A dissertation completed at
The Center for Education, Widener University,
One University Place, Chester, PA 19013 		 
THE MEASUREMENT OF LEADERSHIP: THE CONSTRUCT VALIDITY OF THE LEATHERMAN LEADERSHIP QUESTIONNAIRE AND ITS SUBSCALE CORRELATION WITH ASSESSMENT CENTER MEASURES

A. Dianne Thomas, Ed. D.
May 1998

RETURN
edited 01/11/01

 


ABSTRACT This is a study of the construct validity of the Leatherman Leadership Questionnaire (LLQ). Canonical correlation was used to determine whether the 27 scales of the LLQ taken collectively would predict 10 assessment/development scales. Canonical correlation is a tool for measuring construct validity. This statistical method evaluates convergent and divergent validity between LLQ scales and assessment center scales simultaneously.

This study also investigated whether the LLQ was equally valid for males and females. In the past validation studies of the LLQ were investigated only with white males.

The LLQ is a knowledge-based paper-and-pencil test that is designed to test what individuals know and don't know in relation to key behaviors in 27 leadership skill areas. The LLQ consist of two booklets that are divided into 27 skill areas and a total of 339 multiple choice questions that test actual knowledge in skills like Persuading, Influencing, Problem Solving and Decision Making.

The assessment/development center is a method, not a place, and consists of a series of management simulation exercises which take place over a two-day period and are designed to simulate various aspects of management and professional responsibilities. Six individuals participate in the simulation activities and are observed by a team of three trained assessors who relate behaviors observed during the exercises to 10 professional skills or dimensions. Participants are evaluated in terms of their managerial strengths and areas needing improvement.

This study consisted of a diverse group which included non Caucasians, 104 participants, 32 females and 72 males. The participants served at least two years as a sales representative and the average years of service was 8.8 years.

The results of this study revealed that the LLQ can not be used to predict how well or how poorly participants might score on Vi assessment/development center tasks- or exercises. The assessment/development center subtests were better predictors of the LLQ and there were no LLQ predictors that were better as predicting assessment/development center subtests.

This research did indicate that there might be alternative interpretations of LLQ scores that have yet to be fully investigated that would provide companies and other institutions with other uses for the LLQ. These results might also encourage these same companies and institutions to include the assessment/development center methodology on their list of methods for presenting potential supervisory and managerial success.

This study will make a contribution to the field of leadership assessment because of its unique focus. It evaluated the relationship between the LLQ and the much-studied assessment center measuring technique. In addition, the LLQ had studied only white males while the population of this study was a diverse group which included nonCaucasians and females. And it was also revealed whether the LLQ was valid for women. As a result of this study institutions and organizations Vi will be able to determine if or how they should use an the LLQ and which candidates would participate assessment/development center.

TO TOP