Development of Three Scales to Measure Leader Accountability
James A. Wood Jr.
The focus of this research is on the construction of three unidimensional summated rating scales—the Responsibility, Openness, and Answerability Scales—to measure the collective constructs that comprise Wood and Winston's (2005) definition of leader accountability. The development of these scales creates the possibility of moving the conversation about accountability from a reactive to a proactive posture, can guide organizations in selection or promotion decisions, and can provide organizations a framework by which to train and develop leaders in this critical area. A literature review that focused on the behavior of accountable leaders yielded pools of 26, 21, and 19 items respectively. The items were submitted to a panel of six experts, who reviewed them for relevance to the construct and readability, and who made suggestions for other items and general improvement of the scales. The scales were then pilot tested online by 148 participants. Factor analyses revealed that the item pools measured one construct in each of the scales. Reliability analysis revealed coefficient alpha scores of .9791 (Responsibility), .9884 (Openness) and .9844 (Answerability). In order to create a more efficient and parsimonious instrument and a familiar 100-point range of scoring, I reduced the number of items in the final scales to 10 each. I did so by eliminating items that appeared redundant or may be more difficult to understand. Follow-up principal components analyses also revealed that each 10-item scale measures one factor. Alpha scores for the 10-item scales were .9671 (Responsibility), .9684 (Openness), and .9755 (Answerability).
The fact that the overwhelming number of respondents was Caucasian (93%), and the population of leaders being evaluated predominantly male (71%) suggests the need for more research among a broader spectrum of cultural participants, with more studies of female leaders, and other means of data collection that do not require computer or Internet expertise. Additional needed research includes a series of convergence and discriminability studies, research into other expressions of accountability not measured by these scales, and a variety of causative and correlation studies that can help explain the role that leader accountability serves in leadership theory and organizational life.
Regent students, staff, and faculty: Available in full text from Regent University Library
Non-Regent researchers: Available in full text from UMI Dissertation Services