Analysis of Motorcycle Crashes in Texas with Multinomial Logit Model

Abstract

62and motorcyclist fatalities increased by 31% (from 407 to 530) during the same period of time.Motorcyclists represent the most vulnerable of road users and are particularly susceptible to being seriously injured if involved in a crash. From 2003 through 2008, motorcycle injuries increased from 6,061 to 9,708 (60%). Motorcyclists made up 4% of all traffic-related crash injuries in 2008, doubling from 2% in 2003.Multiple factors contribute to motorcyclists' crash and injury severity including speeding, alcohol use, lack of helmets, and risky and negligent behaviors by motorcyclists and other road users. Risk seekers are naturally attracted to the sport of motorcycling and their attitudes may contribute to crash risk and injury severity. Mannering and Grodsky found that risk seekers knowingly engage in risky behavior in spite of understanding that the behavior may increase their crash risk (1). They also significantly underestimate the consequences of a crash (1).When compared with national averages, rates of Texas riders are higher for improper licensure, lack of helmet use, and alcohol consumption. Data from NHTSA's Fatality Analysis Reporting System show that in Texas in 2008• The proportion of fatally injured motorcycle riders who were not properly licensed was 31% compared with 26% for the whole country,• The proportion of fatally injured motorcycle riders who were impaired (with a blood alcohol concentration greater than or equal to 0.08 g/dL) was 39% compared with 28% nationwide, and • The proportion of fatally injured motorcycle riders who were not wearing a helmet was 57% compared with 41% nationwide.Because motorcyclists make up an increasing proportion of total traffic fatalities in Texas, the development of effective methods to reverse these trends and mitigate motorcycle crash and injury severity becomes increasingly important. Knowledge of the associated causes of motorcycle crashes and the factors that contribute to the severity of injuries to crash-involved motorcyclists is useful in suggesting approaches for reducing the frequency and severity of such crashes.In this study, crash data from police-reported motorcycle crashes in Texas were used to estimate multinomial logit (MNL) models for identifying differences in factors likely to affect the severity of motorcyclists' crash injuries. Because previous studies have documented important differences in crash severity between urban and rural roadways (2, 3), probabilistic models of motorcyclists' injury severity in urban and rural crashes were estimated. PREVIOUS STUDIESIncreases in motorcycle crashes and resulting deaths and injuries shed light on the need to more fully understand factors that contribute to motorcycle crash involvement and how differences in factors likelyMotorcyclists accounted for 15% of all traffic-related deaths in Texas in 2008. This proportion increased threefold during the past decade. Knowledge of the associated causes of motorcycle crashes and the factors that contributed to the severity of injuries to motorcyclists involve...

A comparative study on machine learning based algorithms for prediction of motorcycle crash severity

“…Geedipally et al . [54] investigated motorcycle crash severity in Texas, USA, using a multinomial logit model. Other numerous studies that used multinomial in their studies include Shaheed et al .…”

“…Moreover, the settlement type: such as city, town, or rural; also affects motorcycle crash severity [26,29,31,61,67]. Previous studies discovered that the location of the crash: such as the intersection; was associated with the increasing severity of motorcycle crashes [55,61,70,72], while others found that the road description was related to injury severity [31,54,57,59,63,72].…”

Understanding the Contributing Factors to Nighttime Crashes at Freeway Mainline Segments

“…The benefit of using MNL is that each injury outcome has an individual function while the ordered model has the same coefficient of the same variable for the injury outcomes. It is likely that the ordered model might overestimate the probability of the high injury level while underestimates the low injury level [7]. Moreover, the same variable might significantly impact one injury level and not the other one.…”

Analysis of factors affecting serious multi-fatality crashes in China based on Bayesian network structure

“…Except for the abovementioned mathematical models, other models used for accident analysis include ordered logit models, [8][9][10] multinomial logit, or probit models. 11,12 However, these statistical methods consider all factors to be independent rather than related. In reality, multiple factors contribute to serious multi-fatality crashes and these factors are often interrelated.…”