Author: Imen Keskes
Supervisors: Jose M. Sallan and Pep Simo
Defense date: 1st October, 2014
The thesis of Imen Keskes is a theoretical and empirical study of the relationship between transformational leadership (TL) and organizational commitment (OC). These two concepts are examined as multidimensional constructs and are empirically tested within a French speaking population. The main objective of this study is to examine the mechanisms through which dimensions of TL influence different forms of OC by testing the possible mediating role of leader member exchange (LMX). The present thesis consists of six chapters:
The first chapter introduces the concepts of TL, OC and LMX and the motivation for the research. It provides a brief summary of previous scientific work on the topic and it outlines the purpose and specific objectives of each chapter.
In the second chapter, a study and discussion of the relationship between leadership styles and OC dimensions is carried out. An extensive literature research is performed in order to understand leadership and OC as well as the relationship between these two concepts. Although there is considerable research which suggest that TL is positively associated with OC in a variety of organizational settings and cultures, there is little empirical research focusing on the precise ways in which style of leadership impacts employee OC. Some critics about previous empirical and theoretical studies are presented and a number of areas for future research are developed.
In the third chapter, the English TL Questionnaire (TLQ) is translated into French using a back-translation technique. The validity and reliability of this instrument within two samples of French speaking populations (Tunisia and France) is tested. Internal consistency is assessed using Cronbach alpha coefficient. The construct validity is verified using the exploratory and confirmatory factor analysis. The results support a four-factor model of TLQ which consists of vision, supportive leadership, intellectual stimulation and personal recognition. These findings have shown adequate psychometric properties and have provided preliminary support for the construct validity of the proposed French version of the TLQ.
In the fourth chapter, the English OC Questionnaire (OCQ) is translated into French and the validity and reliability of this instrument within the two samples of French speaking population is tested. The results have supported a 4-factor model of OCQ of the original questionnaire. These findings have shown adequate psychometric properties and have provided preliminary support for the construct validity of the proposed French version of the OCQ.
In the fifth chapter, some models that take into account the diverse dimensions of TL and the different forms of OC are proposed and tested by examining the role of LMX as a mediator between them. To verify the models, two types of validity tests are considered: the convergent validity and the discriminant validity. To test the direct hypothesis, the significance of the beta parameters in the structural models is assessed. To test the mediation hypothesis, two methods are used: Baron and Kenny method and a the comparison of nested structural models. The results confirm the significance of the proposed models. These findings are important since previous research did not test these variables together.
Finally, the last chapter summarizes the contribution of the thesis. The possible directions of further research work are also discussed in brief.