Akinyemi S.G, Bada O and Ahmadu A.O

CITATION: Akinyemi S.G, Bada O and Ahmadu A.O(2025). Loglinear modelling of road accident fatalities in Nigeria (2018-2023). Frontline Professionals Journal 1(5), 16-23

ABSTRACT

In recent years, road accidents have claimed many lives. This fatalities on the road are caused by poor road conditions, lack of drivers’ licenses, alcohol, and other factors. However, this paper focuses on the number of fatalities caused by road accidents in Nigeria from 2018 to 2023. Data were collected from the office of road safety regarding cases of fatalities from road accidents across the six geopolitical zones from 2018 to 2023. The data were analyzed using log-linear modeling, which helped us select the best model to represent the data. R Statistics was used to analyze and fit various models based on the year, state and geopolitical zones. A Chi-square test for independence was conducted between the Year of Death, State and Geopolitical Zone. The fitted model was selected based on the AIC (Akaike Information Criterion) generated by the R package for categorical modeling. The findings showed that the model with the three factors (Geopolitical Zones, State, and Year) best explains the collected. It was concluded that the years of death by road accident are dependent of the Geopolitical zone. It is recommended that attention should be paid to road accident prevention to avoid unplanned expenses through supplementary budgeting for the unexpected rise in accident casualties.

Key Words: Geopolitical Zones, Akaike Information Criterion (AIC), Loglinear Modelling, Categorical Data, Simulation