Evaluating the Use of Contractor-Performed Test Results in Highway Construction and Material Acceptance Decisions

Abstract

Transportation agencies around the world are using contractor-performed test results in construction and material acceptance and pay decisions. This practice is termed contractor acceptance testing (CAT) in this paper. Although CAT has helped transportation agencies to deal with a shrinking workforce and intensive construction schedules, it has been controversial. This paper provides insights into the technical and conceptual pitfalls of CAT. The technical pitfalls are revealed through Monte Carlo–based computer simulations of pavement specifications that use CAT. These simulations show that the verification process, which relies on the statistical F-test and t-test, is unreliable and can lead to erroneous pay decisions. The conceptual pitfalls are revealed by demonstrating that CAT encourages a quality-control approach that focuses on defect detection and containment, rather than defect prevention. Finally, potential alternatives and improvements to CAT are provided.