Providing a model for finding the contribution of each HSE factors affecting crash severity

Abstract

Background and Purpose: A better understanding of the factors affecting crash severity to cause engineers enables to take nessessary measures to and reduce crash severity. Unfortunately, in Iran, regard to increase capacity of the roads and constructing new roads, but mortality rate due traffic accidents is higher than other countries. In investigation traffic accidents, in addition to incidents, also the crash severity should be considered. The study of the relationship between crash severity and the various factors affecting it, is great importance. Therefore, The purpose of this study was to provide a model for determining the contribution of each HSE factors affecting crash severity by drivers and Priority setting of the factors by traffic safety experts. Materials and Methods: At first, factors affecting crash severity were collected using a literature review to design the first questionnaire. Then, for content validity of the questionnaire (factors), it was sent to 20 traffic safety experts to comment on the content validity of the questionnaire. Then, each of the final factors was weighed and prioritized with the Analytical Network Process (ANP) method with the help of the Super Decisions software. In the next step, using the approved factors by traffic safety experts, a two-part questionnaire (including demographic characteristics and factors affecting crash severity) was made to complete by 300 drivers. Finally, to assess the reliability of the questionnaire from the Cronbach's alpha test, To sufficient sample size from the KMO index,to identify the factor structure and check the factor loads from the EFA test and to determine the relationships between variables from Mimic model was used, which this model has both a reflective and formative shape, in which we can observe the relationship between observed and latent variables as well as predictive variables with latent variables. Finally, calculation of CFI, TLI, RMSEA and chi-square has been used to evaluate the model. Result: As a result of library studies, 61 factors affecting crash severity were gathered in four main criteria and 61 sub-criteria. In the next step, as a result of content validity, 2 sub-criteria were eliminated. Finally, the approved factors were categorized in a questionnaire with four main criteria and 59 sub-criteria. Then, as a result of risk factors weighting, prioritizing the factors affecting crash severity were safety, other factors, health and environment, respectively. Based on the EFA test results, the variables in each group were classified as follows: 1- Human group: two factors, 2- Environment group: four factors, 3- Vehicle group: one factor, 4- other factors group: one factor. In internal consistency reliability, internal reliability for all groups and the whole questionnaire (α <0.6) were confirmed. Also, according to the results of the factor load, the criteria like "third-party insurance", "road surface conditions", " aim using of the vehicle ", "driver's education", "driver's gender" were excluded from the study. Finally, in the designed model, only the predictive variable of the vehicle type could have a significant relationship with latent variables, especially with variables of vehicle and accident characteristics. Also, based on the fitting results of themodel, the fitting indices for the model did not achieve the desired amount and final model was confirmed. Conclusion: Based on the present study, the most important factors affecting crash severity were collected, prioritized and finally presented in the form of a model and questionnaire. Although the model was not approved; the studies that are carried out in the future can using the developed theory in this study, carry out a confirmatory factor analysis so that they may be able to validate this theory and, using of it, to evaluate drivers’ perspective about the factors affecting crash severity. Prioritized factors from the perspective of experts can be used to priority setting control measures to reduce crash severity.