Integrating leadership research: a meta-analytical test of Yukl’s meta-categories of leadership
| Date | 05 September 2016 |
| DOI | https://doi.org/10.1108/PR-07-2014-0145 |
| Published date | 05 September 2016 |
| Pages | 1340-1366 |
| Author | Lars Borgmann,Jens Rowold,Kai Christian Bormann |
| Subject Matter | HR & organizational behaviour,Global HRM |
Integrating leadership research:
a meta-analytical test of Yukl’s
meta-categories of leadership
Lars Borgmann, Jens Rowold and Kai Christian Bormann
Center of Continuing Education,
TU Dortmund University, Dortmund, Germany
Abstract
Purpose –The purpose of this paper is to test the theoretical proposition, put forward by Yukl et al.
(2002), that the leadership constructs of transformational and transactional leadership, laissez-faire,
consideration, and initiating structure can be arranged into three meta-categories of leadership,
namely, relations, task, and change-oriented leadership.
Design/methodology/approach –After a comprehensive literature search, 1,402 correlations from
286 sources were obtained which referred to the above mentioned leadership constructs and outcome
criteria of effective leadership.
Findings –It was foundthat three meta-categoriesof leadership (i.e. relations,task, and change-oriented
leadership) were sufficient to explain the leadership constructs of transformational and transactional
leadership, laissez-faire, consideration, and initiating structure. Moreover, it was also found that
change-orientedleadershipwasmosteffectiveinpredictingthejobsatisfactionoffollowers.Incontrast,
relations-oriented leadership accounted for most of the variance in both commitment and job performance.
Research limitations/implications –In combination, these results allow for a more comprehensive,
integrative description of effective leadership behavior.
Practical implications –The results have implications for leadership assessment and development
as well as leader selection and feedback.
Originality/value –Applying Occam’s razor, and in contrast to the majority of prior leadership
studies, the present meta-analytical study provided evidence that three categories of leadership
behaviors allow for a precise and comprehensive description of effective leadership behavior.
Keywords Quantitative, Leadership, Leadership style, Advanced statistical,
Transformational leadership (TL), Leader behaviour
Paper type Research paper
In the past decades, numerous leadership theories have been developed, but these
theories and their respective leadership constructs were largely researched in relative
isolation from each other (Yukl, 1989; House and Aditya, 1997). Avolio (2007) criticize s
the lack of integration between leadership theories, and the way in which new theories
are often created without adequately assessing and comparing the content of existing
models. Also, several scholars have emphasized that leadership constructs such as
transformational leadership are not necessarily new, as their content has been used in
earlier studies and also in various current leadership theories (Humphreys and
Einstein, 2003; Antonakis and House, 2002; Sashkin 2004).
Personnel Review
Vol. 45 No. 6, 2016
pp. 1340-1366
©Emerald Group Publishing Limited
0048-3486
DOI 10.1108/PR-07-2014-0145
Received 8 July 2014
Revised 6 April 2015
23 August 2015
Accepted 6 October 2015
The current issue and full text archive of this journal is available on Emerald Insight at:
www.emeraldinsight.com/0048-3486.htm
This researchwas supported by the German ResearchCouncil (Deutsche Forschungsgemeinschaft,
No Ro 3058/5-1, Principal Investigator: Jens Rowold). The support of Ron Piccolo, Timothy Judge
with data collection assistance is gratefully acknowledged, as is the support of Anneke Soepper,
Tobias Antonik, and Jasmin Laurenawitz with data analysis. The helpful comments from Gary
Yukl on this manuscript were greatly appreciated.
1340
PR
45,6
In face of the multitude of leadership theories, significant overlap exists between
different theories, questioning the construct validity of these theories (Antonakis and
House, 2002; House and Aditya, 1997; Hughes, 1997; Hunt and Conger, 1999; Sashkin, 2004;
Yukl, 1989, 1999, 2002): for example, Seltzer and Bass (1990) reported high correlations
between subscales of transformational leadership and consideration (0.47 oro0.69;
po0.01). Geyer and Steyrer (1998) found a positive relationship between transformational
leadership and initiating structure. Antonakis and House (2002), Sashkin (2004), Hunt and
Conger (1999) also confirmed strong overlap between leadership constructs.
If different leadership constructs exhibit similarities and show significant overlap, it can
be argued that these constructs represent the same phenomenon, or can be reduced to
some broader underlying categories. Graen et al. (2010) explicitly called for an integration
of leadership constructs such as transformational and transactional leadership,
consideration and initiating structure, and leader-member exchange (LMX), since these
leadership constructs have been researched most extensively in the past six decades. For
example, the Web of Science (years 1955-2015) yielded 2,553 hits for transformational
leadership and 1,362 hits for LMX, respectively. In contrast, authentic leadership, a
construct that has only been researched recently, yielded only 225 hits. The importance of
the leadership constructs of transformational and transactional leadership ( Judge and
Piccolo, 2004), consideration and initiating structure ( Judge et al., 2004) and LMX (Gerstner
and Day, 1997) is further supported by the fact that meta-analyses were conducted for
these leadership theories, but not for other leadership constructs.
As an attempt to integrate existing findings on leadershipbehavior, Yukl et al. (2002)
presented a hierarchical taxonomy with three meta-categories (task, relations, and
change behavior)of leadership. This taxonomy provides a parsimonious and meaningful
conceptual framework that shows how the leadership behaviors (i.e. transformational
and transactional leadership, laissez-faire, consideration, and initiating structure as
well as LMX) are interrelated theoretically. It combines the simplicity of a few
meta-categorieswith the explanatory power of specificcomponent behaviors. Potentially,
this taxonomy could advance leadership theory since it would allow for a more
parsimonious classification of leadershipconstructs and theories. However, no empirical
test has been conducted to test this taxonomy. Moreover, Yukl’s classification is limited
in that it lacks possible mediators between leadership constructs and the criteria of
leadership effectiveness. As such, the present meta-analytical study is aimed at
contributing to the leadership literature in several ways: by exploring the factorial
structure of the meta-categories of leadershipwhich are in line with Yukl’s taxonomy, by
considering the relations with relevant organizational criteria, and by testing an
extension of Yukl’s taxonomy (i.e. possible mediating mechanisms).
The following steps were deemed necessary for this purpose. First, the leadership
theories and constructs that are under focus are reviewed on a macro-level. Next, a
theoretical discussion about differences and similarities between these leadership
theories is provided. Also, it is clarified that the leadership constructs included in this
paper fit Yukl’s meta-categories of leadership behavior well. Next, the empirical
literature on transformational and transactional leadership, laissez-faire, consideration,
and initiating structure as well as LMX is then meta-analyzed. Meta-analytic structural
equation modeling (MASEM, Hunter and Schmidt, 2004; Viswesvaran and Ones, 1995)
was used to explore potential overlap between the different leadership constructs under
investigation, in order to test Yukl’s taxonomy of leadership and to investigate
relations to criteria of leadership efficiency. The resulting categories of leadership were
tested for relative criterion validity.
1341
Integrating
leadership
research
To continue reading
Request your trial