Analysing workplace violence towards health care staff in public hospitals using alternative ordered response models: the case of north-eastern Turkey.

The main objective of this article is to determine key factors that may have a significant effect on the verbal abuse, emotional abuse and physical assault of health care workers in north-eastern Turkey. A self-administered survey was completed by 450 health care workers in three well-established hospitals in Erzurum, Turkey. Because of the discrete and ordered nature of the dependent variable of the survey, the data were analysed using four distinctive ordered response models. Results revealed that several key variables were found to be a significant determinant of workplace violence, such as the type of health institution, occupational position, weekly working hours, weekly shift hours, number of daily patient contacts, age group of the respondents, experience in the health sector, training against workplace violence and current policies of the hospitals and the Turkish Ministry of Health.