Modelling Road Traffic Crashes Using Spatial Autoregressive Model With Additional Endogenous Variable

Abstract

Road traffic crashes have become a global issue of concern because of the number of deaths and injuries. The model of interest is a linear cross sectional Spatial Autoregressive (SAR) model with additional endogenous variables, exogenous variables and SAR disturbances. The focus is on RTC in Oyo state, Nigeria. The number of RTC in each LGA of the state is the dependent variable. A 33×33 weights matrix; travel density; land area and major road length of each LGA were used as exogenous variables and population was the IV. The objective is to determine the hotspots and examine whether the number of RTC cases in a given LGA is affected by the number of RTC cases of neighbouring LGAs and an instrumental variable. The hotspots include Oluyole, Ido, Akinyele, Egbeda, Atiba, Oyo East, and Ogbomosho South LGAs. The study concludes that the number of RTC in a given LGA is affected by the number of RTC in contiguous LGAs. The policy implication is that road safety and security measures must be administered simultaneously to LGAs with high concentration of RTC and their neighbours to achieve significant remedial effect.

Analyzing urban crash incidents: An advanced endogenous approach using spatiotemporal weights matrix

“…Wang, Chen, et al (2019) To the best of the author's knowledge, there is limited research on the development of an SWM based on endogenous variables, particularly in the context of utilizing these endogenous variables in regression models to analyze urban crashes. Olubusoye and Salisu (2016) constructed an SWM incorporating endogenous variables such as travel density, land area, road length, and population. They subsequently applied these endogenous variables within a spatial autoregressive (SAR) model to identify hotspots and explore the influence of crashes in nearby local government areas on the crash frequency in other areas.…”

“…Studies by Olubusoye and Salisu (2016), Pljakić et al (2019), Wang, Chen, et al (2019), andSandoval-Pineda et al (2022) have utilized endogenous variables to generate SWM and analyze crash data at a macro-scale traffic analysis zone (TAZ). These SWMs are uniformly weighted for all neighboring TAZ units.…”

A robust spatial autoregressive scalar-on-function regression with t-distribution

“…As the SAR model is representative among these spatial models and has been widely studied, we mainly focus on the SAR model and formulate our new model. Applications of the SAR model can be seen in Case (1991), Topa (2001), and Olubusoye et al (2016), among others. Estimation methods for the SAR model are described in Ord (1975), Lee (2004), Kelejian and Prucha (2001), Lee (2007) and Lesage and Pace (2009).…”

A flexible spatial autoregressive modelling framework for mixed covariates of multiple data types

“…It has become popular to model activities in regional economies, social networks and spatial geography (Case (1991); Topa (2001); Olubusoye et al (2016)). Using a spatial weight matrix and a spatial lag parameter, the SAR model incorporates the network structure into a classical linear model.…”