A team is using a predictive model to determine possible cost overrun scenarios for a disruptive technology project. The data being used is pooled from various projects implemented within the organization. What should the project manager do to increase accuracy?

To increase the accuracy of predictive models for cost estimation, such as analogous or parametric estimating, the historical data used must be highly relevant and comparable to the current project. According to PMI standards, these estimating techniques rely heavily on historical information from projects with similar parameters, scope, and complexity. Using data from projects with similar parameters ensures that the underlying variables driving costs are aligned, thereby improving the reliability and validity of the cost overrun predictions. Including dissimilar projects would skew the statistical model and reduce its predictive accuracy.

A. Including public domain projects introduces external variables, such as different organizational cultures, resource rates, and accounting methods, which can distort the model and reduce accuracy.
B. Projects still in progress lack final, actual cost data, making them incomplete and unreliable for accurately modeling final cost overruns.
C. Using only successful projects introduces survivorship bias, eliminating the very data (projects with actual cost overruns) needed to accurately model overrun scenarios.

1. Project Management Institute (PMI). (2017). A Guide to the Project Management Body of Knowledge (PMBOK® Guide) (6th ed.). Section 7.2.2.2 Analogous Estimating (p. 244) and Section 7.2.2.3 Parametric Estimating (p. 244). (Specifies that estimating is most reliable when previous activities are similar in fact and not just in appearance).

2. Project Management Institute (PMI). (2021). A Guide to the Project Management Body of Knowledge (PMBOK® Guide) (7th ed.). Section 2.4.3 Estimating (p. 55). (Discusses the reliance on historical data and comparable project parameters for accurate forecasting).