| Article ID | Journal | Published Year | Pages | File Type |
|---|---|---|---|---|
| 589212 | Safety Science | 2014 | 8 Pages |
•We collect 1769 questionnaire to gather information on driver’s characteristics.•We present four indicators to characterize driver based on their properties.•We employ cluster analysis and significant test to analyze the information.•We associate the indicators with driving behavior, accident and ticket rates.•We present a 4-level driver risk conditions as driver risk index from safety respect.
ObjectiveThis paper intends to investigate the existing relationship between drivers’ characteristics and their aberrant driving behavior (lapses, errors, and violations), accident and ticket rates. To achieve this, risky drivers s groups are identified with introducing driver’s risk index (DRI).Methods1769 questionnaires were collected from Iranian drivers to gain information on drivers’ personality, age, gender, education, driving behaviors (lapses, errors and violations), accident and ticket rates. Four indicators were used to describe the driver’s characteristics so that the whole combinations of driver’s characteristics and their relationship could be taken into consideration. K-means clustering and a non-parametric test were implemented to group the combinations within the homogeneous categories based on driving behavior, accident and ticket rates.ResultsThe mean age of respondents was 36.53 (Standard Deviation (SD) = 11.33) with mean driving experience of 10.50 (SD = 9.63) years. The mean kilometers driven was 24875.89 km (SD = 24658.73) for 3 years. Results of the significance test (p-values) showed that there are no differences among lapses and errors with pairwise comparison across the whole clusters, however, other factors showed the most significant differences for resulting clusters by k = 4. Consequently, an ordinal 4-level risk index from 1 “safe” to 4 “risky” were identified. Also, a validation was performed by 158 questionnaires in order to confirm the results.ConclusionThese ordinal levels can be used as a driver’s risk index (DRI) to assess the effect of driver’s characteristics on safety. The risk index would help to identify and target high risk drivers with safety Prevention programs.
