| dc.description.abstract | Two surveys 13 years apart (2002 and 2015) using identical questionnaires collected data on crew from a Norwegian ship owner. The 2002 survey used reliability-tested scales from previous research, and the 2015 research used the same scales as the 2002 research. The surveys measured 10 different independent variables: 1) Satisfaction with safety activities, 2) Fatalism, 3) Knowledge/competence, 4) Conflict between work and safety, 5) Reporting culture, 6) Communications, 7) Management attitude towards safety, 8) Job satisfaction, 9) Attitudes towards safety rules and 10) Learning culture. Significant positive changes appeared for factor 1, 3, 4, 5, 6, 8 and 10 between 2002 and 2015, a significant negative change was found for factor 2 and no significant change was found for the factors 7 and 9. Models with the 10 independent “safety culture” variables were tested on four behavioural scales: 1) Positive management behaviour, 2) Precautionary behaviour, 3) Laissez-faire behaviour and 4) Laissez-faire behaviour under pressure, using regression analysis. The 2002 study produced reliable scales and provided a valid construct when tested on a sample of seafarers. The 2015 study largely confirmed both the scales and the scales effect on the behavioural variables used in the 2002 study. | nb_NO |