using a 41-page data collection tool aimed at gathering project-specific information on 76 time-dependent variables as well as a variety of project characteristics and project performance and schedule information. Companies were given approximately two weeks to compile as much of the information as possible on past projects. Interviews were then conducted with key project participants to gather additional information on the projects.

To assist in the data analysis process, a computer tool was developed using Microsoft® Excel and FoxPro, software. This data analysis tool was designed so that it can be used for collection of additional project data and enable project managers to assess the current performance of their projects relative to all other projects in the database. The name of the software product is CAPP® Continuous Assessment of Project Performance. The program operates in four modules that: 1) enable input of project data; 2) normalize data into equal increments for all projects; 3) provide graphical output of variable information to aid in pre-project planning and project assessment; and 4) perform statistical analysis to ensure the significance of the analyses being performed.

To validate the process developed, one variable was analyzed extensively. This analysis was done in three steps. The first step was to identify one variable most suitable for in-depth analysis. The second step differentiated the selected variable data by a variety of project parameters (e.g., construction type) to identify which parameters most impact the variable’s ability to discriminate between successful and less-than-successful projects. Finally, data were differentiated by multiple parameters to validate the source document that a database containing more well-defined projects would allow better discrimination between successful and less-than-successful projects.

The source document concludes with a discussion of how the tool can be used by industry. Many organizations collect continuous data on their construction projects. The process developed here enables them to normalize that data so that projects with different characteristics can be combined to generate curves for successful and less-than-successful projects. These curves can be used during pre-project planning as a new way of looking at the planned expenditure of resources over the course of a project. The curves can also be used to assess project performance. By comparing an ongoing project with the averaged curves, companies can begin to predict when a project is performing similar to projects previously built whose outcome was less than desired.