Research - Others' Research - Detail

Leadership Impact: An Empirical Test of A Predictive Framework

Download a Printer Friendly Version (PDF)
TITLE: Leadership Impact: An Empirical Test of A Predictive Framework
RESEARCHER: Paula J. Crnkovish and William S. Hesterly
Department of Management
School of Business
University of Utah
Salt Lake City, Utah 84112
Presented at the 24th Annual Meeting of the Decision Sciences Institute, March 1993.

Hypothesized that leader practices, managerial discretion, intent to change, and romance of leadership are associated with high impact on performance.

Seniors (N = 129) provided information on their current or most recent manager, across a variety of industries. There were 93 male and 36 female managers ranging in age from 20 to 84 (mean = 40.1) years. Impact included four items designed to assess the positive impact a manager has in his/her organization (alpha = .83). Romance of leadership was measured with an 11-item shortened version of the romance of leadership scale (Ehrlich, Meindl, and Viellieu, 1990) (alpha = .82). Discretion measured with six items the latitude the manager had to give rewards, make decisions, and allocate resources (alpha = .72). Need for change measured with four items the employee's perceived need for the organization to make sweeping changes (alpha = .69). Leader practices was measured as a composite scale using the 30-items from the LPI (alpha = .95).

Leader practices were strongly correlated with leadership impact (.75), and moderately correlated with discretion (.24) and romance of leadership (.19), but not need for change (-.13). All five individual components of the LPI were significantly correlated with leadership impact. Romance was not significantly correlated with leadership impact. Multiple regression analyses revealed that leadership impact as predicted by the other variables was highly significant (R2 = .63, p < .001), although much of the predictive power stemmed from leadership practices. The age of the manager did not seem to play in role in predicting leadership impact, not did the manager's gender.

Post hoc analyses revealed that modeling and enabling contributed the highest standardized beta weights in predicting impact. The authors also comment: "One implication of this study that may be useful for future studies was the failure to find a strong positive relationship between romance and practices observed. Previous criticisms that the romance of leadership may inflate or deflate the followers' view of leadership do not appear to be an issue here" (p. 14).