A multiple correspondence analysis of at‐fault motorcycle‐involved crashes in Alabama

Abstract

According to the U.S. National Highway Traffic Safety Administration, in 2012, more than 4950 motorcyclists were killed in traffic accidents. Compared to passenger car occupants, mile for mile, motorcyclists are more than 26 times more at risk to dying in crashes. Due to the high fatality rate associated with motorcycle crashes, factors contributing to this type of crash must be identified in order to implement effective safety countermeasures. Given that the available datasets are large and complex, identifying the key factors contributing to crashes is a challenging task. Using multiple correspondence analysis, as an exploratory data analysis technique to determine the dataset structure, we identified the roadway/environmental, motorcycle, and motorcyclist-related variables influencing at-fault motorcycle-involved crashes. This study used the latest available dataset (2009 to 2013) from the Critical Analysis Reporting Environment database to study motorcycle crashes in the state of Alabama. The most significant contributors to the frequency and severity of at-fault motorcycle-involved crashes were found to be light conditions, time of day, driver condition, and weather conditions.

Linking Public Transport User Satisfaction with Service Accessibility for Sustainable Mobility Planning

“…MCA has been chosen as the statistical method to analyze data, given the nature of the datasets and the easiness to interpret the results in a graphical way. In the transport sector, MCA is widely used in road safety analysis, to evaluate the main contributing factors causing accidents [56][57][58] and investigating safety perception [59,60]. MCA has also been used to obtain insight into the quality aspects of PT systems: Lombardo et al [61] computed a composite indicator of customer satisfaction based on the key characteristics for assessing transport service quality found in literature; Grison et al [62] analyzed users' experience of PT route choice on the basis of context and users profiles; citizens' satisfaction for PT through a survey has been investigated in a study conducted for the city of Bologna in Italy [63]: MCA showed that punctuality, frequency of service and efficiency of buses were among the main factors that affected the overall satisfaction.…”

Investigating Underage Alcohol-Intoxicated Driver Crash Patterns in Louisiana

“…Most of the police-reported crash data are categorical in nature. Few studies have concentrated on MCA, an exploratory multivariate analysis technique used for handling large and complex nominal crash datasets without any predetermined hypotheses (31)(32)(33). Since development in 1970, MCA has been modified several times to faciltate recognition of variable attribute patterns and their interconnections in multidimensional datasets (34).…”

Uncovering road traffic crashes typologies using multiple correspondence analysis (MCA), in a low-resource setting

“…Over the years, one of the techniques that have proved to be useful in identifying RTC patterns and typologies in several regions of the world is the multiple correspondence analysis (MCA) [14,15]. MCA is a statistical visualization method for picturing the association between the levels of categorical variables (CVs).…”

“…In social sciences [16], marketing [17], economy [18], and health sciences [19], MCA has been used to detect individual and event profiles or typologies, alone or in combination with another technic such as hierarchical clustering. Several studies on road safety have highlighted the value of MCA in studying persons and events (accidents) typologies [14,15]. Some of the advantages of MCA that make it suitable for RTA research in low resource settings are the following: 1) it is specifically designed to deal with nominal or ordinal categorical data, the type of information usually collected in community-based surveys; 2) the availability of several low-cost or even free licensed statistical software that allows its application, and 3) the MCA low learning curve.…”