Page Title

Exploring the Relationship of Leadership Skills and Knowledge to Leader Performance

October 25, 2001 

"Leadership characteristics and capabilities have long been a topic of interest in the leadership literature.  In sight about the nature of the leader capacities that impact leader behavior and performance is invaluable for selecting and developing necessary skills for leaders to further the goals of the organizations in which they work.  Many studies have taken multivariate approaches, examining the relationships of leader characteristics to a variety of criteria, including measures of leader performance.  These empirical studies have examined leader capacities as cognitive abilities, motivation, and personality attributes required for effective leadership performance. 

The goals of the current study addresses several issues.  First, This study addresses the criterion-related validity of constructed response measures of key leader capabilities in a large sample of leaders drawn from multiple organizational levels within the U.S. Army.  These types of measures offer a more direct assessment of leader skills and knowledge than do subordinate measures.  Second, this study provides initial empirical tests of care aspects of the Mumford, Zaccaro, Harding, Jacobs, and Fleishman (2000) leader capabilities model, a mediated model of leadership.  Third, and finally, the results of this study are discussed in the context of related leadership research to examine their generalizability to an Army civilian leader sample. 

The sample of leaders used in this study included only a portion of the 1,807 Army officers who participated in a broader research effort.  These officers are distributed across six grade levels and ten Army posts throughout the United States, the majority of whom are from combat, combat service, and combat service support in other branches of the service.  About 5% of the officers are distributed across positions in other branches.  Demographics for this sample such as age, gender, source of commission, and time in service are described in Zaccaro, Mumford, Connelly, Marks and Gilbert (2000).  Officers typically participated in this study during the time they completed courses in various Army schools.

Leaders' complex problem-solving skills are measured using two different constructed response measures.  Both of these instruments present leadership scenarios for which some action is called.  The first one presents a hypothetical combat scenario and subjects are asked to (1) respond to a series of questions asking them to define the problem, (2) list what information is needed to solve the problem and how different information is related, and (3) describe action plans, implementation strategies and follow-up strategies.  Subjects responded to a total of ten questions designed to tap nine aspects of problem solving, from problem definition and identification of key information, solution implementation and follow-up.  The second problem-solving measure presented two leadership scenarios, one combat oriented, and one dealing with reorganization of a hypothetical Army.  No specific questions or cues were provided other than asking the question, 'What would you do in this situation?'  Each scenario was rated on problem-solving constructs including attending to multiple pieces of information and coordinating and explicating relationships among multiple elements of the problem. 

The third constructed response measure tapped solution construction skills, and was intended to tap leaders' skills in attending to situational restrictions critical in problem solving - time frame for solving the problem, and the nature of the leader's personal and/or organizational goals. 

Several general cognitive abilities are assessed in this study.  The verbal reasoning sub-test of the Employee Aptitude Survey or EAS contains six five-item logic problems designed to measure verbal reasoning (Ruch & Ruch, 1980). Divergent thinking skills are measured using Christensen, Merrifield, and Guilford's (1953) Consequences-A test with an adapted scoring procedure.  The third measure of cognitive abilities, the Alternative Headlines Test (Guilford & Hoepfner, 1966) assesses officers' writing with respect to three dimensions of writing delineated by Hayes and Flowers (1986): planning, generation, and revision, as well as quality and originality. 

Two primary types of leadership criteria are employed in this study.  The first criterion was a self-report biodata measure of career achievements in which most of the items were capable of being verified.  Effective problem solving, or quality ratings of solutions to the cued and uncued leadership problems, was used as a second criterion measure.  The sampling strategy limited use of this second criterion in some of the data analyses. 

As expected, the leader skills and knowledge measures all correlated significantly with the leader achievement and solution quality criteria, producing moderate to large positive correlations ranging from .17 to .51.  Cognitive abilities had moderate to strong positive correlations with the criteria ranging from.13 to .59, with the exception of the correlation of verbal reasoning and leader achievement.  The correlations of motivation and personality variables with leader achievement and solution quality were low to moderate in size.  

The findings in this study supported two of the hypotheses relating to providing additional validity evidence for constructed response measures of critical leaders skills and to testing aspects of Mumford et al.'s (2000) leader capacities model of leadership.  Another important implication of this type of skills-based leadership model is that it provides clues about why leaders do well across a variety of challenging situations.  Leaders tailor their behavior to the specific situations they encounter.  

Implications of these results should be viewed in light of several caveats.  First, the leadership model and measurement framework employed in this study do not include all possible individual capabilities that might influence leader performance.  Second, leaders at different organizational levels are not compared to assess potential differences in the relative contributions of leadership skills and knowledge to performance.  Third, results in this study were not cross-validated.  Fourth, there is a possibility that results of analyses using the solution quality criterion were influenced by method bias since predictors and the criterion were both constructed response measures.  Fifth, casual inferences cannot be made from the mediation analyses because requirements for making such statements were not met (James, Muliak, & Brett, 1982).  A full model would need to be tested via casual analysis.  This might be a useful next step in further delineating relationships among the predictor variables and the leadership criteria.  Although the size of the samples used in this study are not large, similarity in the results for analyses using the leader problem-solving skills and knowledge measures lends credibility to the stability of these findings.   (Connelly, Gilbert, Zaccaro, Threlfall, Marks and Mumford, 2000, pp. 65-86)." 

Reference:  Connelly, M.S.; Gilbert,  J.A.; Zaccaro S.J.; Threlfall, K.V.; Marks, M.A.; & Mumford, M.D., (2000).  Exploring the relationship of leadership skills and knowledge to leader performance.  The Leadership Quarterly, 11(1). 

-----------------------------------------------------------------------------------------------------------------

The Leadership Quarterly is available on loan from the OSU Leadership Center.  A complete listing of all the Leadership Center's resources is available on our website http://leadershipcenter.osu.edu/ 

Leadership Discoveries is a free monthly e-mail mailing about leadership research.  If you have any colleagues who would like to receive Leadership Discoveries, please have them send an e-mail to flynn.61@osu.edu with the message, Subscribe Leadership Discoveries.

Back to Top >

Created: 2008-02-26, Updated: 2009-02-18