Optimal Scaling Categorical Principal Components Analysis: Road Traffic KSI Car Accidents in England (STATS19)

Abstract

Categorical principal component analysis (CATPCA) technique was applied in the road killed or seriously injured (KSI) car accidents in England based on STATS19 data so that the categorical variables of KSI car accidents can be transferred into few components with reduction of dimensionality. Finally selected 20 variables in KSI car accident database were divided to create four principal components by applying “optimal scaling CATPCA” procedure in SPSS. The statistically significant KSI car accident variables, particularly the most accountable categorical variables, were identified and quantified for developing models as well as leading to aims to reduce as well as to prevent the car accidents, particularly the KSI car accidents. It also leads to map out the possible safety improvement strategies as well as to inform the policymakers on how best to reduce the number and severity of car crashes.