Logistic Models in Crash Severity Classification on the Basis of Chosen Road Characteristics.

Auteur(s)
Nowakowska, M.
Jaar
Samenvatting

The paper discusses the influence of some road characteristics of two-directional one-lane roadways on traffic safety. The research work is done onthe basis of real data of multi-vehicle accidents without pedestrians on non-urban roads for a chosen region of Poland that is in charge of one road administration and that can be considered as homogenous in a geographical and climatic sense. The data about accidents come from the Police Road Accident Database whereas the data about roads come form the computer Road Dataset Bank, paper road documentation and local visions of important places. Data cleaning (including cross-checking) was processed before the investigation starts. In order to complete the research tasks, computer programs have been elaborated which enabled to join together data about an accident with road data of the accident location. Logistic regression in theform of proportional odds ratio model and continuation ratio logit modelswas applied to the analysis. The accident severity put in an ascending order (from minor, through serious to fatal one) is a target variable in thelogistic models. The set of explanatory variables consists of road designelements (e.g. vertical and horizontal curve) as well as road access (e.g. bus stops, public and private access lanes) and land use (e.g. gas stations, parking-places) information. The results indicate which features have statistically significant influence on an accident severity and what values of these factors create a certain threat level on a road.

Publicatie aanvragen

7 + 0 =
Los deze eenvoudige rekenoefening op en voer het resultaat in. Bijvoorbeeld: voor 1+3, voer 4 in.

Publicatie

Bibliotheeknummer
C 48301 (In: C 47949 DVD) /80 / ITRD E854005
Uitgave

In: Compendium of papers DVD 89th Annual Meeting of the Transportation Research Board TRB, Washington, D.C., January 10-14, 2010, 19 p.

Onze collectie

Deze publicatie behoort tot de overige publicaties die we naast de SWOV-publicaties in onze collectie hebben.