Evaluation of EVM/ES forecasting methods in hospital construction projects

Abstract

The objective is to evaluate 22 time and 12 cost forecasting methods based on Earned Value (EVM/ES) under three assumptions of future performance and
in terms of the opportunity, accuracy, and stability of the forecasts. Little attention has been paid in the literature to future performance assumptions and
evaluation under these parameters. Method: A total of 6,951 entry data points were used applying statistical measures of error and dispersion based on
empirical evidence from 34 hospital construction projects in Mexico. Findings: Using contractor payment estimates as Actual Cost (AC), the more predictive
time forecasting methods are those under the assumption that future performance will be as planned and, for cost, under the SPI(t) Schedule Performance Index. Relevance: The research shows the conditions under which the diverse methods offer timely and realistic predictions to be used as a reference in the monitoring and control of hospital construction projects. The limitation of the research consists of using durations and costs from completed projects without considering the possible changes formalized in contractual modifications throughout project’s execution. Future research using EVM/ES input data from the project owner's perspective is suggested.