Sample Size Analysis for Asphalt Pavement Quality Control

Abstract

A review of the quality assurance practices of Alberta Transportation & Utilities was undertaken for each of three quality test measures: (1) Asphalt content; (2) percent compaction; and (3) aggregate gradation. The primary objective of the study was to determine how reducing the number of tests per lot affect the lot mean values and the degree of confidence in the quality results. The secondary objective was to determine the effect a change in the sample size would have on the quality-related price adjustments to the contractor. Two methods were used to compare the number of samples and their resulting mean values. First, the minimum number of samples required to stay within the variance allowed in the specifications, given the variance of the samples, was determined. Second, the mean values derived from sample sizes of five, four, and three were compared using one-way ANOVA and independent T-tests. The study found that the minimum number of tests for asphalt content, percent compaction, and aggregate gradation could be reduced. The effect on the price adjustments to the contractor given the reduced number of test samples was negligible.