Factors Affecting the Accuracy and Variability of Pavement Surface Evaluations and Ratings

Abstract

This research identified factors influencing the accuracy and variability of pavement surface evaluations and ratings (PASER). Survey and PASER data obtained from workshop participants were used to develop statistical models estimating three definitions of rater accuracy: (1) bias, which is the absolute value of the difference between a participant’s rating and the segment’s true rating as determined and agreed upon by PASER-certified instructors; (2) average participant bias and bias standard deviation; and (3) good/fair/poor category identification. The results indicate that pavement in good condition was rated more accurately, while pavement on the poor/fair condition boundary was rated less accurately. Participants were more accurate in assigning PASER ratings after receiving PASER-specific training. Additionally, raters that were more accurate were also more consistent in performing PASER ratings. Participants with engineering roles, such as engineer, engineer technician, or engineer assistant, were more accurate in assigning PASER ratings. In contrast, participants with leadership roles, such as supervisor, manager/foreman, or team leader/elected official, or less than 1 year of PASER rating experience were less accurate in assigning PASER ratings.