Moving ahead by going back: improving assessment of transformational leadership with profile scores
University of New Brunswick
The assessment of transformational leadership has been plagued by two distinct, yet persistent criticisms. First, researchers have criticised transformational leadership assessment as being leader-centric, while positioning the follower as a variable that is acted upon (e.g., the follower is an outcome or a moderator variable). Secondly, researchers have criticised the method of transformational leadership assessment and the inability, through decades of research, to find a stable factor structure. The solution to this problem has been to use a single total transformational score in place of individual component scores. The dissertation presented here sought to improve the assessment of transformational leadership by addressing both of these criticisms. First, the follower was integrated into the assessment process by incorporating leader categorization theory, a follower-centric approach. Although this study was unable to provide support for this approach, the method by which researchers can incorporate both leader-centric and follower-centric approaches into a single hybrid approach was presented. Second, this study proposed that the contextual nature of leadership leads to the inability to identify a single stable factor structure when assessing transformational leadership. As an alternative to traditional factor analysis and subsequent regression analyses between latent factors and criterion variables, this study also used profile analysis to investigate the relationship between transformational leadership and student performance. The profile analysis identified unique leadership profiles that were able to distinguish between instructors in a high-average student performance group from instructors in a lower-average student performance group. The implications of the profile analysis approach for the assessment of transformational leadership is discussed regarding education and training.