ABSTRACT
This research aimed to study the relationship between speed limits and speed variance in the traffic stream and to explore the significant factors that might affect speed variance on multilane highways in Jordan. This study also developed models to get the most accurate prediction with a higher coefficient of determination and to help agencies better-set speed limits to reduce speed variance and their associated crashes. Field measurements and observations were performed at twenty-five multilane highway segments during off-peak hours. Data include spot speeds, speed limits, lane width, number of lanes, and presence of roadside barriers, which were used to determine if its effect was statistically significant on speed variance.
Multiple Linear Regression and One-Way ANOVA were used to analyze the data. It was found that the speed variance was mainly dependent on design speed minus speed limit (DS-SL), the presence of roadside barriers, and the number of lanes with R2 = 0.835 at α = 0.05. Also, results indicated that as the number of lanes increases, the speed variance decreases. Moreover, the presence of roadside barriers reduces speed variance. This paper provides beneficial information to authorities and policymakers who set speed limits on multilane highways.
