Content uploaded by Michael B. Harari
Author content
All content in this area was uploaded by Michael B. Harari on Aug 27, 2021
Content may be subject to copyright.
Journal of Occupational and Organizational Psychology (2021)
©2021 The British Psychological Society
www.wileyonlinelibrary.com
Self-leadership: A meta-analysis of over two
decades of research
Michael B. Harari*, Ethlyn A. Williams, Stephanie L. Castro and
Katarina K. Brant
Florida Atlantic University, Boca Raton, Florida, USA
Self-leadership has been the subject of dozens of empirical investigations over the past
several decades and has emerged as a pivotal construct in the self-influence literature.
Despite the interest in the construct, the myriad of disparate variables studied and the
absence of a quantitative systematic review summarizing findings have combined to limit
our ability to cohesively interpret and draw meaningful conclusions from this large
literature. To address this, we carried out a meta-analysis of the nomological network of
self-leadership, encompassing effect sizes from 101 studies and 111 independent samples.
Drawing on social cognitive theory to frame our research questions and hypotheses, we
evaluate global self-leadership and its constituent strategies (i.e., behaviour-focused,
constructive thought, natural rewards) as predictors of job performance, self-efficacy, and
job attitudes. In addition to evaluating zero-order correlations, we use regression and
relative-weight analyses to evaluate the three strategies’ effects on the various outcomes
simultaneously, delineating their relative contributions. Our meta-analysis examines the
Five-Factor Model of personality traits as antecedents. We also observed evidence
suggesting that self-leadership’s relationships were moderated by national power
distance. Implications for research and practice are discussed.
Practitioner points
Self-leadership is meaningfully associated with conscientiousness, openness, extraversion, and
transformational leadership.
Fostering employee self-leadership may promote productive cognition, attitudes, and behaviors.
Self-leadership training programs can target specific strategies for training based on the desired
outcome.
Self-leadership –a process through which individuals exert self-influence over their
thoughts, feelings, and behaviours at work –is advanced as a pivotal construct in the
organizational sciences (Houghton & Neck, 2002; Manz, 2015; Manz & Neck, 2004;
Stewart, Courtright, & Manz, 2019). Whereas traditional forms of leadership focus on a
leader motivating subordinates, self-leadership involves employee self-motivation rather
than reliance on an external influence. In their reviews of self-leadership research, Neck
and Houghton (2006), Stewart, Courtright, and Manz (2011), and Stewart et al. (2019)
suggest that individuals can leverage self-leadership strategies to improve their work
*Correspondence should be addressed to Michael B. Harari, Florida Atlantic University, DeSantis Pavilion 201, 777 Glades Rd.,
Boca Raton, FL, USA (email: mharari@fau.edu).
An earlier version of the manuscript has been presented at the 2019 Southern Management Associated annual conference.
DOI:10.1111/joop.12365
1
attitudes and performance. With the recent focus on employee self-development (Holt,
Hall, & Gilley, 2018; Reichard & Johnson, 2011), research has emphasized self-leadership
as a central mechanism through which employees take personal initiative (Manz, 1992;
Stewart et al., 2011).
Self-leadership represents an internal source of leadership (Manz, 1986; Manz & Sims,
1980), operationally described as a process through which employees influence
themselves to achieve self-direction and self-motivation that they deem to be beneficial
to their work-relevant cognition, attitudes, and behaviours (Houghton & Neck, 2002).
There are three self-leadership strategies that can be used to achieve this self-direction and
self-motivation (Houghton & Neck, 2002; Manz & Neck, 2004; Neck, Manz, & Houghton,
2017): (1) Behaviour-focused strategies that heighten self-awareness for the management
of behaviours, including self-cueing, self-observation, self-goal setting, self-reward, and
self-punishment (self-correcting feedback); (2) Constructive thought pattern strategies
that focus on the power of positive thinking to form constructive habitual ways of
thinking that emphasize positive outcomes through positive self-talk, visualizing
successful performance, and evaluating beliefs and assumptions; and (3) Natural reward
strategies that involve building in inherently enjoyable aspects of work into activities to
help an individual feel self-controlling, competent, and purpose.
Reviews of the literature suggest that substantial progress has been made in defining
self-leadership, explicating its constituent strategies, and producing a large volume of
empirical research linking these strategies to various antecedents (such as conscien-
tiousness; Stewart et al., 2019) and work outcome s (such as job performance, self-efficacy,
and job satisfaction; Manz, 2015; Stewart et al., 2011). Yet, a number of gaps exist in the
literature. For example, Neck and Houghton (2006) review the self-leadership literature,
describing the normative model and its effects on outcomes. They examine self-leadership
in-depth at various levels, finding that self-efficacy appears to be a mechanism for the
influence of self-leadership. They also conclude that the area is under-investigated
empirically and conceptually, and notably advocate for examining effects of culture as
well as an expanded investigation of self-leadership and personality. Stewart et al. (2011)
conducted a broad review at various levels and affirmed the effects at the individual level
on attitudinal and performance outcomes. They concluded that self-leadership (and its
constituent strategies) is generally beneficial at the individual level and called for research
to examine more outcomes such as organizational commitment and to understand if more
self-leadership is always better. Finally, the review by Stewart et al. (2019) updates the
findings of Stewart et al. (2011) and goes on to examine paradoxes of self-leadership (e.g.,
self vs. others, resource strengthening vs. depletion) that should inform its practice. They
conclude that research on outcomes of self-leadership appears to be consistently positive
at the individual level and encourage research that seeks to further understand the role of
internal forces (such as traits) and the role of external forces (such as supervision and
culture).
In addition to these identified gaps, we also note that existing reviews have not utilized
systematic search strategies intended to identify all available studies dealing with self-
leadership (Neck & Houghton, 2006; Stewart et al., 2011, 2019). Thus, studies reviewed
might not be representative of the entire body of work (Harari, Parola, Hartwell, &
Riegelman, 2020). Further, these reviews have also been qualitative. Although valuable,
qualitative reviews lack empirical methods for objectively synthesizing and indexing
heterogeneity in findings across studies, limiting their utility in explicating the full state of
the literature.
2Michael B. Harari et al.
Although research findings suggest that self-leadership is generally positively
associated with favourable outcome variables, there remain both conceptual and
empirical gaps in our understanding of the self-leadership concept. We believe that many
of these identified gaps can be addressed through a systematic meta-analytic review of
empirical self-leadership studies. Our meta-analysis is guided by two main research
questions: What are the key traits of individuals who are self-leaders and what are the
strengths of the relationships between self-leadership and its constituent strategies with
outcomes? Meta-analytically addressing these research questions will substantially
advance understanding of self-leadership in multiple ways. Drawing upon insights from
social cognitive theory (SCT), we build hypotheses highlighting the way that self-
leadership reflects personal agency in relation to antecedents, moderators, and outcomes.
Thus, the primary purpose of this study is the utilization of meta-analysis, which will
deepen our understanding of correlates of self-leadership and its constituent strategies.
Our first contribution is the application of meta-analysis to empirically synthesize existing
research. Utilizing a comprehensive systematic search strategy, statistically combining
findings across studies, accounting for the biasing effects of statistical artifacts, and
indexing heterogeneity in effects, our quantitative meta-analytic review will help to clarify
population correlations between self-leadership and its various antecedents and
outcomes (Schmidt & Hunter, 2015).
Our second contribution relates to job performance, one of the most central
constructs in all of Industrial, Work, and Organizational Psychology (Harari & Viswes-
varan, 2018) and the focus of much self-leadership research. Even while a clear majority of
self-leadership studies point to a positive effect of self-leadership on performance (Manz,
2015), there remains ambiguity concerning population effect size magnitudes, hetero-
geneity, and the influence of methodological characteristics and substantive moderators.
The literature involving job performance, for example, reports null relationships (e.g.,
Andressen, Konradt, & Neck, 2012; van Kortenhof, 2013) and variation in rating sources
(e.g., self- vs. other-ratings of performance), which have not been considered in narrative
reviews. Further still, there have been limited efforts to explicate and test mechanisms
accounting for the self-leadership-job performance link. Based on SCT, which emphasizes
the role of personal agency in affecting outcomes such as performance, and following up
on a model initially proposed and tested in Prussia, Anderson, and Manz (1998), we test
the role of self-efficacy as such a mediator.
Another way in which we advance research on self-leadership is by conducting a more
fine-tuned analysis, investigating the relative contribution of each self-leadership strategy.
Self-leadership is a multidimensional construct, comprised of three strategies. Although
narrative reviews have described self-leadership and its relationships with various criteria
in broad terms (e.g., Manz, 2015; Neck & Houghton, 2006; Stewart et al., 2011), the
relative contribution of each strategy has not received systematic treatment. Based upon
predictions derived from SCT –that individuals exercise control over personal,
environmental, and behavioural factors –we anticipate that the various self-leadership
strategies differentially contribute towards self-leadership’s relationship with particular
outcome variables (e.g., behavioural strategies contributing more towards regulating
behavioural outcomes, cognitive thought strategies for cognitive belief outcomes, and
natural reward strategies contributing towards more emotional or attitudinal outcomes).
In order to understand why self-leadership influences various criteria, it is necessary to
understand which strategies drive the association of self-leadership with said criteria.
Further, as a normative model of self-influence (i.e., it prescriptively emphasizes ‘how
something should be done’; Neck & Houghton, 2006, p. 275), it is important to
Self-leadership meta-analysis 3
understand if these three strategies are all necessary to account for the influence of self-
leadership on outcomes, or if a more parsimonious prescription can be offered. We test
our hypotheses concerning relative contribution via meta-analytic regression and relative
weight analyses.
Finally, although variation in findings observed in the literature can be due to statistical
artifacts, it is also the case that findings could be affected by moderators that vary across
studies. Particularly relevant is the influence of cultural differences in power distance
(reflecting the extent to which autonomy is accepted in power-sharing; Hofstede,
Hofstede, & Minkov, 2005). Further, despite observations that cultural differences can
shape the influence of self-leadership in organizations (Stewart et al., 2019), robust
insights are not offered in the empirical data currently available. We again draw on SCT
and its focus on personal agency, suggesting that individuals respond to such
environmental factors; the role of power distance requires empirical testing (Stewart
et al., 2011, 2019).
Self-leadership literature review
Conceptualization, theoretical frameworks, and state of the field
Self-leadership distinctly emphasizes the ‘...purposeful leadership of self toward personal
standards ...’ (Manz, 1986, p. 858) and the personal meaningfulness and ownership of
one’s individual standards. While self-leadership involves some self-management actions,
it supersedes self-management. Rather than the self-management focus on reduction of
deviations from operating standards as defined by external sources (managers or the
organization), self-leadership is anchored in self-control processes that are internally
generated or superordinate standards of behaviour (Markham & Markham, 1998). Self-
leadership addresses what is to be done, why, and how to do it.
Self-leadership is grounded in part in SCT (Stewart et al., 2019), which explains human
behaviour by emphasizing personal agency with reciprocal links between personal,
environmental, and behavioural factors. People can respond to and exercise control over
these personal, environmental, and behavioural factors through five basic human
capabilities (SCT; Bandura, 1986). We first note how the five SCT capabilities map onto
the three self-leadership strategies, and then we use this to guide our review of the three
strategies below. The first two capabilities are symbolizing (processing visual models into
cognitive models that serve as a guide for future actions) and vicarious learning (learning
by observing the performance of others and the consequences of their actions), both of
which map on to constructive thought strategies as individuals use cognitive processing
to promote positive thoughts that can translate into positive future actions. Two
additional SCT capabilities, self-regulation (self-control of actions by setting internal
standards and evaluating the discrepancy between the standard and one’s performance to
make improvements) and self-reflection (reflect on past actions using perceptions to
determine the strength of their beliefs that they can achieve success in the future), map on
to behavioural strategies, as individual set standards for improvement and evaluate how to
improve on past actions. And the fifth SCT capability, forethought (self-determination
through planning actions, anticipating consequences, and determining desired perfor-
mance level), maps on to natural reward strategies, as individuals plan how to incorporate
aspects of their work or work setting into their tasks that make them have positive feelings
about what they are doing. With these linkages in mind, below we further develop how
employees might use these basic capabilities to self-influence using the three self-
4Michael B. Harari et al.
leadership strategies –for encouraging oneself to act in positive ways that lead to
successful personal and professional outcomes (Houghton & Neck, 2002).
Behaviour-focused self-leadership strategies. Behaviour-focused self-leadership strate-
gies are enabled by two SCT capabilities: self-regulation and self-reflection. In SCT, self-
regulation capability enables individuals to exercise control over their thoughts and
actions by using self-set standards and self-evaluative reactions to behaviours exerted
(Bandura, 2001). Self-reflection involves assessing the adequacy of actions by evaluating
their results (Bandura, 2001), helping people make sense of their experiences. These two
SCT capabilities (individual self-reflective and self-regulatory capabilities) are employed in
self-leadership behaviour-focused strategies such as self-observation, goal-setting, self-
punishment, and self-reward (Manz, 1986). The behaviour-focused action of self-
observation (raising one’s self-awareness about engaging in specific behaviours in terms
of when and why) is a manifestation of both self-reflection and self-regulation, as self-
observation involves standard-setting and evaluation of results, important for eliminating
unproductive behaviours (Manz & Neck, 2004) and focusing on productive behaviours
(Stewart et. al., 2011). Self-goal setting (involving identifying targets for an individual’s
efforts) is used as individuals set challenging and specific goals based on self-reflection to
raise their performance levels (Zimmerman, Bandura, & Martinez-Pons, 1992). Self-
punishment involves self-correcting feedback, a manifestation of both self-reflection
(introspection leading to identification of unacceptable past behaviour) and self-
regulation (adaptation of behaviours based on self-correcting feedback). Self-rewards
(support and provide incentives for desired behaviours and goal-attainment) is the final
behaviour-focused component, important for performance (Neck et al., 2017), and is a
manifestation of self-reflection, providing a positive outcome in response to meeting self-
set standards.
Constructive thought pattern self-leadership strategies. Constructive-thought self-
leadership strategies are enabled by two SCT capabilities: symbolizing and vicarious
learning. In SCT, the importance of self-perception of our own effectiveness is stressed
through the use of imagination and choice of thought patterns (Bandura, 2001). The SCT
capability of symbolizing, which is critical to allow processing and transforming visual
experiences into cognitive models that can guide individuals’ actions, allows individuals
to test possible solutions symbolically first in order to eliminate or accept them based on
thought processes (Bandura, 1986). The SCT capability of learning by observation
(vicariously) is also involved in self-perception, enabling individuals to accumulate guides
for initiating and controlling behaviour and thought patterns (Bandura, 1977) important
for the self-direction involved in constructive thought strategies. Constructive thought
strategies or thought self-leadership involves individuals managing their thinking
tendencies to direct focus to positive mental imagery and self-talk, helping to maintain
self-direction and self-motivation. Self-talk refers to what people say to themselves
covertly; mental self-evaluations to suppress negative internal dialogue and replace it with
positive or constructive self-talk for optimistic internal dialogue, encouraging persistence
in positive self-directed actions. The identification of alternative beliefs to challenge
currently held dysfunctional beliefs involves examining current thought patterns and
confronting and replacing beliefs that are dysfunctional and irrational with more positive
Self-leadership meta-analysis 5
constructive thoughts, contributing to a sense of personal control (Manz, Adsit, Campbell,
& Mathison-Hance, 1988).
Natural reward self-leadership strategies. Natural reward self-leadership strategies are
enabled through the SCT capability forethought. Forethought capability involves
individuals planning future favourable courses of action that will result in positive,
desirable outcomes (Bandura, 1986). Per SCT, competence and self-control are central to
forethought, providing autonomy to act on your own volition (Bandura, 2001; Deci, 1975;
Gagn
e & Deci, 2005). These elements are necessary for natural reward strategies, as
individuals must have competence to identify naturally motivating activities, and self-
control to incorporate these naturally motivating activities and tasks into their work.
When individuals focus on positive aspects of their work (natural rewards), they are
thought to feel greater satisfaction (Neck & Manz, 1996). Natural reward strategies involve
using forethought to build natural motivation into tasks to create motivation (and goal
internalization) based on inherently enjoyable aspects of an activity. The emphasis is on
intrinsic rewards that encourage individuals to perform work for its own value rather than
an external reward. These create feelings that translate into positive attitudes (Frese & Fay,
2001; Manz, 2015).
As described above, the five basic SCT capabilities underlie the three self-leadership
strategies, as well as the role of these strategies in understanding the fundamentals of self-
leadership. As we have discussed, SCT theorizes that individuals exercise control over
personal, environmental, and behavioural factors, and both personal and environmental
factors motivate individuals to self-lead. Stewart et al. (2011), Stewart et al. (2019) note
that some individuals have a natural inclination for self-leadership, and it is, therefore,
important to understand the factors that provide impetus. Thus, it is important to examine
personal factors represented by personality traits and environmental factors representing
context that might play a role in self-leadership (Stewart et al., 2019). In the sections
below, we review anticipated correlates of self-leadership, organized as antecedents (to
examine personality traits), outcomes, and cultural context (to examine environment),
which will be subject to integration in the present meta-analysis. We provide a selective
review of studies to present hypotheses before reporting our comprehensive meta-
analytic synthesis.
Self-leadership correlates
Personality antecedents
Theoretically relevant personal antecedent variables take the form of individual
differences in personality. As elaborated below, we draw upon the Five-Factor Model
(FFM) of personality –a long-established consensus model of normal personality traits. We
do not deny or downplay the importance and likely role of various compound personality
traits (e.g., proactive personality, core self-evaluations). However, the taxonomy of basic
traits offered by the FFM has proven robust and valuable across time, cultures, and
research domains (Digman, 1990) and offers ‘a useful delineation of traits that will provide
adequate coverage across the personality spectrum’ (Connelly, Ones, & Hulsheger, 2018,
p. 325).
The five personality factors include extraversion (e.g., enthusiastic, assertive),
emotional stability (e.g., even-tempered), agreeableness (e.g., courteous, cooperative),
6Michael B. Harari et al.
conscientiousness (e.g., diligent, achievement-oriented), and openness (e.g., curious,
creative; DeYoung, 2015; John, Naumann, & Soto, 2008). It is well understood that these
traits play important roles in motivational processes (Judge & Ilies, 2002), including those
related to self-regulation, social cognition, and intrinsic motivation –which are central to
self-leadership (Hart, Stasson, Mahoney, & Story, 2007; Houghton, Bonham, Neck, &
Singh, 2004; McCrae & L€
ockenhoff, 2010; Schaub & Tokar, 2005).
While some empirical evidence has found extraversion and conscientiousness to have
the strongest associations with self-leadership (e.g., Bailey, Barber, & Justice, 2016;
Furtner & Rauthmann, 2010; Houghton et al., 2004), we posit that the personality factor
with the strongest theoretical tie to self-leadership is conscientiousness (Guzzo, 1998;
Stewart et al., 2011). Due to tendencies associated with impulse control in the pursuit of
long-term goals, conscientious individuals might be likely to utilize the behavioural
strategies that are characteristic of self-leaders such as goal setting, employ planning
emphasized in natural reward strategies, and a have a positive focus as emphasized in
constructive thought strategies (DeYoung, 2015; John et al., 2008; Markham & Markham,
1998). We therefore hypothesize:
Hypothesis 1. The effect of the FFM traits on self-leadership is predominantly driven by
conscientiousness.
Behaviour-focused strategies. We expect that the effect of personality on self-
leadership would be differentially driven by different personality factors as a function
of the self-leadership strategy in question. We anticipate that behavioural strategies would
be primarily accounted for by conscientiousness. Conscientiousness speaks to goal-
directed impulse control (John et al., 2008), which highlights the importance of self-
regulation and self-reflective capabilities in SCT. Conscientious individuals are disciplined
and self-directed as the trait serves the role of ‘governing behaviour across long time
spans’, in order to achieve long-term goals (DeYoung, 2015, p. 45). Indeed, the
tendencies, motivations, and proclivities underpinned by conscientiousness manifest as
behaviours that are closely tied to behavioural strategies, such as goal setting (Judge &
Ilies, 2002). This leads us to hypothesize:
Hypothesis 2. The effect of the FFM traits on behaviour-focused strategies is predominantly
driven by conscientiousness.
Constructive thought strategies. We anticipate that constructive thought strategies
would be accounted for primarily by the personality factor that speaks to mental control:
openness. Those who are open are interested in ideas and enjoy engaging in information
processing (which requires motivation to expend required effort; DeYoung, 2015; John
et al., 2008). Open employees are therefore more likely to be motivated to examine and
replace destructive thought patterns with constructive ones (Connolly et al., 2018).
Further, because openness is associated with cognitive functioning (DeYoung, Peterson,
& Higgins, 2005), these individuals might also have a greater ability to monitor and manage
their thoughts and imagine themselves performing tasks before doing so. This
combination of ability and motivation could promote greater engagement in these types
Self-leadership meta-analysis 7
of self-leadership strategies as individuals employ vicarious learning and symbolizing
capabilities described in SCT. We therefore suggest:
Hypothesis 3. The effect of the FFM traits on constructive thought strategies is
predominantly driven by openness.
Natural rewards strategies. In terms of natural rewards, we anticipate that extraversion
would have the greatest effect. At the heart of extraversion (but particularly its enthusiasm
facets) is positive affect (Harari, Thompson, & Viswesvaran, 2018; Watson & Clark, 1997).
Perhaps due to this ‘crucial affective component’ (DeYoung, Quilty, & Peterson, 2007, p.
883), extraversion predicts affective states likely to be experienced at any given point in
time. Due to the manner in which positive affective states influence perceptions of
external stimuli and can guide attention towards affect-congruent elements of an
environment (Barsky, Kaplan, & Beal, 2011), natural rewards strategies should ensue. In
fact, due both to extraversion’s affective grounding and its reflection of a tendency
towards exploration of the physical world (and sensitivity to rewards therein, DeYoung,
2015), it is likely that extraverts would employ forethought capability in SCT to (1) seek
out pleasurable aspects of their environment and (2) be more likely to identify rewarding,
pleasurable aspects in their environment. We therefore hypothesize:
Hypothesis 4. The effect of the FFM traits on natural rewards strategies is predominantly
driven by extraversion.
Outcomes
Research on self-leadership frequently conceptualizes the construct globally, while
limited research has examined the effects of each of the three self-leadership strategies –
behaviour-focused, natural reward, and constructive thought. We anticipate that different
self-leadership strategies will account for a larger proportion of predicted variance in
different outcome variables, as discussed below.
Job performance. Job performance has been defined as ‘...scalable actions, behaviour
and outcomes that employees engage in or bring about that are linked with and contribute
to organizational goals’ (Viswesvaran & Ones, 2000, p. 216). Although job performance is
multidimensional, it can be characterized broadly, as dimension scores are saturated by a
substantive general factor (Harari, Reaves, & Viswesvaran, 2016; Viswesvaran, Schmidt, &
Ones, 2005). The focus in self-leadership on behaviours that influence goal attainment,
intrinsic motivation to attain goals, and thought patterns that encourage persistence
towards goals (Neck & Manz, 1996) suggests that self-leadership is an important
determinant of performance (important for personal agency in SCT).
A number of studies show positive relationships between various self-leadership
dimensions and job performance (Amundsen & Martinsen, 2014; Frayne & Geringer,
2000; Neck & Manz, 1996; Roberts & Foti, 1998). Other studies such as Andressen et al.
(2012) and Konradt, Andreben, and Ellwart (2009) show no association. Despite the
equivocality, self-leadership involves reflection and regulating one’s behaviour in a
8Michael B. Harari et al.
manner that should facilitate positive work-related behavioural outcomes (e.g., goal
setting; Locke, Shaw, Saari, & Latham, 1981). Given the very focus of the definition of self-
leadership on performance (‘... the process of influencing oneself to establish the self-
direction and self-motivation needed to perform’: Neck & Manz, 1992, p. 682), we
anticipate a positive association with job performance.
Hypothesis 5. Self-leadership is positively correlated with job performance.
In SCT, behavioural self-regulation assumes that when faced with discrepancies in the
progression towards goals there is a search for alternative goals and a promotion focus that
operates on the basis of accomplishments (Carver & Scheier, 1998). Controlled activities
include self-observation of goal attainment, intentional self-goal setting to increase effort,
the creation of self-reward contingencies, self-correction to avoid derailing behaviours,
and self-cueing to enhance personal agency over goal attainment (Neck & Houghton,
2006). Given the focus of self-leadership on behaviours that increase goal attainment, we
expect the behavioural strategies would primarily account for the association of self-
leadership strategies with performance outcomes.
Hypothesis 6. Behavioural strategies primarily account for the association of self-leadership
with performance.
Self-efficacy. Self-efficacy influences expectations of success and determines how much
effort will be expended in pursuit of a goal (Bandura, 1977). Social cognitive theory
postulates that individuals are not merely passive reactors but actively shape their
environment (Bandura, 1986). Studies by Norris (2008) and Sahin (2015) reported that
there is a positive relationship between self-leadership strategies and general self-efficacy.
Because self-leadership facilitates the adoption of productive cognition (creating new
knowledge and replacing destructive thinking patterns with constructive ones), we
expect a positive relationship with general self-efficacy. We therefore hypothesize:
Hypothesis 7. Self-leadership is positively correlated with self-efficacy.
Constructive thought strategies involve the self-influence of thoughts and align with
SCT in terms of self-reactive influences for self-assessment of capabilities to perform a task.
In SCT, there is a system of discrepancy production followed by discrepancy reduction
that involves using thought patterns to raise the level of motivation (Neck & Houghton,
2006). Bandura (1986) noted that cognitive simulations that involve individuals
visualizing themselves successfully executing activities raised confidence and subsequent
performance. Constructive-thought self-leadership involves mental imagery of perfor-
mance, constructive self-talk, and identification of alternative beliefs to currently held
dysfunctional beliefs; Stewart et al. (2011) note that this thought self-leadership can foster
self-efficacy. Therefore, we expect constructive thought strategies to predominantly
account for the relationship of self-leadership with self-efficacy.
Self-leadership meta-analysis 9
Hypothesis 8. Constructive thought strategies primarily account for the association of self-
leadership with self-efficacy.
Job attitudes. Job satisfaction refers to an employee’s attitude towards the job including
negative and positive evaluations of the work environment and the degree to which the
job is liked (Iverson & Mcguire, 2000). Houghton and Jinkerson (2007), Sesen, Tabak, and
Arli (2017), and Toglaw (2006) reported positive effects of self-leade rship strategies on job
satisfaction, whereas a few studies such as Konradt et al. (2009) and Roberts and Foti
(1998) reported no association. Organizational commitment describes a person having
strong beliefs in adopting organizational objectives and values, making large efforts for the
organization, and having the desire to maintain organizational membership (Mowday,
Steers, & Porter, 1979). It reflects the relationship between the organization and the
employee (Meyer & Allen, 1991). Chung, Chen, Lee, Chen, and Lin (2011) reported
positive associations between commitment and self-leadership strategies. Because job
satisfaction and organizational commitment reflect positive responses towards work, and
those who self-lead work take personal initiative towards engendering positive cognitive
and attitudinal outcomes (focusing on positive aspects of the job and work environment),
we expect job satisfaction and organizational commitment to be positively associated
with self-leadership.
Hypothesis 9. Self-leadership is positively correlated with (a) job satisfaction and (b)
organizational commitment.
Naturalreward strategies in self-leadershipinvolve usingforethought to focus onnatural
(or intrinsic) rewards that result from performing a task or engaging in an activity (Manz,
1986). Research on SCT suggests that intrinsic rewards or feelings of competence that are
experienced when facing a challenge act as positive feedback, driving positive feelings and
evaluations (Bandura, 2001). Focusing on feelings of purpose and valued aspects of work
drivefeelingsofjob satisfactionandcommitmenttotheorganization(Sesenetal., 2017).We,
therefore, expect natural reward strategies to primarily account for self-leadership’s
relationship with work attitudes (i.e., job satisfaction, organizational commitment).
Hypothesis 10. Natural reward strategies primarily account for the association of self-
leadership with (a) job satisfaction and (b) organizational commitment.
The primary mechanism driving performance: Self-efficacy. The focus of self-
leadership strategies is the enhancement of self-efficacy perceptions, especially with
respect to employing forethought, symbolizing, and vicarious learning in natural reward
and constructive thought strategies (Neck & Manz, 1992). As individuals exercise
personal agency their self-efficacy increases; research suggests this is the primary
mechanism through which self-leadership influences performance (Neck & Houghton,
2006). Self-leadership strategies build confidence in the ability to attain goals; this self-
efficacy, in turn, promotes greater effectiveness as individuals mobilize their efforts
towards performance consistent with their capability beliefs (Bandura & Cervone, 1986).
The model described here (i.e., self-efficacy as a mediator of the self-leadership-job
10 Michael B. Harari et al.
performance relationship) appears in the literature and was initially tested in Prussia et al.
(1998). In the present work, we provide a robust test of this model, as our analysis uses
correlations estimated in our meta-analysis versus those from a primary study. We,
therefore, hypothesize that self-efficacy will be an important mechanism through which
self-leadership strategies affect job performance:
Hypothesis 11. The relationship between self-leadership and job performance is mediated
by self-efficacy.
Cultural context dimension of power distance as a moderator
As noted by Alves et al. (2006), variation in culture might impact views and practice of self-
leadership. Although there are several dimensions of national culture that might explain
the way that various self-leadership strategies are employed, perhaps the dimension most
relevant to understanding the practice of self-leadership is power distance (Stewart et al.,
2011). Power distance refers to the degree of equality or inequality accepted between
people in a country’s society (Hofstede, 2001). There are some inconsistencies in the
power distance literature regarding self-leadership. While Alves et al. (2006) suggested
that individuals in high power distance cultures are likely to engage in a restricted form of
self-leadership (employing fewer strategies), Neubert and Wu (2006) demonstrated that
numerous self-leadership behaviours including goal setting, visualizing successful
performance, self-talk, self-reward, self-punishment, and natural rewards appear to
generalize to Chinese (high power distance) culture. To further explore this, we note that
the assumption that high power distance cultures (i.e., accepting of inequality in power)
are restrictive and limit one’s discretion, suggests that high power distance might weaken
the association between self-leadership and outcomes (Ho & Nesbit, 2014). On the other
hand, while self-leadership is encouraged in the West (Neck & Houghton, 2006) and
described as normal practice, this normalcy might result in less variance, limiting the
demonstration of strong effects. This suggests that while fewer strategies might be
employed in high power distance cultures, their effects might be stronger due to more
variation in practice given their novelty (Ho & Nesbit, 2014). The role of power distance,
in terms of the association between self-leadership and outcomes, will be interesting to
note (Ho & Nesbit, 2014). Because fewer strategies are utilized, will high power distance
weaken the association between self-leadership and outcomes due to restricted practice,
or will the strategies enacted have more impact because of their novelty –given it is
considered a Western normative model? Given that the current literature (albeit sparse)
suggests that power distance likely weakens the association between self-leadership and
outcomes, we pose the following research question:
Research Question 1: Will high power distance weaken the association between self-
leadership and outcomes?
Method
Literature Search
The systematic search protocol was carried out following current best practice
recommendations (see Harari et al., 2020) as follows. First, in 2017, we searched a range
Self-leadership meta-analysis 11
of databases (i.e., Social Science Premium Collection, ABI Inform, ProQuest Dissertations
and Theses, International Bibliography of the Social Sciences, Sociology Collection,
SciTech Premium Collection, Applied Social Sciences Index & Abstracts, Education
Collection, Education Database, Social Science Database, ERIC, Military Database,
PsycINFO, and PsycArticles) for articles that contained ‘self-leadership’. Our search
included all studies available until the date of the search being carried out with no date
constraints. Second, we forward searched the following self-leadership scale develop-
ment papers: Anderson and Prussia (1997), Cox (1993), Houghton and Neck (2002) , and
Houghton, Carnes, and Ellison (2014). Third, we searched 10 years of Society for
Industrial and Organizational Psychology and Academy of Management annual confer-
ence programs (i.e., 2009-2018). Finally, we sought additional unpublished studies by (1)
posting a call for studies on the AOM OB Listserv and (2) contacting 48 self-leadership
researchers to request unpublished data (the names and contact information for these
researchers were extracted from relevant studies identified in the first three steps of our
search).
A total of 3,235 articles were identified. First, we excluded studies that were not
written in English (k=72) and that were not primary quantitative studies (e.g., qualitative
studies, literature reviews, conceptual; k=1,036). Next, the study had to measure self-
leadership consistent with the operational definition offered earlier, based on the work of
Manz (1986, 1992) as a process through which people influence themselves to achieve
self-direction and self-motivation that they deem to be beneficial to their work-relevant
cognition, attitudes, and behaviours (Houghton & Neck, 2002). As our meta-analysis is
addressing overall and strategy-level relationships, studies were included if at least one
self-leadership strategy was measured. Failing to measure self-leadership was the primary
reason for the exclusion of initially screened studies (k=1,973); in many cases, the term
simply appeared in the study or references section, so it was captured in our search.
Studies were excluded that took place at the team-level of analysis (k=6) in examining
self-leadership. Finally, we excluded an additional 61 for (a) not measuring any needed
correlates or (b) not reporting zero-order correlations or statistics that could be used to
compute said correlations (where the needed data were not made available by the authors
upon request). Where studies reported correlations between multiple measures of the
same variables, we computed composites following the methods outlined in Schmidt and
Hunter (2015). Further, following the recommendations in Schmidt and Hunter (2015),
composite reliabilities were calculated using the Mosier (1943) formula. Ultimately, 86
studies (30% unpublished) were included.
As part of a revision effort, we updated our database by carrying out a backward search
and updated database search through 2020. For the backward search, we reviewed
references in self-leadership review articles (Neck & Houghton, 2006; Stewart et al., 2011,
2019), identifying k=305. Studies were excluded for not including primary quantitative
data (k=150), not including a measure of self-leadership (k=141), not including
correlates relevant to the present study (k=7), being carried out that the team-level
(k=2), correlations were not reported and could not be obtained by contacting the
authors (k=1), and for otherwise meeting our inclusion criteria but having been identified
in our initial search and already coded (k=4), resulting in k=2 relevant studies.
For the updated database search, we repeated the procedures outlined earlier for
studies available between 2018 and 2020 (i.e., the year in which the search was carried
out). We identified 379 studies. Among them, 23 did not report primary quantitative data,
one was written in a language that could not be interpreted by the authors, 300 did not
measure self-leadership, eight were non-employee samples. Among the studies dealing
12 Michael B. Harari et al.
with self-leadership, 26 did not measure needed correlates, five did not report needed data
(and the data were not made available upon request), and three were captured in our prior
literature search, resulting in k=13 studies, and therefore resulting in k=101 studies
(111 independent samples; 26% unpublished) in total coded for our meta-analysis.
Included studies are reported in Supporting Information.
Coding
All studies were coded by the lead author and split between the second through fourth
authors. Thus, all studies were coded twice and independently. Inter-rater agreement was
97% across all coded characteristics and disagreements (which tended to be minor
mistakes such as typographical errors or misreading of correlation tables) were resolved.
Authors coded correlations, reliabilities (i.e., coefficient alpha), and study characteristics
(i.e., self-leadership scale, study design, power distance). The measures included in the
study were all individual level. Job performance measures included self- and other-ratings
of job performance and measures of overall or dimension-level (e.g., task, OCB)
performance; composite correlations and reliabilities were calculated where multiple
dimensions were reported. Similarly, organizational commitment measures included
those that reflected dimension-level (e.g., affective, continuance) and overall. Where
dimensions were provided without an overall score, composite correlations, and
reliabilities were calculated. Where only a single dimension was reported (e.g., affective
commitment), as an indicator of the latent organizational commitment variable, it was
coded and included in analyses.
Analyses
Analyses were carried out in Excel following the psychometric meta-analysis formulas
outlined in Hunter and Schmidt (2004). For each relationship, we computed sample size-
weighted mean correlations. To correct this estimate for the biasing effect of measure-
ment error, using artifact distributions comprised of coefficient alphas from primary
studies, we calculated corrected sample size-weighted mean correlations (i.e., q). To
evaluate heterogeneity in effects, we calculated %var (i.e., the percentage of variance in
observed correlations accounted for by sampling error). A value greater than 75% would
support homogeneity –that is, that the relationship is not affected by moderators
(Schmidt & Hunter, 2015). We also calculated 80% credibility intervals –the range in
which the distribution of underlying true score correlations would likely fall (Schmidt,
Viswesvaran, Ones, & Le, 2017). Finally, we computed 95% confidence intervals around
our estimate of q(Viswesvaran, Schmidt, & Ones, 2002).
We evaluated the moderating effect of power distance using categorical moderator
analysis procedures, which involves grouping studies together based on their level of the
moderator and repeating meta-analytic calculations within each level of the moderator.
Results can be compared by evaluating differences in the point estimate at each level of
the moderator and overlap in their associated 95% confidence intervals. If the 95%
confidence intervals around qestimated at each level of the moderator do not overlap, this
would suggest that power distance acted as a moderator (Schmidt & Hunter, 2015). Using
scores from Hofstede et al. (2005), we classified studies included in our meta-analysis as
either high (i.e., power distance scores greater than 50) or low (i.e., power distance scores
less than 50) power distance (see Courtright, Thurgood, Stewart, & Pierotti, 2015; Harari,
Manapragada, & Viswesvaran, 2017), and carried out categorical moderator analyses
Self-leadership meta-analysis 13
where k=3 studies were available. Although this involves using a continuous variable
(i.e., power distance scores) to form categories, resulting in some loss of information (i.e.,
introduction of coarseness), such an approach is widely used and, on balance, is
recommended as a superior alternative to meta-regression (see Schmidt, 2017, for a
detailed discussion).
Regression and relative weight analysis
To test hypotheses concerning (1) the effect of multiple antecedents on self-leadership
(i.e., hypotheses 1–4), (2) the effect of multiple self-leadership dimensions on outcomes
(hypotheses 6, 8, and 10), and (3) mediation (Hypothesis 11), we carried out meta-analytic
regression and relative weight analyses (RWA). This involved building a correlation matrix
that included all variables used in the regression models (Viswesvaran & Ones, 1995). The
sample size used in analyses was the harmonic mean of pooled Ns across cells. The
correlations between the FFM traits were derived from Ones, Viswesvaran, and Reiss
(1996). The FFM trait-self-efficacy correlations were drawn from Judge, Erez, Bono, and
Thoresen (2002)
1
. When predictors in a regression model are intercorrelated, the beta-
weights do not accurately indicate the contribution made by each predictor towards
explained variance in the criterion (Tonidandel & LeBreton, 2011). Thus, to test
hypotheses concerning relative importance of antecedents, we carried out RWA as a
supplement to regression analyses using RWA-web (Tonidandel & LeBreton, 2015).
Mediation analyses were carried out using path analysis following the two-step
procedure recommended in Hayes and Rockwood (2017). Model 1 estimated the apath
(i.e., self-leadership to self-efficacy) by regressing self-efficacy onto self-leadership. Model
2 estimated the b(i.e., self-efficacy to job performance, controlling for self-leadership) and
c(i.e., self-leadership to job performance, controlling for self-efficacy) paths by regressing
job performance onto self-efficacy and self-leadership. Although bootstrapping 95%
confidence intervals as a method for evaluating the statistical significance is recom-
mended, this is not possible when using a me ta-analytically derived correlation matrix. In a
simulation study, MacKinnon, Lockwood, Hoffman, West, and Sheets (2002) observed
that a joint significance test was the most effective of available alternatives in terms of its
balance of type 1 and type 2 error rates, and this approach is applicable and utilized in
meta-analysis (e.g., Harari, Reaves, Beane, Laginess, & Viswesvaran, 2018). Following this
procedure, when both the aand bpaths are statistically significant, the indirect effect is
statistically significant as well.
Sensitivity analyses
Publication bias
To evaluate robustness of our findings, we carried out various sensitivity analyses. In terms
of publication bias, following the recommendations of Schmidt and Hunter (2015), we
utilized Cumulative Meta-Analysis (CMA) procedures. Large sample size studies are least
likely to be affected by suppression of null effects, as large sample sizes increase the
probability of a statistically significant finding, even if the effect is small. Suppression of
null effects is more likely in small sample size studies. CMA involves carrying out the meta-
analysis iteratively, adding studies to the analysis in order of the sample sizes (i.e., largest to
1
We corrected observed mean correlations for measurement error using artifact distributions from the present study.
14 Michael B. Harari et al.
smallest). Taking the fifth iteration of CMA (i.e., results including the five studies with the
largest sample sizes) allows us to produce an alternative effect size estimate –q
prec
–that
we can compare to our na
€
ıve estimate to evaluate the presence of publication bias (e.g.,
Harrison, Banks, Pollack, O’Boyle, & Short, 2017; Kepes, Banks, McDaniel, & Whetzel,
2012; Kepes & McDaniel, 2015). We carried out these analyses when at least k=10
studies were available.
Study quality
We address two characteristics that speak to study quality. First, several scales exist to
measure self-leadership consistent with our operational definition. Studies have initially
relied on ad hoc measures. In a dissertation, Cox (1993) developed a self-leadership scale,
constructing items following the conceptualization offered by Manz and colleagues (e.g.,
Manz, 1986; Manz & Sims, 1987). However, Cox’s scale, perhaps owing to it being
unpublished, has not been widely used. Subsequently, the Self-Leadership Questionnaire
was developed in Anderson and Prussia (1997). However, it was criticized for poor
psychometric properties. For example, several subscales produced low alphas and in
some instances items loaded weakly onto intended factors. These issues prompted the
development of the Revised Self-Leadership Questionnaire (RSLQ; Houghton & Neck,
2002), which is currently the most construct valid and reliable measure available.
Although the RSLQ demonstrates superior psychometric properties, the RSLQ and SLQ
largely conceptualize self-leadership strategies similarly. Both measure natural rewards
using a single subscale and constructive thoughts via visualizing successful performance,
self-talk, and evaluating beliefs and assumption subscales. The dimensionality of
behaviour-focused strategies exhibits minor differences. Both scales utilize self-goal
setting, self-reward, self-punishment, self-observation, and self-cueing, whereas the SLQ
also includes a self-withholding subscale. Because the RSLQ is a superior measure, we
carry out a set of sensitivity analyses whereby we report findings from only those studies
that used the RSLQ
2
.
Second, we address study design as an indicator of study quality. We carry out
categorical moderator analysis procedures with two levels: cross-sectional/self-report
(CS/SR) versus multi-wave/multi-source (MW/MS). Although valid conclusions can be
drawn from CS/SR studies (e.g., Brannick, Chan, Conway, Lance, & Spector, 2010;
Spector, 2006), when such studies are published in leading I-O Psychology and
Management journals, the design is frequently noted as a limitation (Brutus, Aguinis, &
Wassmer, 2013). Further, a MW/MS study may signify a greater concern for methodolog-
ical rigour on the part of investigators, as well as greater attention to detail. Thus, it is not
only a belief that MW/MS data collection methods are likely superior to CS/SR methods
that support the use of this moderator but also that it can be considered as serving as a
general indicator of study quality that might correlate with other, less discernable or
unreported indicators of study quality (Aguinis, Ramani, & Alabduljader, 2018).
2
Note that nearly all studies used the RSLQ (or an adapted version, e.g., translation; 84%). Thus, it is not possible to use categorical
moderator analysis to compare findings from studies that used the RSLQ vs. other measures. Unless noted otherwise, where an
RSLQ subgroup analysis is not presented, this is because all studies included in the main analysis used the RSLQ.
Self-leadership meta-analysis 15
Results
Hypothesis 1, dealing with the role of the FFM traits as antecedents of self-leadership,
suggested that the combined effect of said traits is predominantly driven by conscien-
tiousness. Results of our meta-analysis of personality-self-leadership correlations are
reported in Table 1 and results of regression and relative weight analyses are reported in
Table 2. Per Table 2, the FFM traits accounted for 34% of the variance in overall self-
leadership scores and, as predicted, this was predominantly driven by conscientiousness
(R-RW =38.80%). We also observed large contributions from openness (R-RW =32.82%)
and extraversion (R-RW =22.62%).
As predicted by Hypothesis 2, the overall effect of the FFM traits on behaviour-focused
strategies was predominantly driven by conscientiousness (R-RW =53.02%; Table 2).
Hypothesis 3, which predicted that the effect of the FFM traits on constructive thought
strategies was predominantly driven by openness, was supported (R-RW =51.43%;
Table 2). Finally, hypothesis 4 –that the effect of the FFM traits on natural rewards
strategies is predominantly driven by extraversion –was not supported. Although
extraversion made a strong contribution to explained variance (R-RW =31.77%),
openness to experience was the strongest driver (R-RW =36.62%; Table 2).
Now turning towards the outcomes of self-leadership, hypothesis 5 predicted a
positive relationship between self-leadership and job performance. The relationship was
positive across rating sources (q=.44 for overall self-leadership; see Table 3) and, as
shown in Table 4, was positive whether work performance was evaluated using self
(q=.53) or other (q=.12) ratings; though, the effect was much smaller when other
performance ratings were used.
Hypothesis 6 predicted that the self-leadership dimension-level relationships with job
performance would be primarily accounted for by behaviour-focused strategies. The
correlations between self-leadership strategies used in our regression and relative weight
analyses (which ranged from q=.60 to .68) appear in Supporting Information (Table S1).
As reported in Table 5, hypothesis 6 was not supported. In fact, natural rewards strategies
(followed by behaviour-focused strategies) accounted for the largest proportion of
predicted variance in job performance regardless of the rating source. Behaviour-focused
strategies uniformly accounted for the second largest proportion of predicted variance.
Hypothesis 7 predicted a positive relationship between self-leadership and self-
efficacy, which was supported (q=.43 for overall self-leadership; see Table 6).
Hypothesis 8 predicted that, among the self-leadership dimensions, explained variance
in self-efficacy was principally accounted for by constructive thought strategies. As
reported in Table 7, findings supported this hypothe sis (R-RW =34.39%). However, both
behaviour-focused (R-RW =32.79%) and natural rewards (R-RW =32.82%) strategies
accounted for similarly large proportions of predicted variance. Thus, although our
hypothesis was technically supported, it appears that all three dimensions are important
drivers of self-efficacy.
Turning to work attitudes (i.e., job satisfaction, organizational commitment),
hypothesis 9 predicted a positive relationship, which was supported by our findings
(q=.29 and .37 for overall self-leadership with job satisfaction and organizational
commitment, respectively; see Table 8). As specified in hypothesis 10, we anticipated
that the effect of self-leadership dimensions on job attitudes would be primarily accounted
for by natural reward strategies. As reported in Table 7, we observed mixed support. For
organizational commitment, natural rewards strategies did account for the largest
proportion of predicted variance (R-RW =41.73%). For job satisfaction, although natural
16 Michael B. Harari et al.
Table 1. Meta-analysis of self-leadership-personality correlations
kN r qr
q
%Var CV
L
CV
U
CI
L
CI
U
q
prec
Conscientiousness
(with overall)
16 4,294 .29 .35 .12 24.14 0.20 0.50 0.28 0.42 .35
RSLQ 14 3,909 .31 .38 .06 51.70 0.30 0.45 0.33 0.42 .35
CS/SS 9 2,792 .33 .40 .06 48.24 0.33 0.48 0.34 0.46
MW/MS 7 1,502 .20 .25 .14 25.10 0.07 0.42 0.13 0.37
Behaviour 12 3,580 .36 .45 .14 16.51 0.27 0.62 0.36 0.53 .38
Constructive 10 3,185 .20 .26 .06 54.94 0.18 0.33 0.20 0.31 .25
Natural 11 3,294 .22 .30 .18 14.09 0.06 0.54 0.18 0.42 .31
Extraversion
(with overall)
9 2,273 .25 .30 .22 9.22 0.01 0.58 0.14 0.45
RSLQ 6 1,715 .25 .31 .22 8.21 0.02 0.59 0.12 0.49
CS/SS 6 1,715 .25 .31 .22 8.21 0.02 0.59 0.12 0.49
MW/MS 3 558 .23 .27 .22 12.46 0.01 0.55 0.00 0.54
Behaviour 6 1,715 .24 .30 .18 12.27 0.07 0.53 0.14 0.45
Constructive 6 1,715 .18 .22 .12 25.97 0.07 0.37 0.11 0.33
Natural 7 2,066 .28 .37 .31 5.30 0.02 0.76 0.14 0.61
Openness
(with overall)
7 1,782 .27 .34 .12 26.71 0.19 0.49 0.23 0.44
RSLQ 5 1,397 .31 .39 .00 100.00 0.39 0.39 0.35 0.42
CS/SS 4 1,289 .32 .39 .00 100.00 0.39 0.39 0.36 0.42
MW/MS 3 493 .16 .19 .17 23.19 0.03 0.41 0.03 0.41
Behaviour 4 1,289 .24 .30 .00 100.00 0.30 0.30 0.27 0.33
Constructive 4 1,289 .30 .38 .00 100.00 0.38 0.38 0.33 0.43
Natural 4 1,289 .27 .38 .00 72.92 0.38 0.38 0.29 0.46
Emotional Stability
(with overall)
9 2,506 .04 .04 .16 17.35 0.16 0.24 0.07 0.16
RSLQ 8 2,216 .01 .01 .13 22.55 0.16 0.18 0.10 0.11
CS/SS 6 1,715 .02 .02 .15 18.15 0.17 0.21 0.11 0.16
MW/MS 3 791 .08 .09 .16 17.42 0.11 0.29 0.11 0.29
Behaviour 8 2,394 .01 .02 .16 15.39 0.19 0.23 0.11 0.14
Constructive 7 2,108 .01 .01 .11 28.47 0.16 0.13 0.11 0.08
Natural 7 2,108 .10 .14 .20 12.84 0.11 0.40 0.02 0.30
Agreeableness
(with overall)
5 1,579 .13 .17 .04 70.43 0.11 0.22 0.09 0.24
RSLQ 4 1,289 .14 .17 .06 56.88 0.10 0.24 0.08 0.26
Behaviour 4 1,289 .12 .15 .06 58.09 0.08 0.22 0.06 0.24
Constructive 4 1,289 .08 .10 .05 65.12 0.04 0.16 0.02 0.18
Natural 4 1,289 .15 .21 .05 69.19 0.16 0.27 0.12 0.30
k=number of independent samples included i analysis, N=pooled sample size, r=observed sample
size-weighted correlation, q=sample size-weighted corrected correlation, r
q
=sample size-weighted
standard deviation of corrected correlations, %Var =percent variance accounted for in correlations by
sampling error, CV =80% credibility intervals, CI =95% confidence intervals, q
prec
=corrected
correlation calculated using the five effects with the greatest precision, Overall =Overall Self-
Leadership, Behaviour =Behaviour-Focused, Constructive =Constructive Thought, Natural =Natu-
ral Rewards, RSLQ =Revised Self-Leadership Questionnaire, CS/SS =Cross-Sectional/Single-Source,
MW/MS =Multi-Wave/Multi-Source.
Self-leadership meta-analysis 17
Table 2. Regression and relative weight analyses of personality traits on self-leadership (H1–H4)
Overall (H1) Behaviour-focused (H2) Constructive thoughts (H3) Natural rewards (H4)
BRW R-RW (%) BRW R-RW (%) BRW R-RW (%) BRW R-RW (%)
Conscientiousness .41*.13 38.80 .53*.22 53.02 .34*.08 28.72 .33*.09 25.38
Extraversion .27*.08 22.62 .29*.08 19.64 .19*.04 14.21 .32*.11 31.77
Openness .35*.11 32.82 .32*.09 21.87 .40*.15 51.43 .35*.13 36.62
Emotional stability .18*.01 2.42 .22*.01 3.43 .19*.01 4.35 .07*.00 1.21
Agreeableness .02 .01 3.34 .02*.01 2.04 .02 .00 1.29 .05*.01 5.02
R(R
2
) .58 (.34)*.64 (.42)*.53 (.28)*.59 (.35)*
RW =relative weight, R-RW =rescaled-relative weight.
*Statistically significant at p<.05.
18 Michael B. Harari et al.
Table 3. Meta-analysis of self-leadership-job performance correlations (any rating source: H5)
kN r qr
q
%Var CV
L
CV
U
CI
L
CI
U
q
prec
Overall 38 10,031 .38 .44 .26 5.13 0.11 0.77 0.36 0.53 .50
RSLQ 31 9,065 .38 .44 .25 5.08 0.13 0.76 0.35 0.53 .50
CS/SS 23 7,403 .46 .54 .20 5.68 0.27 0.80 0.45 0.62 .57
MW/MS 15 2,628 .15 .18 .22 13.53 0.10 0.45 0.06 0.30 .12
Behaviour 15 3,618 .29 .34 .23 8.51 0.05 0.63 0.22 0.46 .34
RSLQ 12 3,305 .32 .38 .20 8.65 0.12 0.64 0.26 0.50 .34
Constructive 15 3,797 .24 .32 .24 8.81 0.00 0.63 0.19 0.45 .35
RSLQ 12 3,352 .28 .36 .23 8.83 0.07 0.65 0.22 0.50 .35
Natural 14 3,756 .26 .34 .21 10.45 0.06 0.61 0.22 0.46 .30
RSLQ 13 3,591 .27 .35 .21 10.20 0.08 0.62 0.22 0.47 .30
k=number of independent samples included in analysis, N=pooled sample size, r=observed sample
size-weighted correlation, q=sample size-weighted corrected correlation, r
q
=sample size-weighted
standard deviation of corrected correlations, %Var =percent variance accounted for in correlations by
sampling error, CV =80% credibility intervals, CI =95% confidence intervals, q
prec
=corrected
correlation calculated using the five effects with the greatest precision, Overall =Overall Self-
Leadership, Behaviour =Behaviour-Focused, Constructive =Constructive Thought, Natural =Natu-
ral Rewards, RSLQ =Revised Self-Leadership Questionnaire, CS/SS =Cross-Sectional/Single-Source,
MW/MS =Multi-Wave/Multi-Source.
Table 4. Meta-analysis of self-leadership-job performance correlations (by rating source: H5)
kN r qr
q
%Var CV
L
CV
U
CI
L
CI
U
q
prec
Self-rated 29 8,095 .45 .53 .21 6.47 0.26 0.79 0.45 0.61 .57
RSLQ 27 7,587 .44 .51 .20 6.80 0.25 0.77 0.43 0.59 .57
CS/SS 23 7,403 .46 .54 .20 5.68 0.28 0.80 0.45 0.63 .57
MW/MS 6 692 .35 .41 .20 18.81 0.15 0.66 0.23 0.58
Behaviour 6 1,479 .42 .50 0.08 32.69 0.39 0.60 0.41 0.58
Constructive 6 1,526 .36 .43 .14 17.36 0.25 0.60 0.30 0.55
Natural 6 1,188 .40 .52 .04 56.32 0.47 0.57 0.44 0.61
Other-Rated 11 2,211 .10 .12 .17 17.30 0.11 0.34 0.002 0.23 .09
RSLQ 6 1,753 .11 .13 .18 11.65 0.11 0.36 0.03 0.28
Behaviour 6 1,249 .09 .10 .19 14.70 0.14 0.35 0.07 0.27
RSLQ 3 936 .12 .14 .21 8.92 0.12 0.41 0.10 0.39
Constructive 6 1,381 .08 .09 .19 14.57 0.15 0.33 0.07 0.25
RSLQ 3 936 .11 .13 .21 9.03 0.13 0.40 0.11 0.38
Natural 5 1,678 .11 .14 .14 19.01 0.04 0.32 0.00 0.28
RSLQ 4 1,513 .12 .15 .15 15.46 0.04 0.34 0.01 0.31
k=number of independent samples included in analysis, N=pooled sample size, r=observed sample
size-weighted correlation, q=sample size-weighted corrected correlation, r
q
=sample size-weighted
standard deviation of corrected correlations, %Var =percent variance accounted for in correlations by
sampling error, CV =80% credibility intervals, CI =95% confidence intervals, q
prec
=corrected
correlation calculated using the five effects with the greatest precision, Overall =Overall Self-
Leadership, Behaviour =Behaviour-Focused, Constructive =Constructive Thought, Natural =Natu-
ral Rewards, RSLQ =Revised Self-Leadership Questionnaire, CS/SS =Cross-Sectional/Single-Source,
MW/MS =Multi-Wave/Multi-Source.
Self-leadership meta-analysis 19
rewards strategies accounted for a large proportion of predicted variance (R-
RW =36.87%), a similar, though slightly larger, value was estimated for behaviour-
focused strategies (R-RW =38.24%).
Table 5. Regression and relative weight analyses of self-leadership strategies on job performance (H6)
Any source Self-rated Other-rated
BRW R-RW (%) BRW R-RW (%) BRW R-RW (%)
Behaviour-focused .15*.05 34.18 .25*.11 35.28 .02*.00 20.00
Constructive thoughts .11*.04 28.68 .07*.07 21.36 .00*.00 14.77
Natural rewards .18*.05 37.15 .32*.14 43.36 .13*.01 65.24
R(R
2
) .38 (.15)*.57 (.32)*.14 (.02)*
RW =relative weight, R-RW =rescaled-relative weight.
*Statistically significant at p<.05.
Table 6. Meta-analysis of self-leadership-self-efficacy correlations (H7)
kN r qr
q
%Var CV
L
CV
U
CI
L
CI
U
q
prec
Overall 19 5,202 .36 .43 .16 13.12 0.23 0.64 0.36 0.51 .43
CS/SS 11 3,205 .42 .50 .16 11.25 0.30 0.70 0.40 0.60 .50
MW/MS 8 1,997 .27 .33 .08 42.92 0.23 0.42 0.25 0.40
Behaviour 11 3,791 .33 .40 .19 8.81 0.16 0.63 0.28 0.51 .38
Constructive 11 3,791 .33 .40 .16 11.44 0.19 0.60 0.29 0.50 .38
Natural 10 3,635 .30 .39 .12 20.06 0.24 0.55 0.31 0.48 .34
k=number of independent samples included in analysis, N=pooled sample size, r=observed sample
size-weighted correlation, q=sample size-weighted corrected correlation, r
q
=sample size-weighted
standard deviation of corrected correlations, %Var =percent variance accounted for in correlations by
sampling error, CV =80% credibility intervals, CI =95% confidence intervals, q
prec
=corrected
correlation calculated using the five effects with the greatest precision, Overall =Overall Self-
Leadership, Behaviour =Behaviour-Focused, Constructive =Constructive Thought, Natural =Natu-
ral Rewards, RSLQ =Revised Self-Leadership Questionnaire, CS/SS =Cross-Sectional/Single-Source,
MW/MS =Multi-Wave/Multi-Source.
Table 7. Regression and relative weight analyses of self-leadership strategies on self-efficacy and job
attitudes (H8, H10)
Self-efficacy Job satisfaction
Organizational
commitment
BRW R-RW (%) BRW R-RW (%) BRW R-RW (%)
Behaviour-focused .16*.07 32.79 .18*.05 38.24 .16*.05 35.24
Constructive thoughts .18*.07 34.39 .07*.03 24.89 .07*.03 23.24
Natural rewards .18*.07 32.82 .16*.05 36.87 .21*.06 41.73
R(R
2
) .46 (.21)*.36 (.13)*.38 (.15)*
RW =relative weight, R-RW =rescaled-relative weight
*Statistically significant at p<.05.
20 Michael B. Harari et al.
Hypothesis 11 predicted that the relationship between self-leadership and job
performance is mediated through self-efficacy. We observed support for this prediction.
Both models were statistically significant (R
2
=.05 and .19 for model 1 and 2,
respectively), and results indicated that the indirect effect of self-leadership on job
performance through self-efficacy (ab =.10) was statistically significant following the
joint significance test, outlined earlier (MacKinnon et al., 2002). Detailed results are
reported in Supporting Information (Table S2).
Finally, research question 1 dealt with the effect of cultural differences in power
distance on self-leadership’s relations with the dependent variables addressed in the
present study. Findings are reported in Table 9. We observed some noteworthy effects.
First, population self-leadership-job performance correlations differed across levels of the
power distance moderator. Namely, the effect for overall self-leadership with perfor-
mance across rating sources was stronger in high (q=.58) versus low (q=.32) power
distance cultures.
In terms of self-efficacy and job attitudes, although a consistent pattern emerged, 95%
confidence intervals around effects estimated at different levels of the power distance
moderator overlapped. Nonetheless, the point estimates for overall self-leadership with
self-efficacy (q=.37 vs. .46), job satisfaction (q=.23 vs. .40), and organizational
commitment (q=.26 vs. .47) for low versus high power distance cultures, respectively,
were consistent with the pattern of results observed for job performance. Thus, although
conventional levels of statistical significance were not reached , we interpret results across
analyses as pointing to a tendency for effects to be stronger in high versus low power
distance cultures.
Table 8. Meta-analysis of self-leadership-job attitude correlations (H9)
kN r qr
q
%Var CV
L
CV
U
CI
L
CI
U
q
prec
Job satisfaction 17 3,457 .25 .29 .14 22.11 0.10 0.47 0.21 0.37 .34
RSLQ 12 2,756 .26 .30 .15 17.89 0.10 0.50 0.20 0.40 .32
CS/SS 10 2,171 .31 .36 .11 28.59 0.22 0.50 0.28 0.44 .41
MW/MS 7 1,286 .14 .16 .10 42.87 0.04 0.29 0.07 0.26
Behaviour 7 1,729 .28 .33 .13 20.91 0.15 0.50 0.21 0.44
Constructive 8 1,992 .24 .29 .12 24.72 0.14 0.45 0.19 0.39
Natural 4 1,303 .25 .32 .00 98.18 0.32 0.32 0.25 0.39
Commitment 8 2,890 .32 .37 0.13 13.53 0.19 0.54 0.27 0.47
Behaviour 6 2,580 .29 .34 .14 12.29 0.16 0.51 0.22 0.45
Constructive 6 2,580 .26 .30 .11 17.23 0.16 0.45 0.20 0.41
Natural 5 2,489 .28 .35 .12 14.60 0.20 0.51 0.23 0.48
k=number of independent samples included in analysis, N=pooled sample size, r=observed sample
size-weighted correlation, q=sample size-weighted corrected correlation, r
q
=sample size-weighted
standard deviation of corrected correlations, %Var =percent variance accounted for in correlations by
sampling error, CV =80% credibility intervals, CI =95% confidence intervals, q
prec
=corrected
correlation calculated using the five effects with the greatest precision, Overall =Overall Self-
Leadership, Behaviour =Behaviour-Focused, Constructive =Constructive Thought, Natural =Natu-
ral Rewards, RSLQ =Revised Self-Leadership Questionnaire, CS/SS =Cross-Sectional/Single-Source,
MW/MS =Multi-Wave/Multi-Source.
Self-leadership meta-analysis 21
Sensitivity analyses
Publication bias
We observed evidence of modest publication bias for only one distributio n of correlations:
those involving other ratings of job performance. Specifically, q
prec
(.09) was 22% smaller
than our na€
ıve estimate (.12; see Table 4). Note that the conclusion, that self-leadership
shares a small positive relationship with job performance, remains unchanged.
In light of this observation, we suggest that research is needed to take greater care to
distinguish between performance dimensions, understand which strategies are likely to
relate to each dimension, and which dimensions of performance are unlikely to be
affected by self-leadership strategies. The ability for future meta-analyses to carry out
publication bias analyses for this relationship with larger kvalues and more homogenous
subsamples will shed further light on this issue.
Study quality
In terms of self-leadership measures, excluding non-RSLQ measures did not alter any
substantive conclusions. In terms of the study design moderator (i.e., CS/SR vs. MW/MS),
we observed a consistent effect. Findings from lower quality studies (i.e., CS/SR) were
larger than findings from higher quality studies (MW/MS). We agree with Spector and
colleagues (Brannick et al., 2010; Spector, 2006; Spector, Rosen, Richardson, Williams, &
Johnson, 2017) that CS/SR studies are not necessarily less valid than those that are MW/
MS, but the latter produces estimates that are more conservative.
Table 9. Results of power distance moderator analyses (RQ1)
kN r qr
q
%Var CV
L
CV
U
CI
L
CI
U
q
prec
Job performance (any source)
Low PD 18 3,118 .28 .32 .20 14.15 0.07 0.58 0.22 0.42 .31
High PD 16 5,406 .50 .58 .20 5.07 0.32 0.84 0.48 0.68 .62
Job performance (self-rated)
Low PD 13 2,660 .32 .37 .18 14.01 0.15 0.60 0.27 0.48 .32
High PD 14 4,919 .52 .61 .18 5.45 0.37 0.84 0.51 0.71 .68
Self-efficacy
Low PD 9 1,779 .31 .37 .12 28.19 0.22 0.52 0.28 0.46
High PD 5 1,858 .39 0.46 .14 12.58 0.29 0.64 0.33 0.59
Job satisfaction
Low PD 12 1,671 .20 .23 .12 38.75 0.08 0.38 0.14 0.32 .26
High PD 4 1,476 .34 .40 .08 29.05 0.30 0.50 0.30 0.50
Commitment
Low PD 4 637 .23 .26 .12 34.31 0.11 0.41 0.12 0.40
High PD 3 1,572 .41 .47 .08 18.84 0.36 0.57 0.36 0.57
k=number of independent samples included in analysis, N=pooled sample size, r=observed sample
size-weighted correlation, q=sample size-weighted corrected correlation, r
q
=sample size-weighted
standard deviation of corrected correlations, %Var =percent variance accounted for in correlations by
sampling error, CV =80% credibility intervals, CI =95% confidence intervals, q
prec
=corrected
correlation calculated using the five effects with the greatest precision, PD =Power Distance.
22 Michael B. Harari et al.
Discussion
Research on self-leadership continues to emerge, providing new and novel insights
concerning the role of self-influence in organizations (Brown & Fields, 2011; Steinbauer,
Renn, Taylor, & Njoroge, 2014). Our research complements narrative reviews on the
topic, employing SCT as an overarching framework, summarizing the state of the
literature quantitatively, and conceptualizing self-leadership at both the global and
strategy levels (Stewart et al., 2011, 2019). By pooling findings across studies, accounting
for the biasing effects of statistical artifacts on observed correlations, our meta-analysis
helps to draw generalizable conclusions from the existing body of work, as well as bring
forth new insights.
Perhaps our most striking findings were those related to job performance. Following
Cohen’s (1992) rules-of-thumb for interpreting effect size magnitude, the self-leadership-
job performance relationship was small in magnitude when job performance was
measured using other-ratings and large in magnitude when measured using self-ratings.
Thus, a broad overview of the literature that does not delineate between these rating
sources can lead to inaccurate conclusions concerning how much job performance might
improve by engaging in self-leadership strategies. Other- versus self-ratings of job
performance measure vastly different constructs. For example, ratings from peers and
supervisors (i.e., other-raters) converge with one another, whereas self-ratings appear to
measure a distinct construct (Scullen, Mount, & Goff, 2000; Viswesvaran et al., 2002).
Although errors and biases affect other-ratings of performance, the construct validity has
been well-researched and is favourable (see Harari & Viswesvaran, 2018, for a detailed
review). As an example, supervisor ratings of sales performance correlate with objective
sales performance records at .71 (Bommer, Johnson, Rich, Podsakoff, & Mackenzie,
1995). Conversely, the construct validity of self-ratings is poorly understood and, as
observed here, can misstate the effect (in this case, by a large margin).
Theoretical implications
Uncertainty remained in the self-leadership literature related to personality (Guzzo, 1998;
Markham & Markham, 1998; Stewart et al., 2011). Shedding light on this issue, our
analyses indicated that three personality traits in particular facilitated engagement in self-
leadership: conscientiousness, openness, and extraversion. Our strategy-level regression
and relative weight analyses elucidate why these three particular traits are so important for
self-leadership. These traits promote the behaviours consistent with the five basic
capabilities described in SCT (Bandura, 1986, 2001). Conscientiousness facilitates
engagement in behaviour-focused strategies as impulse control employs self-regulative
and self-reflective capabilities to persist towards goals. Openness promotes constructive
thought strategies as individuals employ symbolizing (visualizing) and learning capabil-
ities for managing their thoughts. Reflecting on the results for hypothesis 4, contrary to
expectations we found openness to be a stronger driver of natural reward strategies than
extraversion (with both playing important roles explaining variance (R-RW of 36.62% and
31.77%, respectively). Utilizing SCT, we posit that this is because openness helps
individuals use forethought capabilities to imagine and plan for pleasurable aspects of
their work while extraversion allows individuals to employ forethought in seeking out and
identifying pleasurable aspects of their work. Thus, as a set, these three traits facilitate the
enactment of the full spectrum of self-leadership strategies.
Self-leadership meta-analysis 23
Our analyses at the global self-leadership level were consistent with our expectations
of positive effects on outcomes of job performance, self-efficacy, and job attitudes
(satisfaction and commitment). At the strategy levels, consistent with SCT, we found that
each strategy largely accounted for different outcomes, however, findings were not
uniformly in line with our expectations.
In reflecting on hypothesis 6, although we anticipated that behaviour-focused
strategies would primarily drive self-leadership’s effect on job performance, natural
rewards did. Reflecting on SCT, forethought capability might play a larger role than
assumed in executing on goals in terms of planning than even self-regulation and self-
reflection. This suggests that forethought is important in planning for performance
followed by self-regulation and self-reflective capabilities as individuals determine needed
changes, set specific goals, and motivate themselves to future performance with self-
rewards. It appears that one confound could be variability in job performance dimensions
studied; despite empirical overlap across dimensions, they are not redundant (Harari et al.,
2016). It is possible that performance dimensions dealing with performance on in-role job
tasks (i.e., task performance) are driven by behaviour-focused strategies (following the
logic outlined earlier in the manuscript), whereas dimensions that are driven more so by
intrinsic motivation of job tasks (i.e., creative and innovative performance/CIP; Amabile,
1988) are predominantly driven by natural rewards strategies. Ancillary analyses reported
in Supporting Information (Tables S3 and S4) provide support for differences across
performance dimensions. Behaviour-focused strategies made the largest contribution to
predicted variance in task performance (R-RW =67.71%), whereas natural rewards
strategies did so for CIP (R-RW =40.22%). Further performance dimension-level research
appears to be warranted based on these findings.
Our results confirm the importance of cognitive visualization and vicarious learning
capabilities for raising confidence (self-efficacy) in performance abilities as well as the
importance of forethought capability in planning and anticipating results for creating
positive feelings of organizational commitment. Our results suggest that, while natural
rewards strategies account for a large proportion of variance in job satisfaction, the larger
proportion was accounted for by behaviour-focused strategies. In reflecting on these
results related to hypothesis 10, it appears that, while forethought helps with being
satisfied as plans are realized, self-regulation and self-reflective capabilities might provide
more tangible evidence that behavioural control and realization of goals (when self-
rewards are applied) are bearing fruit, creating feelings of satisfaction.
Support for the mediating effect of self-efficacy in linking self-leadership and job
performance confirms that performance outcomes are the result of a process involving
thoughts, feelings, and behaviours that increase expectations of success in the pursuit of a
goal. This also confirms that research on self-leadership that extends current models and
examines more comprehensive process models is needed (Stewart et al., 2011, 2019). Self-
leadership behaviours might be converted into action through the confidence that
accompanies feelings of self-efficacy. This suggests a causal relationship that can be
extended in future research. Having established self-efficacy as a mediator in the self-
leadership-job performance relationship suggests that self-efficacy might also be an
important mediator for outcomes not considered in this research.
Our findings with respect to power distance merit further consideration. With
international research on self-leadership taking off (e.g., Ho & Nesbit, 2014), we are in the
early stages of understanding the role of power distance in the practice of self-leadership.
Our results suggest that self-leadership has a stronger association with outcomes in high
power distance cultures. This is an important insight in light of conceptual work
24 Michael B. Harari et al.
suggesting that self-leadership might be restricted in employees in high power distance
cultures (Alves et al., 2006). Apparently, self-leadership is effective in high power distance
cultures (i.e., the effectiveness of self-leadership is not detrimentally influenced by such
contexts). The finding that contextual factors (e.g., power distance) may strengthen the
effects of self-leadership has implications for theorizing on the role of national culture in
understanding self-leadership. Local and culture-specific interpretations of self-leadership
can be explored to reveal how self-leadership is conceptualized and practiced in different
cultures. While Western contexts might interpret self-leadership as expressing autonomy,
in high power distance contexts, it might be interpreted as being a responsible employee
who acts to increase personal productivity. Thus, encouraging its practice across cultural
contexts.
Practical implications
Although our study is correlational, we nonetheless bring to light a few implications for
the development of self-leadership interventions. First, the meaningful zero-order
correlations between self-leadership and proposed outcomes (across study design
characteristics) provide some initial insight suggesting that training in self-leadership
might be effective in promoting productive cognition, attitudes, and behaviours among
employees. Second, our cross-cultural analyses suggest that self-leadership has meaning-
ful effects on studied outcomes globally, but that efforts to train self-leadership might pay
particular dividends in high power distance cultures. Third, when establishing a self-
leadership training programme, developers must choose which strategies to train
(Furtner, Sachse, & Exenberger, 2012; Stewart, Carson, & Cardy, 1996; Unsworth &
Mason, 2012). Our analyses suggested that different outcome variables are predominantly
affected by different self-leadership strategies. Although broadly covering the three
strategies might result in the broadest benefit, it might not be possible when considering
constraints on time, money, and other resources. Thus, strategies can be chosen for
training based upon the desired outcome. For example, efforts to improve self-effica cycan
emphasize constructive thought strategies, as informed by our findings. Given our finding
that self-efficacy is a key mechanism influencing the relationships between self-leadership
and job performance, this emphasis can result in significant benefits to organizations.
Limitations and strengths
Most findings were consistent with predictions from the literature on self-leadership,
though the relationship between self-leadership and other-rated job performance was
small and, as we observed evidence of publication bias, likely an overestimate of the true
effect. This effect was also deeply heterogeneous. In other words, results suggested that
self-leadership can, but not necessarily will, affect job performance. Efforts to elucidate
when self-leadership does predict job performance are needed to better understand the
role of self-leadership in organizational settings and combat publication bias in this
literature.
As a meta-analytic review, we cannot consider characteristics that vary between-
people that would be relevant for informing our understanding of boundary conditions
around self-leadership relationships (Stewart et al., 2011). Although our moderator
analyses dealing with power distance (a characteristic that varied between-study and was
thus amenable to meta-analytic methods) yielded novel insights, other characteristics that
vary between-people versus between-study could not be explored.
Self-leadership meta-analysis 25
While power distance is a theoretically relevant cultural dimension as it pertains to self-
leadership, we did not have strong theory to suggest what the specific nature of this
moderating effect would be. We, therefore, encourage future primary studies to attempt
to replicate the effect observed here. However, we do not believe that this finding reflects
capitalization on chance, given that results were consistent across outcome variables.
Although the FFM constitutes a consensus taxonomy of normal personality, there are
alternate means of conceptualizing personality that have value and can aid further in
understanding the dispositional basis of self-leadership. We discuss this further in the
future directions section, below.
Future directions
Our use of SCT to develop the hypotheses that guided this meta-analytic study has
highlighted opportunities for additional future research. Being grounded in SCT, self-
leadership involves self-reflection and self-regulation. Traits that govern one’s self-views,
influencing their judgments about themselves and their capabilities, should play a role. In
particular, we are referring to core self-evaluations (CSEs; Judge, Locke, & Durham, 1997)
–a compound trait dealing with the assumptions that individuals make about their
capabilities. CSEs affect behaviours via diminished avoidance motivations; individuals
with high CSEs are confident in their capabilities and thus resilient in the face of setbacks
(Connelly et al., 2018; Ferris et al., 2011). Given this, we were surprised to see limited
empirical investigations attempting to empirically eva luate the link between CSEs and self-
leadership; we encourage research to explore this.
Social cognitive theory emphasized the role of self-efficacy, as expectancies about
anticipated outcomes drive behaviours (Bandura, 1986). Our finding that self-efficacy is
indeed an important mediator of the self-leadership to performance relationship can be
extended to understand the mediating role in relation to other outcomes of self-
leadership, such as work attitudes (and resulting reactions, such as absenteeism and work
anxiety –Frayne & Latham, 1987; Saks & Ashforth, 1996), career expectations, and career
success (Murphy & Ensher, 2001). More comprehensive process models can increase our
understanding of how self-efficacy impacts organization members and outcomes through
additional mediators, improving our ability to engage high-performing members.
Research is also needed to determine the broader mediating effects of self-efficacy, as
well as potential facilitating relationships within other, more complex relationships (e.g.,
mediated moderated relationships).
National culture has been discussed with observations that self-leadership can be
found in a variety of cultural contexts. With no strong theory to suggest the specific nature
of the potential moderating effect of cultural dimensions, more research is needed to
explore self-leadership in non-Western samples to determine the generalizability of the
strategies to better understand its effects. Hofstede (1984) suggested that employees in
high power distance cultures who also hold high power distance values might feel lost in
an organization that embraces low power distance. They might rely on superiors
providing structure rather than autonomy. Expanding on the findings here, tests that
include power distance orientation of the organization might prove useful from an SCT
perspective, explicating differences between organizational environments as it pertains
to power distance, allowing for expanded theory building.
Work is needed to consider theoretically relevant moderators of self-leadership
relations that vary between-people versus between-study. The role of the work context
poses one such opportunity. For example, to the extent that one’s supervisor takes a top-
26 Michael B. Harari et al.
down approach to management and prefers oversight of employee activities versus
independence, relations involving self-leadership and job performance might be reduced
(Meyer, Dalal, & Hermida, 2010). Whereas in an autonomous environment, engagement
in self-leadership strategies might be a prerequisite to performance and its effects would
be more likely to show up in observer ratings of job performance. We encourage a
consideration of the broader work context at the primary study level.
Although carried out as a study quality sensitivity analysis, our study design moderator
analysis –which demonstrated larger correlations from CS/SR studies as compared to
MW/MS studies –has implications for future work. Although point estimates differ, even
under the more robust MW/MS designs, our expectations were largely supported in the
data. The observation of anticipated relationships in the more rigorous study designs
provides more robust support for the contention that self-leadership is an important
construct within organizational settings.
Because fewer resources are needed to conduct a CS/SR study as compared to more
rigorous designs, CS/SR studies can be particularly useful early in the development of a
literature. Though, such work is less useful once a sufficient body of work exists
suggesting covariation (Spector, 2019). At this stage of the self-leadership literature, CS/SR
research dealing with correlates addressed in the present meta-analysis is of limited
further value. To continue to advance the literature, MW/MS designs should be preferred.
Finally, research that considers the complementarity among self-leadership strategies
(Neck & Houghton, 2006) is also merited. SCT suggests that self-efficacy is a primary
mechanism for producing performance outcomes and as suggested by our results,
behavioural strategies play an important role. The way in which behavioural strategies
complement each other or interact (even with other self-leadership strategy categories) to
produce outcomes is an area that can expand our understanding of the way the self-
leadership process influences work responses.
Conclusion
Research on employee self-leadership has discussed a variety of implications for
organizations, especially with respect to job performance. Selective and narrative reviews
note the positive effects on attitudes and work performance, however, these reviews
could not resolve discrepancies in results, facilitate empirical comparisons of discrep-
ancies in findings between the three self-leadership strategies, or comprehensively
synthesize the entirety of empirical research dealing with self-leadership. Our results
support the value of self-leadership in organizations and affirm the importance of
considering the role of each self-leadership strategy.
Conflicts of interest
All authors declare no conflict of interest.
Author contributions
Michael B Harari: Data curation (equal); Formal analysis (equal); Methodology (equal);
Writing –original draft (equal); Writing –review & editing (equal). Ethlyn Williams:
Conceptualization (equal); Data curation (equal); Writing –original draft (equal); Writing
–review & editing (equal). Stephanie Castro: Data curation (equal); Writing –review &
Self-leadership meta-analysis 27
editing (equal). Katarina Brant: Data curation (equal); Writing –review & editing
(equal).
Data availability statement
Research data are not shared.
References
Aguinis, H., Ramani, R. S., & Alabduljader, N. (2018). What you see is what you get? Enhancing
methodological transparency in management research. Academy of Management Annals,12
(1), 83–110. https://doi.org/10.5465/annals.2016.0011
Alves, J. C., Lovelace, K. J., Manz, C. C., Matsypura, D., Toyasaki, F., & Ke, K. G. (2006). A cross-
cultural perspective of self-leadership. Journal of Managerial Psychology,21(4), 338–359.
https://doi.org/10.1108/02683940610663123
Amabile, T. M. (1988). A model of creativity and innovation in organizations. In B.S. Cummings (Ed.),
Research in organizational behavior (pp. 123–167). Greenwich, CT: JAI Press.
Amundsen, S., & Martinsen, Ø. L. (2014). Empowering leadership: Construct clarification,
conceptualization, and validation of a new scale. The Leadership Quarterly,25, 487–511.
https://doi.org/10.1016/j.leaqua.2013.11.009
Anderson, J. S., & Prussia, G. E. (1997). The Self-Leadership Questionnaire: Preliminary assessment
of construct validity. Journal of Leadership Studies,4(2), 119–143. https://doi.org/10.1177/
107179199700400212
Andressen, P., Konradt, U., & Neck, C. P. (2012). The relation between self-leadership and
transformational leadership: Competing models and the moderating role of virtuality. Journal of
Leadership & Organizational Studies,19(1), 68–82. https://doi.org/10.1177/
1548051811425047
Bailey, S. F., Barber, L. K., & Justice, L. M. (2016). Is self-leadership just self-regulation? Exploring
construct validity with HEXACO and self-regulatory traits. Current Psychology,37(1), 149–161.
https://doi.org/10.1007/s12144-016-9498-z
Bandura, A. (1977). Social learning theory. Englewood Cliffs, NJ: Prentice Hall.
Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory.
Englewood Cliffs, NJ: Prentice-Hall.
Bandura, A. (2001). Social cognitive theory: An agentic perspective. Annual Review of Psychology,
52(1), 1–26. https://doi.org/10.1146/annurev.psych.52.1.1
Bandura, A., & Cervone, D. (1986). Differential engagement of self-reactive influences in cognitive
motivation. Organizational Behavior and Human Decision Processes,38(1), 92–113. https://
doi.org/10.1016/0749-5978(86)90028-2
Barsky, A., Kaplan, S. A., & Beal, D. J. (2011). Just feelings? The role of affect in the formation of
organizational fairness judgments. Journal of Management,37(1), 248–279. https://doi.org/
10.1177/0149206310376325
Bommer, W. H., Johnson, J. L., Rich, G. A., Podsakoff, P. M., & Mackenzie, S. B. (1995). On the
interchangeability of objective and subjective measures of employee performance: A meta-
analysis. Personnel Psychology,48, 587–605. https://doi.org/10.1111/j.1744-6570.1995.tb
01772.x
Brannick, M. T., Chan, D., Conway, J. M., Lance, C. E., & Spector, P. E. (2010). What is method
variance and how can we cope with it? A panel discussion. Organizational Research Methods,
13, 407–420. https://doi.org/10.1177/1094428109360993
Brown, R. T., & Fields, D. (2011). Leaders engaged in self-leadership: Can followers tell the
difference? Leadership,7(3), 275–293. https://doi.org/10.1177/1742715011407383
28 Michael B. Harari et al.
Brutus, S., Aguinis, H., & Wassmer, U. (2013). Self-reported limitations and future directions in
scholarly reports: Analysis and recommendations. Journal of Management,39(1), 48–75.
https://doi.org/10.1177/0149206312455245
Carver, C. S., & Scheier, M. F. (1998). On the self-regulation of behavior. New York, NY: Cambridge
University Press.
Chung, A., Chen, I., Lee, A. Y., Chen, H. C., & Lin, Y. (2011). Charismatic leadership and self-
leadership: A relationship of substitution or supplementation in the contexts of internalization
and identification? Journal of Organizational Change Management,24, 299–313. https://doi.
org/10.1108/09534811111132703
Cohen, J. (1992). A power primer. Psychological Bulletin,112(1), 155–159. https://doi.org/10.
1037/0033-2909.112.1.155
Connelly, B. S., Ones, D. S., & Hulsheger, U. R. (2018). Personality in industrial, work, and
organizational psychology: Theory measurement and applications. In D.S. Ones, N. Anderson, C.
Viswesvaran & H.K. Sinangil (Eds.), The Sage handbook of industrial, work, and
organizational psychology (Vol. 1, pp. 320–365). London, UK: Sage.
Courtright, S. H., Thurgood, G. R., Stewart, G. L., & Pierotti, A. J. (2015). Structural interdependence
in teams: An integrative framework and meta-analysis. Journal of Applied Psychology,100,
1825–1846. https://doi.org/10.1037/apl0000027
Cox, J. F. (1993). The effects of superleadership training on leader behavior, subordinate self-
leadership behavior, and subordinate citizenship (Order No. 9507927). Available from ProQuest
Dissertations & Theses Global. (304092384). Retrieved from http://ezproxy.fau.edu/login?url=
http://resolver.ebscohost.com/openurl?url=https://search.proquest.com/docview/
304092384?accountid=10902
Deci, E. (1975). Intrinsic motivation. New York, NY: Plenum.
DeYoung, C. G. (2015). Cybernetic big five theory. Journal of Research in Personality,56,33–58.
https://doi.org/10.1016/j.jrp.2014.07.004
DeYoung, C. G., Peterson, J. B., & Higgins, D. M. (2005). Sources of openness/intellect: Cognitive
and neuropsychological correlates of the fifth factor of personality. Journal of Personality,73,
825–858. https://doi.org/10.1111/j.1467-6494.2005.00330.x
DeYoung, C. G., Quilty, L. C., & Peterson, J. B. (2007). Between facets and domains: 10 aspects of the
Big Five. Journal of Personality and Social Psychology,93, 880–896. https://doi.org/10.1037/
0022-3514.93.5.880
Digman, J. M. (1990). Personality structure: Emergence of the five-factor model. Annual Review of
Psychology,41(1), 417–440. https://doi.org/10.1146/annurev.ps.41.020190.002221
Ferris, D. L., Rosen, C. R., Johnson, R. E., Brown, D. J., Risavay, S. D., & Heller, D. (2011). Approach or
avoidance (or both?): Integrating core self-evaluations with an approach/avoidance framework.
Personnel Psychology,64, 137–161. https://doi.org/10.1111/j.1744-6570.2010.01204.x
Frayne, C. A., & Geringer, J. M. (2000). Self-management training for improving job performance: A
field experiment involving salespeople. Journal of Applied Psychology,85(3), 361. https://doi.
org/10.1037/0021-9010.85.3.361
Frayne, C. A., & Latham, G. P. (1987). Application of social-learning theory to employee self-
management of attendance. Journal of Applied Psychology,72, 387–392. https://doi.org/10.
1037/0021-9010.72.3.387
Frese, M., & Fay, D. (2001). Personal initiative: An active performance concept for work in the 21st
century. Research in Organizational Behavior,23, 133–187. https://doi.org/10.1016/S0191-
3085(01)23005-6
Furtner, M. R., & Rauthmann, J. F. (2010). Relations between self-leadership and scores on the Big
Five. Psychological Reports,107, 339–353. https://doi.org/10.2466/02.03.14.20.PR0.107.5.
339-353
Furtner, M. R., Sachse, P., & Exenberger, S. (2012). Learn to influence yourself: Full range self-
leadership training. Journal of the Indian Academy of Applied Psychology,38, 294–304.
Gagn
e, M., & Deci, E. L. (2005). Self-determination theory and work motivation. Journal of
Organizational Behavior,26(4), 331–362. https://doi.org/10.1002/job.322
Self-leadership meta-analysis 29
Guzzo, R. A. (1998). Leadership, self-management, and levels of analysis. In F. Dansereau & F.J.
Yammarino (Eds.), Leadership: The multiple-level approaches, classical and new wave (pp.
213–219). Stanford, CT: JAI Press.
Harari, M. B., Manapragada, A., & Viswesvaran, C. (2017). Who thinks they’re a big fish in a small
pond and why does it matter? A meta-analysis of perceived overqualification. Journal of
Vocational Behavior,102,28–47. https://doi.org/10.1016/j.jvb.2017.06.002
Harari, M. B., Parola, H. R., Hartwell, C., & Riegelman, A. (2020). Literature searches in systematic
reviews and meta-analyses: A review, evaluation, and recommendations. Journal of Vocational
Behavior,118, 103377. https://doi.org/10.1016/j.jvb.2020.103377
Harari, M. B., Reaves, A. C., Beane, D. A., Laginess, A. J., & Viswesvaran, C. (2018). Personality and
expatriate adjustment: A meta-analysis. Journal of Occupational and Organizational
Psychology,91, 486–517. https://doi.org/10.1111/joop.12215
Harari, M. B., Reaves, A. C., & Viswesvaran, C. (2016). Creative and innovative performance: A meta-
analysis of relationships with task, citizenship, and counterproductive job performance
dimensions. European Journal of Work and Organizational Psychology,25, 495–511. https://
doi.org/10.1080/1359432X.2015.1134491
Harari, M. B., Thompson, A. H., & Viswesvaran, C. (2018). Extraversion and job satisfaction: The role
of trait bandwidth and moderating effect of status goal attainment. Personality and Individual
Differences,123(1), 14–16. https://doi.org/10.1016/j.paid.2017.10.041
Harari, M. B., & Viswesvaran, C. (2018). Individual Job Performance. In D.S. Ones, N. Anderson, C.
Viswesvaran & H.K. Sinangil (Eds.), The SAGE handbook of industrial, work & organizational
psychology: V1: Personnel psychology and employee performance (pp. 55–72). London, UK:
Sage.
Harrison, J. S., Banks, G. C., Pollack, J. M., O’Boyle, E. H., & Short, J. (2017). Publication bias in
strategic management research. Journal of Management,43, 400–425. https://doi.org/10.
1177/0149206314535438
Hart, J. W., Stasson, M. F., Mahoney, J. M., & Story, P. (2007). The Big Five and achievement
motivation: Exploring the relationship between personality and a two-factor model of
motivation. Individual Differences Research,5, 267–274. https://psycnet.apa.org/record/
2007-18918-003
Hayes, A. F., & Rockwood, N. J. (2017). Regression-based statistical mediation and moderation
analysis in clinical research: Observations, recommendations, and implementation. Behaviour
Research and Therapy,98,39–57. https://doi.org/10.1016/j.brat.2016.11.001
Ho, J., & Nesbit, P. L. (2014). Self-leadership in a Chinese context: Work outcomes and the
moderating role of job autonomy. Group & Organization Management,39, 389–415. https://
doi.org/10.1177/1059601114539389
Hofstede, G. (1984). The cultural relativity of the quality of life concept. Academy of Management
Review,9, 389–396. https://doi.org/10.2307/258280
Hofstede, G. (2001). Culture’s consequences: Comparing values, behaviors, institutions and
organizations across nations. Thousand Oaks, CA: Sage Publications.
Hofstede, G., Hofstede, G. J., & Minkov, M. (2005). Cultures and organizations: Software of the
mind, Vol. 2. New York, NY: Mcgraw-hill.
Holt, S., Hall, A., & Gilley, A. (2018). Essential components of leadership development programs.
Journal of Managerial Issues,30, 214–153. https://go.openathens.net/redirector/fau.edu?url=
https://www.proquest.com/scholarly-journals/essential-components-leadership-developme
nt/docview/2186639172/se-2?accountid=10902
Houghton, J. D., Bonham, T. W., Neck, C. P., & Singh, K. (2004). The relationship between self-
leadership and personality. Journal of Managerial Psychology,19, 427–441. https://doi.org/
10.1108/02683940410537963
Houghton, J. D., Carnes, A., & Ellison, C. N. (2014). A cross-cultural examination of self-leadership:
Testing for measurement invariance across four cultures. Journal of Leadership and
Organizational Studies,21, 414–430. https://doi.org/10.1177/1548051813515753
30 Michael B. Harari et al.
Houghton, J. D., & Jinkerson, D. L. (2007). Constructive thought strategies and job satisfaction:
Preliminary examination. Journal of Business and Psychology,22(1), 45–53. https://doi.org/
10.1007/s10869-007-9046-9
Houghton, J. D., & Neck, C. P. (2002). The revised self-leadership questionnaire: Testing a
hierarchical factor structure for self-leadership. Journal of Managerial Psychology,17, 672–
691. https://doi.org/10.1108/02683940210450484
Hunter, J. E., & Schmidt, F. L. (2004). Methods of meta-analysis: Correcting error and bias in
research findings (2nd ed.). Newbury Park, CA: Sage.
Iverson, R. D., & Maguire, C. (2000). The relationship between job and life satisfaction: Evidence
from a remote mining community. Human Relations,53, 807–839. https://doi.org/10.1177/
0018726700536003
John, O. P., Naumann, L. P., & Soto, C. J. (2008). Paradigm shift to the integrative big-five trait
taxonomy: History, measurement, and conceptual issues. In O.P. John, R.W. Robins & L.A.
Pervin (Eds.), Handbook of personality: Theory and research (pp. 114–158). New York: NY:
Guilford Press.
Judge, T. A., Erez, A., Bono, J. E., & Thoresen, C. J. (2002). Are measures of self-esteem, neuroticism,
locus of control, and generalized self-efficacy indicators of a common core construct? Journal of
Personality and Social Psychology,83, 693–710. https://doi.org/10.1037/0022-3514.83.3.693
Judge, T. A., & Ilies, R. (2002). Relationship of personality to performance motivation: A meta-
analytic review. Journal of Applied Psychology,87, 797–807. https://doi.org/10.1037/0021-
9010.87.4.797
Judge, T. A., Locke, E. A., & Durham, C. C. (1997). The dispositional causes of job satisfaction: A core
evaluations approach. Research in Organizational Behavior,19, 151–188.
Kepes, S., Banks, G. C., McDaniel, M., & Whetzel, D. L. (2012). Publication bias in the organizational
sciences. Organizational Research Methods,15, 624–662. https://doi.org/10.1177/
1094428112452760
Kepes, S., & McDaniel, M. A. (2015). The validity of conscientiousness is overestimated in the
prediction of job performance. PLoS One,10(10), e0141468. https://doi.org/10.1371/journal.
pone.0141468
Konradt, U., Andreben, P., & Ellwart, T. (2009). Self-leadership in organizational teams: A multilevel
analysis of moderators and mediators. European Journal of Work and Organizational
Psychology,18, 322–346. https://doi.org/10.1080/13594320701693225
Locke, E. A., Shaw, K. N., Saari, L. M., & Latham, G. P. (1981). Goal setting and task performance:
1969–1980. Psychological Bulletin,90(1), 125–152. https://doi.org/10.1037/0033-2909.90.1.
125
MacKinnon, D. P., Lockwood, C. M., Hoffman, J. M., West, S. G., & Sheets, V. (2002). A comparison of
methods to test mediation and other intervening variable effects. Psychological Methods,7,83–
104. https://doi.org/10.1037/1082-989X.7.1.83
Manz, C. C. (1986). Self-leadership: Toward an expanded theory of self-influence processes in
organizations. Academy of Management Review,11, 585–600. https://doi.org/10.5465/amr.
1986.4306232
Manz, C. C. (1992). Self-leadership. The heart of empowerment. The Journal for Quality and
Participation,15,80–85.
Manz, C. C. (2015). Taking the self-leadership high road: Smooth surface or potholes ahead? The
Academy of Management Perspectives,29, 132–151. https://doi.org/10.5465/amp.2013.0060
Manz, C. C., Adsit, D., Campbell, S., & Mathison-Hance, M. (1988). Managerial thought patterns and
performance: A study of perceptual patterns of performance hindrances for higher and lower
performing managers. Human Relations,41, 447–465. https://doi.org/10.1177/
001872678804100603
Manz, C. C., & Neck, C. P. (2004). Mastering self-leadership: Empowering yourself for personal
excellence (3rd ed.). Upper Saddle River, NJ: Prentice-Hall.
Manz, C. C., & Sims, Jr, H. P. (1980). Self-management as a substitute for leadership: A social learning
theory perspective. Academy of Management Review,5(3), 361–367.
Self-leadership meta-analysis 31
Manz, C. C., & Sims, Jr, H. P. (1987). Leading workers to lead themselves: The external leadership of
self-managing work teams. Administrative Science Quarterly,32, 106–129. https://doi.org/10.
2307/2392745
Markham, S. E., & Markham, I. S. (1998). Self-management and self-leadership reexamined: A levels
of analysis perspective. In F. Dansereau & F.J. Yammarino (Eds.), Leadership: The multiple-level
approaches, classical and new wave (pp. 193–210). Stanford, CT: JAI Press.
McCrae, R. R., & L€
ockenhoff, C. E. (2010). Self-regulation and the five-factor model of personality
traits. In R.H. Hoyle (Ed.), Handbook of personality and self-regulation (pp. 145–168). Wiley-
Blackwell.
Meyer, J. P., & Allen, N. J. (1991). A three-component conceptualization of organizational
commitment. Human Resource Management Review,1(1), 61–89. https://doi.org/10.1016/
1053-4822(91)90011-Z
Meyer, R. D., Dalal, R. S., & Hermida, R. (2010). A review and synthesis of situational strength in the
organizational sciences. Journal of Management,36(1), 121–140. https://doi.org/10.1177/
0149206309349309
Mosier, C. I. (1943). On the reliability of a weighted composite. Psychometrika,8(3), 161–168.
Mowday, R. T., Steers, R. M., & Porter, L. W. (1979). The measurement of organizational
commitment. Journal of Vocational Behavior,14, 224–247. https://doi.org/10.1016/0001-
8791(79)90072-1
Murphy, S. E., & Ensher, E. A. (2001). The role of mentoring support and self-management strategies
on reported career outcomes. Journal of Career Development,27, 229–246. https://doi.org/
10.1177/089484530102700401
Neck, C. P., & Houghton, J. D. (2006). Two decades of self-leadership theory and research: Past
developments, present trends, and future possibilities. Journal of Managerial Psychology,21,
270–295. https://doi.org/10.1108/02683940610663097
Neck, C. P., & Manz, C. C. (1992). Thought self-leadership: The influence of self-talk and mental
imagery on performance. Journal of Organizational Behavior,13(7), 681–699.
Neck, C. P., & Manz, C. C. (1996). Thought self-leadership: The impact of mental strategies training
on employee cognition, behavior, and affect. Journal of Organizational Behavior,17(5), 445–
467. https://doi.org/10.1002/(SICI)1099-1379(199609)17:5<445:AID-JOB770>3.0.CO;2-N
Neck, C. P., Manz, C. C., & Houghton, J. D. (2017). Self-leadership: The definitive guide to personal
excellence. Thousand Oaks: CA. SAGE Publications.
Neubert, M. J., & Wu, J. C. (2006). An investigation of the generalizability of the Houghton and Neck
Revised Self-Leadership Questionnaire to a Chinese context. Journal of Managerial
Psychology,21(4), 360–373. https://doi.org/10.1108/02683940610663132
Norris, S. E. (2008). An examination of self-leadership. Emerging Leadership Journeys,1,43–61.
Ones, D. S., Viswesvaran, C., & Reiss, A. D. (1996). Role of social desirability in personality testing for
personnel selection: The red herring. Journal of Applied Psychology,81, 660–679. https://doi.
org/10.1037/0021-9010.81.6.660
Prussia, G. E., Anderson, J. S., & Manz, C. C. (1998). Self-leadership and performance outcomes: The
mediating influence of self-efficacy. Journal of Organizational Behavior,19, 523–538. https://
doi.org/10.1002/(SICI)1099-1379(199809)19:5<523:AID-JOB860>3.0.CO;2-I
Reichard, R. J., & Johnson, S. K. (2011). Leader self-development as organizational strategy. The
Leadership Quarterly,22(1), 33–42. https://doi.org/10.1016/j.leaqua.2010.12.005
Roberts, H. E., & Foti, R. J. (1998). Evaluating the interaction between self-leadership and work
structure in predicting job satisfaction. Journal of Business and Psychology,12(3), 257–267.
https://doi.org/10.1023/A:1025067128193
Sahin, F. (2015). The convergent, discriminant, and concurrent validity of scores on the abbreviated
self-leadership questionnaire. The Journal of Human Work,2,91–104. https://doi.org/10.
18394/iid.25158
Saks, A. M., & Ashforth, B. E. (1996). Proactive socialization and behavioral self-management.
Journal of Vocational Behavior.,48, 301–323. https://doi.org/10.1006/jvbe.1996.0026
32 Michael B. Harari et al.
Schaub, M., & Tokar, D. M. (2005). The role of personality and learning experiences in social
cognitive career theory. Journal of Vocational Behavior,66, 304–325. https://doi.org/10.
1016/j.jvb.2004.09.005
Schmidt, F. L. (2017). Statistical and measurement pitfalls in the use of meta-regression in meta-
analysis. Career Development International,22, 469–476. https://doi.org/10.1108/CDI-08-
2017-0136
Schmidt, F. L., & Hunter, J. E. (2015). Methods of meta-analysis: Correcting error and bias in
research findings. New York, NY: Sage. https://doi.org/10.4135/9781483398105
Schmidt, F. L., Viswesvaran, C., Ones, D. S., & Le, H. (2017). A failed challenge to validity
generalization: Addressing a fundamental misunderstanding of the nature of VG. Industrial and
Organizational Psychology,10(3), 488. https://doi.org/10.1017/iop.2017.47
Scullen, S. E., Mount, M. K., & Goff, M. (2000). Understanding the latent structure of job performance
ratings. Journal of Applied Psychology,85, 956–970. https://doi.org/10.1037/0021-9010.85.6.
956
Sesen, H., Tabak, A., & Arli, O. (2017). Consequences of self-leadership: A study on primary school
teachers. Educational Sciences: Theory & Practice,17, 945–968. https://doi.org/10.12738/
estp.2017.3.0520
Spector, P. E. (2006). Method variance in organizational research: Truth or urban legend?
Organizational Research Methods,9, 221–232. https://doi.org/10.1177/1094428105284955
Spector, P. E. (2019). Do not cross me: Optimizing the use of cross-sectional designs. Journal of
Business and Psychology,34(2), 125–137. https://doi.org/10.1007/s10869-018-09613-8
Spector, P. E., Rosen, C. C., Richardson, H. A., Williams, L. J., & Johnson, R. E. (2017). A new
perspective on method variance: A measure-centric approach. Journal of Management,45,
855–880. https://doi.org/10.1177/0149206316687295
Steinbauer, R., Renn, R. W., Taylor, R. R., & Njoroge, P. K. (2014). Ethical leadership and followers’
moral judgment: The role of followers’ perceived accountability and self-leadership. Journal of
Business Ethics,120, 381–392. https://doi.org/10.1007/s10551-013-1662-x
Stewart, G. L., Carson, K. P., & Cardy, R. L. (1996). The joint effects of conscientiousness and self-
leadership training on employee self-directed behavior in a service setting. Personnel
Psychology,49(1), 143–164. https://doi.org/10.1111/j.1744-6570.1996.tb01795.x
Stewart, G. L., Courtright, S. H., & Manz, C. C. (2011). Self-leadership: A multilevel review. Journal of
Management,37(1), 185–222. https://doi.org/10.1177/0149206310383911
Stewart, G. L., Courtright, S. H., & Manz, C. C. (2019). Self-leadership: A paradoxical core of
organizational behavior. Annual Review of Organizational Psychology and Organizational
Behavior,6,47–67. https://doi.org/10.1146/annurev-orgpsych-012218-015130
Toglaw, S. D. (2006). Organizational commitment in a virtual work environment (Unpublished
doctoral dissertation). University of Technology, Sydney.
Tonidandel, S., & LeBreton, J. M. (2011). Relative importance analysis: A useful supplement to
regression analysis. Journal of Business and Psychology,26,1–9. https://doi.org/10.1007/
s10869-010-9204-3
Tonidandel, S., & LeBreton, J. M. (2015). RWA-Web –A free, comprehensive, web-based, user-
friendly tool for relative weight analyses. Journal of Business and Psychology,30, 207–216.
https://doi.org/10.1007/s10869-014-9351-z
Unsworth, K. L., & Mason, C. M. (2012). Help yourself: The mechanisms through which a self-
leadership intervention influences strain. Journal of Occupational Health Psychology,17, 235–
245. https://doi.org/10.1037/a0026857
van Kortenhof, M. C. (2013). E-working, leadership and employee characteristics: The effect of
transformational leadership and employee characteristics on successful performance in an
E-working environment (Unpublished doctoral dissertation). Eindhoven University of
Technology.
Viswesvaran, C., & Ones, D. S. (1995). Theory testing: Combining psychometric meta-analysis and
structural equations modeling. Personnel Psychology,48, 865–885. https://doi.org/10.1111/j.
1744-6570.1995.tb01784.x
Self-leadership meta-analysis 33
Viswesvaran, C., & Ones, D. S. (2000). Perspectives on models of job performance. International
Journal of Selection and Assessment,8, 216–226. https://doi.org/10.1111/1468-2389.00151
Viswesvaran, C., Schmidt, F. L., & Ones, D. S. (2002). The moderating influence of job performance
dimensions on convergence of supervisory and peer ratings of job performance: Unconfounding
construct-level convergence and rating difficulty. Journal of Applied Psychology,87, 345–354.
https://doi.org/10.1037/0021-9010.87.2.345
Viswesvaran, C., Schmidt, F. L., & Ones, D. S. (2005). Is there a general factor in ratings of job
performance? A meta-analytic framework for disentangling substantive and error influences.
Journal of Applied Psychology,90(1), 108–131. https://doi.org/10.1037/0021-9010.90.1.108
Watson, D., & Clark, L. A. (1997). Extraversion and its positive emotional core. In R. Hogan, J.
Johnson & S. Briggs (Eds.), Handbook of personality psychology (pp. 767–793). San Diego, CA:
Academic Press. https://doi.org/10.1016/B978-012134645-4/50030-5
Zimmerman, B. J., Bandura, A., & Martinez-Pons, M. (1992). Self-motivation for academic attainment:
The role of self-efficacy beliefs and personal goal setting. American Educational Research
Journal,29, 663–676. https://doi.org/10.3102/00028312029003663
Received 27 July 2019; revised version received 13 June 2021
Supporting Information
The following supporting information may be found in the online edition of the article:
Table S1. Meta-analytically derived intercorrelations among self-leadership strategies
Table S2. Results of mediation analyses (H11)
Table S3. Meta-analysis of self-leadership-performance dimension correlations
Table S4. Regression and relative weight analyses of job performance dimensions on
self-leadership dimensions
34 Michael B. Harari et al.
