Potential Biases in Leadership Measures: How Prototypes, Leniency, and General Satisfaction Relate to Ratings and Rankings of Transformational and Transactional Leadership Constructs

Abstract

Eighty-seven respondents completed either a graphic rating or a forced ranking questionnaire describing their immediate superior. Five leadership scales were embedded in each questionnaire. Three represented transformational leadership constructs (charismatic leadership, individualized consideration, and intellectual stimulation); two reflected transactional leadership constructs (contingent reward and management-by-exception). Appended to each questionnaire were five additional scales. The items constituting these scales measured two outcomes-satisfaction with the leader and effectiveness of the leader. The remaining scales measured each participant's leadership prototype, the participant's tendency to be lenient in his/her ratings, and a general measure of satisfaction. As expected, the intercorrelations among the factor scores representing the transformational and transactional leadership constructs were reduced substantially by using the forced rankings as compared with the graphic ratings. Also, the magnitude of the relationships among leadership and outcome factor scores was reduced, on average, when using the forced rankings. Prototypicality factor scores were more highly correlated with factor scores reflecting transformational than were factor scores portraying transactional leadership. The tendency of participants to be more or less lenient in their ratings or rankings and their general level of satisfaction were of little or no consequence to the intercorrelations among the leadership and outcome factor scales.