United States Department of Transportation
i
Evaluating Two Different Traffic Data Methods Based on Data Observed, Analysis of Provided Data – Final Report A
-
2021-11-01
[PDF-1.05 MB]
Details:
-
Creators:
-
Corporate Creators:
-
Contributors:
-
Corporate Contributors:
-
Subject/TRT Terms:
-
Publication/ Report Number:
-
Resource Type:
-
Geographical Coverage:
-
Edition:Final Report
-
Contracting Officer:
-
Corporate Publisher:
-
Abstract:Traditional traffic counts have long been utilized by U.S. DOT, the states, and MPOs to determine volumes of traffic. Recent developments in connected technologies have provided the potential for probe-based, or passive, measurement of traffic volumes that combines information on specific vehicles and advanced analytics algorithms to estimate AADT (Annual Average Daily Traffic). This task considers data provided from an evaluation of one such passive measurement system that produced AADT estimates from both the passive method and the traditional counts in 2019 for a sample of more than 800 roadways over 29 states, including a wide range of functional classifications and volumes. A statistical analysis was performed to determine if the data collected provided evidence that the mean (and median) AADT as determined by the passive method would be expected to be systematically different (higher or lower) than that of AADT based on traditional counts. Mean difference was evaluated through paired sample t-tests of base-10 logarithmically transformed AADT values, while median differences were evaluated through the Wilcoxon signed rank test. Overall, there was adequate evidence to reject the assumption of equality and instead conclude that the passive measurement mean (or median) AADT exceeded that of the traditional counts with at least 95% confidence [geometric mean difference estimated at 6.6 percent (p-value <0.001) / median difference estimated at 3.9 percent (p-value <0.001)]. More detailed comparisons were conducted by urban/rural status, functional classifications, and AADT volume ranges. For these comparisons, the statistical test results were adjusted for multiple comparisons to control the risk of erroneously concluding significant differences to no more than 5 percent. In this context, urban and rural, broad volume ranges (<5000, 5,000-50,000, and 50,000+), and especially toll roads provided strong evidence for higher values in the passive measurement. The comparisons by roadway Functional Class 1, 2, 3, 4 and 5, and for most smaller volume ranges (e.g., AADT range of 5,000), failed to reject the assumption of equal means (medians). The lack of significance for the smaller groups could be related to lower statistical power with fewer samples in each comparison. The overall significant difference between passive and traditional estimates was influenced by the very large observed difference for toll roads.
-
Format:
-
Funding:
-
Collection(s):
-
Main Document Checksum:
-
Download URL:
-
File Type:
