Methods for Accumulating Construction and Material Quality Test Results and Their Effect on Acceptance Decisions

Abstract

Many state departments of transportation are currently shifting more acceptance sampling and testing responsibilities to contractors because of a shortage of personnel at highway agencies and intensive construction schedules. The practice of using contractor-performed test results for acceptance decisions is referred to here as contractor acceptance testing (CAT). Studies have shown the need for improving the power of the statistical process used to validate the contractor’s test results. A potential improvement to CAT is to accumulate contractor’s and agency’s test results of consecutive lots to increase sample size and consequently increase the power of the statistical tests used to verify the contractor’s test results. In this study, two different types of cumulative sampling methods (i.e., continuous cumulative and chain lot) are applied to actual test results obtained from the Kansas Department of Transportation (KDOT), which includes contractor-performed test results and agency-performed test results from independent samples. The percent within limits (PWL) calculated according to KDOT’s acceptance practices was compared to the PWL obtained by the alternative methods for more than 90 projects. The results show that a chain-lot method with three accumulated lots is a plausible way to increase the power of the F-test and t-test. The contribution of this study is a practical sampling method for state DOTs to improve their CAT practices by increasing the power of the F-test and t-test used in the verification process while staying close to current agency’s practice in terms of pay factors.