Abstract
Shipping movements are operated in a complex and high-risk environment. Fatal shipping accidents are the nightmares of seafarers. With ten years' worldwide ship accident data, this study develops a binary logistic regression model and a zero-truncated binomial regression model to predict the probability of fatal shipping accidents and corresponding mortalities. The model results show that both the probability of fatal accidents and mortalities are greater for collision, fire/explosion, contact, grounding, sinking accidents occurred in adverse weather conditions and darkness conditions. Sinking has the largest effects on the increment of fatal accident probability and mortalities. The results also show that the bigger number of mortalities is associated with shipping accidents occurred far away from the coastal area/harbor/port. In addition, cruise ships are found to have more mortalities than non-cruise ships. The results of this study are beneficial for policy-makers in proposing efficient strategies to prevent fatal shipping accidents.
Original language | English |
---|---|
Pages (from-to) | 92-101 |
Number of pages | 10 |
Journal | Accident Analysis and Prevention |
Volume | 76 |
DOIs | |
Publication status | Published - 1 Jan 2015 |
Externally published | Yes |
Keywords
- Logistic regression
- Mortality
- Negative binomial regression
- Shipping accident
ASJC Scopus subject areas
- Human Factors and Ergonomics
- Safety, Risk, Reliability and Quality
- Public Health, Environmental and Occupational Health
- Law