This study aimed at construction workers to investigate the influence of safety leadership on workers' safety behavior and injury events. This study was conducted cross-sectionally, using 500 purposive selected construction workers from southern Taiwan as study subjects. A structured questionnaire was applied to collect information, including demographics, types of workplace safety leadership, safety behavior and injury events. Data analyses used statistical package SAS to perform one-way ANOVA test and logistic regression to examine association of safety leadership and unsafe behavior or injury events. The results showed that positive safety caring and safety management controlling were both correlated with reduction of unsafe behavior and yet the safety management controlling was better than safety caring (r=0.32 and 0.50, both p<0.0001). After combined the factors of safety caring and safety management controlling into four leadership types, the results showed that those who perceived high caring/high controlling from their forefront leaders posed better safety behavior than those who perceived low caring/low controlling or high caring/low controlling (both p<0.0001). Regarding to the association between safety leadership and lost work day injury, the results showed that lost work day injury was not significantly associated with safety caring, but significantly associated with management controlling (OR=3.88, 95%CI=1.86~12.73). Furthermore, in the four types of safety leadership, compared to high caring/high controlling types, those who perceived their forefront leaders as low caring/low controlling had significantly higher risk of lost work day injury (OR=3.75, 95%CI=1.11~12.63). This study found that safety leadership of forefront leaders was associated with subordinate workers' unsafe behavior and occupational injury and that, comparing to safety caring, safety management controlling was more effective. The best type of safety leadership is to combined with both high safety caring and high management controlling.
This study aimed at construction workers to investigate the influence of safety leadership on workers' safety behavior and injury events. This study was conducted cross-sectionally, using 500 purposive selected construction workers from southern Taiwan as study subjects. A structured questionnaire was applied to collect information, including demographics, types of workplace safety leadership, safety behavior and injury events. Data analyses used statistical package SAS to perform one-way ANOVA test and logistic regression to examine association of safety leadership and unsafe behavior or injury events. The results showed that positive safety caring and safety management controlling were both correlated with reduction of unsafe behavior and yet the safety management controlling was better than safety caring (r=0.32 and 0.50, both p<0.0001). After combined the factors of safety caring and safety management controlling into four leadership types, the results showed that those who perceived high caring/high controlling from their forefront leaders posed better safety behavior than those who perceived low caring/low controlling or high caring/low controlling (both p<0.0001). Regarding to the association between safety leadership and lost work day injury, the results showed that lost work day injury was not significantly associated with safety caring, but significantly associated with management controlling (OR=3.88, 95%CI=1.86~12.73). Furthermore, in the four types of safety leadership, compared to high caring/high controlling types, those who perceived their forefront leaders as low caring/low controlling had significantly higher risk of lost work day injury (OR=3.75, 95%CI=1.11~12.63). This study found that safety leadership of forefront leaders was associated with subordinate workers' unsafe behavior and occupational injury and that, comparing to safety caring, safety management controlling was more effective. The best type of safety leadership is to combined with both high safety caring and high management controlling.