The research was focused in a multiple correspondence analysis. We analyzed the
relationships between accident rates and road infrastructure variables in a municipality of
Soacha (with more than five hundred thousand inhabitants in the suburbs of Bogota-
Colombia). In Colombia, the fatality rate is 13 deaths per 100.000 inhabitants. In Soacha
municipality the fatality rate is 10 deaths per 100.000 inhabitants. Soacha reports the
highest number of deaths in the region (ANSV, 2020). In this research we performed a
spatial analysis using a Kernel density graph to identify the areas with the highest
concentration of fatal crashes.
We found that five dimensions explain 70% of the variability of data. We made a crossdimension
analysis by pairs, to obtain information on all possible combinations between
that dimensions and used them as an input of a spatial distribution analysis of crashes in
the municipality. We also calculated the contour curves for density modeling. We found
that higher concentration of fatal crashes is around the highway “Autopista Sur”. These
crashes were associated whit crash hypothesis variables, immediate cause of the crash,
crash class, crash severity and type of vehicle.
