Safety Leadership:
Challenges and Solutions
(ASSE Professional Development Conference
Orlando Florida
June, 2000)

The Research

Terry E. McSween

The result was a much more reasonable field of research. In short, we found 19 field and experimental studies, many of which were conducted by Judy Komaki – one of the principle researchers whose work was instrumental in establishing the field of behavioral safety in the early and mid 70's. These studies examined the differences between effective and mediocre leadership. The research showed that effective leaders have followers who perform better and have more positive attitudes toward both their leaders and their organizations. Most importantly, the research shows what managers should do, when they should do it, and what they should avoid doing.

Many of the studies were field studies that involved recording actual samples of the interactions of leaders with followers (the terminology used in most of this body of research). The researchers then went back over the transcripts and categorized the statements made by the leaders and followers. They then looked for differences in the types of interactions between effective leaders (those with top performing organizational units) and mediocre leaders (those with lower performing organizational units). The studies were done in a wide variety of different organizations such as financial organizations, a motor pool in the US Army, theater groups, and America's Cup racing sailboats. In addition, the studies were done across several different cultures, including the US, Sweden, and Israel.

In classifying the interactions of leaders with followers, the researchers used a classification scheme based on behavioral concepts. They classified the responses as antecedents, monitoring (behavior), or consequences. Antecedents were statements made by leaders that would provide guidance or direction, often instructions, reminders, suggestions, goals, or statements about their expectations. Monitoring was defined as seeking information about performance, which could be done either through direct observations or questions about performance. Consequences were statements that leaders made about performance. They could be positive feedback, indicating approval, neutral feedback, statements that reflected the leader's observation of the followers behavior, or negative feedback, which indicated disapproval or that the follower needed to do something differently. (We should note that this classification scheme was based on practical considerations of the research and was, therefore, based on what was said or done, rather than on its functional impact on the follower. In other words, consequences were identified on the basis of the response topography, not whether the followers' behavior changed in any way.) The researchers refined their definitions and trained observers to achieve a high level of inter-observer reliability, meaning that two observers scoring the same sample agreed on how each interaction was scored more that 90% of the time.

One of the experimental studies in particular illustrates the general findings of this important body of literature. Komaki, Desselles, & Schepman (1988) demonstrated the impact of leadership practices on both the performance and the attitudes of employees. In this study, the employees in a post office were randomly assigned to one of three groups. A manager was then trained either to (1) monitor, or (2) provide feedback without monitoring, or (3) monitor and provide feedback to employees.

The experimenters then examined the impact of each these leadership practices on the accuracy of the employees' mail sorting task. The employees who received feedback with monitoring (in the form of direct observations) performed at a significantly higher level of accuracy than the other two groups. The experimenters then met with employees and discussed their reaction to their supervision. The employees who received feedback with observations made significantly more positive statements and significantly fewer negative comments, as compared with either of the other two groups. In short, when the leader directly observed employee performance and provided specific feedback based on those observations, the employees performed better and had more positive attitudes towards their leadership.

Paper presented at the American Society of Safety Engineers annual Professional Development Conference in Orlando, in June, 2000. Included in the conference Proceedings.