Abstract:
It is assumed that followers only observe a limited set of leader behaviours and, of those behaviours, followers are keenly attuned to leader actions that affect their performance and well-being at work. In this thesis, leader behaviour is explored through the design and psychometric evaluation of a positive supervisor behaviour scale (PSBS) across three studies. In Study One, exploratory factor analysis was conducted on the broader managerial practices survey (MPS) to identify salient supervisor behaviours from the perspective of followers without specific rater training or experience. In Study Two, the hypothesised four-factor structure of the PSBS and its construct validity were assessed using confirmatory factor analyses on office-based employees in New Zealand. In Study Three, the psychometric properties of the PSBS were further examined on office-based employees in the United States. Study Three tested the performance and stability of the measure on a geographically distant sample to re-examine construct validity using a second method and to evaluate the PSBS for measurement invariance between a New Zealand and a United States sample. Using data from Study One and Three, the predictive validity of the PSBS on follower outcomes was also assessed. Lastly, using data from Study Three, the PSBS, which is a measure of the frequency of positive supervisor behaviour was compared against a multi-dimensional scale of negative supervisor behaviour. Results showed that the PSBS reliably measures four behaviours that followers can distinguish without training or systematic observation. These behaviours are relevant to follower attitudes and work performance. Across the three studies used to design the PSBS, the scale showed excellent psychometric properties. Results from the predictive validity studies across geographically distinct samples showed that the PSBS reliably predicts satisfaction with leader and cognitive engagement. Results from a comparative study showed that a negative behaviour scale was superior to the PSBS in predicting follower satisfaction with a leader and task proficiency.