Qualititative Prediction of Estimate Accuracy

The research identified 45 elements to measure the quality of an estimate, including 14 PDRI scope definition elements and 31 elements not directly related to scope definition. The complete list of 45 elements was then subdivided into eleven factor groups; through factor analysis in developing Estimate Score Program (ESP) each of the 45 elements was allocated an element weight and determines the factor weights of the eleven factor groups.  Factor group weights have relative influence the four drivers of the estimate score. (RS131-1, p. 4)
 

The factor analysis and regression analysis performed in development of the ESP identified eleven factor groups with the most impact on estimate accuracy. 

The figure below describes the eleven factors in order of significance. The most significant factor, basic process design, accounts for 25% of the estimate score therefore concludes that a comprehensive and definitive process design is critical to the accuracy of early estimates. Basic process design includes process flow sheets, heat and material balance, project schedule, capacities, equipment list, and P&IDs as the highest factors elements. (RS131-1, p. 9)

This is the second factor of significance in the accuracy of early estimates. Extensive analysis of historical project data revealed that there is a positive correlation between estimated and actual total installed cost based on issues related to the team’s skills and experience, the estimating procedures, and scope definition. It was concluded in situations where scope definition is limited, project teams with good established estimating practices and experienced personnel were able to prepare accurate early estimates. 

The figure below illustrates the hypothesis RT-131 developed regarding the positive impacts of team experience and good procedures with various levels of scope definition. (RS131-1, p. 25)

Graphic showing effect of the team experience and procedures on estimate quality.

This tool had the function to allow users to query the database of completed projects to identify projects against which to compare an early estimate. A Scatter Plot and Cumulative Graph are graphical views of project data returned by the query and display the Estimate Score values or Probability values and corresponding amount of contingency as a percent of the base estimate. Values are displayed based on the queried projects and chosen confidence level. Users can view the results of recommended contingency amounts given their current project estimate score. (RS131-1, p. 19)