The Comprehensive Meta-Analyses of the Nomological Network of

Psychological Capital (PsyCap)

Abstract
This paper presents the most rigorous meta-analysis undertaken to date of empirical literature

examining antecedents and outcomes related to psychological capital (PsyCap), and

moderators of these relationships. We investigated seven leadership styles as antecedents of

PsyCap (authentic, ethical, servant, empowering, transactional, transformational, and abusive

leadership), five outcomes (burnout, turnover intentions, work engagement, performance, and

satisfaction), and the impact of four moderators (country of sample origin, cultural

characteristics, industry type, and research design). Our analysis of PsyCap research (2007-

2020) examined 244 studies (254 independent samples and over 96000 participants), which is

over twice as large as previous PsyCap meta-analyses. To optimise the quality and reliability

of findings, we corrected for artefacts and included heterogeneity, sensitivity, and publication

bias analyses. Our results provide several new findings beyond previous PsyCap meta-

analyses. We found that empowering, servant, transformational, and transactional leadership

were all positively associated with PsyCap, with empowering and transformational leadership

being the strongest antecedents of PsyCap and abusive and transactional leadership being the

weakest. The findings demonstrated PsyCap was positively associated with work engagement,

and negatively associated with burnout. Country of sample origin moderated all the

relationships, except for servant leadership. Additionally, cultural characteristics (e.g., power

distance, masculinity, long-term orientation, and uncertainty avoidance) moderated several

conceptual relationships. Study design was also found to moderate the PsyCap - work

engagement relationship. Collectively, these findings offer new and extended insights into the

antecedents, outcomes, and moderators related to PsyCap, beyond previous meta-analyses. The

theoretical and practical implications of these new findings are also discussed.

Keywords: meta-analysis; psychological capital; PsyCap; leadership; employee outcomes

1
 Introduction

Psychological capital (PsyCap) reflects an individual’s state of positive development

and is characterised by hope, efficacy, resilience, and optimism (Luthans, Youssef, & Avolio,

2007). Research over the past 15 years has demonstrated that PsyCap is a meaningful resource

that can promote desirable employee outcomes (e.g., job satisfaction, organizational

commitment, and employee creativity) and buffer against negative employee outcomes (e.g.,

turnover intentions and cynicism) (Avey, Reichard, Luthans, & Mhatre, 2011; Fontes & Dello

Russo, 2020; Huang & Luthans, 2015). These relationships have now also been established in

a small number of meta-analyses studies (e.g., Avey, Reichard, et al., 2011; Kong, Tsai, Tsai,

Huang, & de la Cruz, 2018; Wu & Nguyen, 2019). However, we argue that these meta-analyses

have been underdeveloped in terms of both scope and methodology, which has led to a

restricted understanding regarding the antecedents of PsyCap, and the factors that may

moderate the relationships between PsyCap and its antecedents and outcomes. We further argue

that the flaws with the design of recent meta-analyses studies (Kong et al., 2018; Wu &

Nguyen, 2019) might limit the reliability of the produced results and the future implications.

Additionally, the ‘progress reports’ on the state of theoretical and empirical knowledge about

PsyCap over time provided by these studies both highlight and reinforce the need to ensure that

such efforts keep pace with the explosive growth of PsyCap literature. For example, a simple

keyword search for “psychological capital” conducted in May 2022 in the Scopus database

found that between 2007 and 2018 (the cut-off data collection point for Wu and Nguyen

(2019)), 717 PsyCap studies had been published. In the 2 years between that study and

December 2020 (the cut-off data collection point for this study), 461 new studies were

published, increasing the volume of PsyCap literature by 64%. Furthermore, the significant

recent increase in empirical research about PsyCap provides a valuable opportunity to both

extend the insights into the antecedents, outcomes related to PsyCap, and moderators of these

2
relationships provided by previous meta-analyses, and to also overcome the limitations of their

methodological approaches. In doing so, we expect to provide a more fine-grained

understanding of PsyCap, which can help inform organisational practices to maximise the

benefits of PsyCap for enhanced employee performance and functioning.

Limitations of Previous PsyCap Meta-Analyses

A small number of PsyCap meta-analyses have been conducted in recent years (Avey,

Reichard, et al., 2011; Kong et al., 2018; Wu & Nguyen, 2019) 1. These studies have aimed to

generate a greater understanding of PsyCap and the factors that influence its emergence, and

the relationship with important work-related outcomes. However, these meta-analyses have

been notably limited in terms of both scope and methodology. For example, despite previous

studies showing leadership to be a significant antecedent of PsyCap (e.g., Avey, 2014;

Bouckenooghe, Zafar, & Raja, 2014), extant PsyCap meta-analyses have rarely gone beyond

investigating authentic leadership (see e.g., Banks, McCauley, Gardner, & Guler, 2016; Hoch,

Bommer, Dulebohn, & Wu, 2018; Kong et al., 2018). Thus, other styles of leadership, such as

empowering, transformational, and transactional leadership, have yet to be fully investigated

in meta-analysis research. Moreover, the strength of these leadership styles in predicting

PsyCap has not been compared in previous PsyCap meta-analysis studies. As such, it is

currently unknown which leadership styles are the strongest predictors of PsyCap. Overall, the

current PsyCap meta-analyses do not draw a comprehensive picture of the relationship between

leadership and PsyCap, and do not compare the strength of different leadership styles in

predicting PsyCap. This is an important point as it has implications for both future research

and practice.

1
It is acknowledged that another meta-analysis of PsyCap has been recently published (Lupșa, Vîrga, Maricuțoiu,
& Rusu, 2020). However, this study examined the effectiveness of PsyCap intervention programs, rather than the
relationships between PsyCap and antecedent and outcome variables. Therefore, discussing this paper is out of
the scope of the current study.

3
 Furthermore, the investigation of moderators of the relationships between PsyCap and

other variables has also been limited in meta-analysis research. Two previous meta-analysis

studies (Avey, Reichard, et al., 2011; Wu & Nguyen, 2019) have investigated moderation

effects in the relationship between PsyCap and employee outcomes. However, neither

investigated potential moderation effects of the relationships between PsyCap and antecedent

variables, nor did they investigate moderation effects in the relationship between PsyCap and

negative employee outcomes (e.g., burnout). Similarly, Avey, Reichard, et al. (2011) only

investigated the moderation effects of sample origin and industry type in the relationship

between PsyCap and a combined group of positive outcomes (e.g., well-being, commitment,

and satisfaction). Thus, the results of this study cannot explain the conditional effect for each

individual relationship between PsyCap and the outcome variables and provide a simplistic

investigation of the moderators by combining the positive outcomes. We also suggest the

US/non-US dichotomy (Avey, Reichard, et al., 2011) represents a simplistic approach for

investigating the potential influence sample origin may have on PsyCap relationships and does

not provide a detailed analysis. Collectively, we argue that these identified omissions and

shortcomings across previous PsyCap meta-analyses limit current understandings of the factors

and boundary conditions of the relationships between PsyCap and antecedent and outcome

variables. Thus, there is still much yet to be discovered in terms of the moderators of PsyCap

relationships.

Besides, it is important to acknowledge that previous PsyCap meta-analyses (Avey,

Reichard, et al., 2011; Kong et al., 2018; Wu & Nguyen, 2019) have incurred consequential

methodological issues, which undermines the reliability of the findings reported. For instance,

failure to correct for the artefacts, particularly in meta-analysis studies that use correlation, can

result in misleading estimates of effect sizes (Schmidt & Hunter, 2015). Additionally, the

omission of heterogeneity analyses, sensitivity analyses, and publication bias analysis can also

4
adversely influence the findings of meta-analysis studies (Cleophas & Zwinderman, 2017). For

example, when sensitivity analyses (e.g., outlier analysis) are not conducted, the influence of

the results from low-quality individual studies on the overall meta-analysis is unknown. This

can lead to an over-estimation in reported findings stemming from the meta-analysis.

Therefore, it is important to investigate how results from lower quality studies may influence

the meta-analysis summary effect size (see Bown & Sutton, 2010; Cleophas & Zwinderman,

2017). Furthermore, it is well established that studies that have obtained significant results, or

strong effect sizes, are more likely to be published in comparison to studies that have not

obtained significant results or obtained weak effect sizes (Card, 2012; Schmidt & Hunter,

2015). In this situation, the published studies will not be representative of all studies (see

Borenstein, Hedges, Higgins, & Rothstein, 2009). To help remedy this, publication bias

analysis can be conducted to enhance the representativeness of the meta-analysis. However,

this important methodological consideration has not been included in any of the PsyCap meta-

analysis studies to date. Overall, in reviewing PsyCap meta-analyses conducted to date, it is

evident that these important methodological steps have not been routinely undertaken,

especially for the most recent studies (Kong et al., 2018; Wu & Nguyen, 2019). All in all, these

methodological and study design flaws undermine the findings of these studies and suggest

that their results need to be interpreted with caution.

The Current Study

This study aims to extend previous meta-analyses by undertaking a series of 13 meta-

analyses to identify significant antecedents and outcomes related to PsyCap, along with the

factors that moderate these relationships. In doing so, this study responds to calls for more in-

depth investigations of antecedents and moderators related to PsyCap (e.g., Luthans &

Youssef-Morgan, 2017; Luthans, Youssef, & Avolio, 2015; Newman, Ucbasaran, Zhu, &

Hirst, 2014) in several ways. First, it will provide an extended scope by including antecedent

5
(e.g., empowering, transformational, and transactionalleadership) and outcome (burnout and

work engagement) variables pertinent to PsyCap that have not been examined in prior meta-

analyses. In addition, it will provide the first meta-analytic comparison of the strength of

leadership styles in predicting employee PsyCap. As such, this study will provide a

comprehensive framework of the relationship between PsyCap and leadership and help

determine the effectiveness of various leadership styles in predicting employee PsyCap through

strong research syntheses.

Second, this study will extend previous investigations of the moderating effects of

sample origin and industry type (Avey, Reichard, et al., 2011), by investigating these effects

on relationships with individual outcome variables rather than grouped outcomes. This will

also enable the first meta-analytical investigation of moderating effects in the relationships

between PsyCap and both positive and negative outcome variables. Moreover, this study will

go beyond using a simple US/non-US dichotomy to examine the influence of sample origin.

Instead, we will code studies for each country and draw on Hofstede’s six-dimensional model

of national culture (Hofstede, Hofstede, & Minkov, 2010) to determine how different cultural

dimensions may influence the relationships between PsyCap and its outcomes and antecedents.

This is a novel application of the sample origin and national culture in a PsyCap meta-analysis

study and will provide a more elaborate understanding of the role of context and culture in the

nomological network of PsyCap. As such, this study will explore the moderation effects of

sample origin, culture, and industry type in the relationships between PsyCap and both

antecedent and outcome variables. Importantly, our study will aim to provide a more

methodologically rigorous meta-analysis of PsyCap by undertaking moderation analyses

(subgroup analysis and meta-regression) for each of the individual investigated relationships

and by employing corrected effect sizes, sensitivity analysis (outlier analysis), and publication

bias analysis.

6
 By conducting a more comprehensive and methodological rigorous meta-analysis of

PsyCap, the findings from this study will provide a deeper understanding regarding the

relationship between PsyCap and antecedent and outcome variables, as well as the moderator

variables of these relationships. Moreover, from a practical perspective, the findings from this

study will provide important insights about the situations in which, and for whom, PsyCap has

the greatest impact, which in turn can enable organisations to maximise the benefits of PsyCap.

The theoretical framework of PsyCap and research questions development

Antecedents of PsyCap

In more recent years, PsyCap research has moved away from solely investigating direct

relationships between PsyCap and outcomes to investigate factors that foster PsyCap (e.g.,

antecedents). Currently, one of the most frequently studied antecedents of PsyCap is leadership

including, empowering (Park, Kim, Yoon, & Joo, 2017), transformational (Gooty, Gavin,

Johnson, Frazier, & Snow, 2009), authentic (Amunkete & Rothmann, 2015; Avey, 2014;

Hystad, Bartone, & Eid, 2014; Malik & Dhar, 2017; Rego, Sousa, Marques, & Cunha, 2012;

Woolley, Caza, & Levy, 2011), servant (Bouzari & Karatepe, 2017; Karatepe & Talebzadeh,

2016), and ethical leadership (Avey, 2014; Bouckenooghe et al., 2014). Findings from these

studies have demonstrated that these leadership styles have a significant, positive influence on

employee PsyCap. Other research has also demonstrated the significant and negative impact of

abusive leadership on employee PsyCap (Agarwal, 2019; Agarwal & Avey, 2020; Ahmad,

Athar, Azam, Hamstra, & Hanif, 2018; Lee & Wu, 2016).

It has been argued that leaders can bring positivity to the organisation by engaging and

expressing behaviours that are either intrinsically positive, or lead to positive outcomes (Cunha,

Rego, Simpson, & Clegg, 2020). This is argued to be most evident in relation to positive forms

of leadership such as ethical, empowering, authentic, transformational, and servant leadership

(see e.g., Adams, Meyers, & Sekaja, 2020; Cunha et al., 2020; Marques, 2020; Stander &

7
Coxen, 2017). However, it is also suggested that behaviours that typify transactional

leadership, which is not considered as a positive leadership style per se, can still lead to positive

employee outcomes (Cunha et al., 2020). For example, Gardner, Avolio, Luthans, May, and

Walumbwa (2005, p. 345) purported that authentic leaders “draw from the positive

psychological states that accompany optimal self-esteem and psychological well-being, such

as confidence, optimism, hope, and resilience, to model and promote the development of these

states in others”. Drawing on this line of argument, we suggest that intrinsically positive

leadership styles, including authentic, ethical, empowering, transformation, and servant

leadership, can positively predict employee PsyCap. Conversely, we posit that negative styles

of leadership, such as abusive leadership, are likely to erode positive psychological states, such

as PsyCap in employees. Importantly, investigation of the relationship between various

leadership styles and PsyCap will enable the determination of those leadership styles with the

strongest (and weakest) association with PsyCap. Therefore, the following research questions

are proposed:

Research Question 1: Is authentic leadership positively associated with employee PsyCap?

Research Question 2: Is ethical leadership positively associated with employee PsyCap?

Research Question 3: Is transformational leadership positively associated with employee

PsyCap?

Research Question 4: Is transactional leadership positively associated with employee PsyCap?

Research Question 5: Is servant leadership positively associated with employee PsyCap?

Research Question 6: Is empowering leadership positively associated with employee PsyCap?

Research Question 7: Is abusive leadership negatively associated with employee PsyCap?

8
Research question 8: Which leadership styles have the strongest and weakest impact on

employee PsyCap?

Outcomes of PsyCap

Research over the past two decades has consistently demonstrated that PsyCap is a

significant predictor of a range of desirable outcomes, including job performance,

organisational citizenship behaviours, and organisational commitment (Nolzen, 2018).

Research has also shown that PsyCap negatively influences undesirable outcomes, including

cynicism (Wang, Chang, Fu, & Wang, 2012), turnover intentions, and burnout (Amunkete &

Rothmann, 2015; Manzano-García & Ayala, 2017).

Conservation of resources theory (Hobfoll, 1989) is the predominant theoretical

framework cited to explain the relationship between PsyCap and outcome variables

(Alessandri, Consiglio, Luthans, & Borgogni, 2018; Karatepe & Karadas, 2015; Newman,

Nielsen, Smyth, Hirst, & Kennedy, 2018). According to the COR theory, individuals strive to

keep, preserve, and build resources. Resources have been defined as “those objects, personal

characteristics, conditions, or energies that are valued by individual or that serve as a means

for attainment of these objects, personal characteristics, conditions, or energies” (Hobfoll,

1989, p. 516). COR theory differentiates between two mechanisms designed to conserve

resources: (1) the seeking of new resources which will aid energetic activation towards the

pursuit of goals (resource gain orientation); and (2) the propensity to prevent resource loss, and

thus retain a sufficient buffer of resources (resource loss orientation; Hobfoll, 1989). Moreover,

COR theory proposes the notion of resource caravans (Hobfoll, 2011; Hobfoll, Halbesleben,

Neveu, & Westman, 2018), whereby psychological resources travel together and interact

synergistically, as is purported with the components of PsyCap – hope, efficacy, resilience,

and optimism (Luthans & Youssef-Morgan, 2017).

9
 Previous studies have demonstrated that gaining personal resources, such as PsyCap,

leads to positive employee outcomes, including job satisfaction, work engagement, and

employee performance (Amunkete & Rothmann, 2015; Mazzetti, Guglielmi, Chiesa, &

Mariani, 2016; Tüzün, Çetin, & Basim, 2018). Additionally, gaining resources has been found

to buffer against negative outcomes. For example, PsyCap has been shown to negatively predict

turnover intentions, burnout (Amunkete & Rothmann, 2015; Manzano-García & Ayala, 2017),

stress (Avey, Luthans, & Jensen, 2009; Siu, Cheung, & Lui, 2015), and job stress (Abbas &

Raja, 2015).

Therefore, drawing on the COR theoretical framework, it is argued that gaining PsyCap

as a resource caravan can help develop positive outcomes, including job satisfaction, employee

performance, and work engagement, and buffer against negative outcomes, such as turnover

intentions and burnout. As such, the following research questions are investigated:

Research Question 10: Is PsyCap positively associated with job performance?

Research Question 11: Is PsyCap positively associated with work engagement?

Research Question 12: Is PsyCap positively associated with job satisfaction?

Research Question 13: Is PsyCap negatively associated with turnover intentions?

Research Question 14: Is PsyCap negatively associated with burnout?

Moderators of the relationship between PsyCap and outcome variables

As discussed earlier, previous PsyCap meta-analysis studies (Avey, Reichard, et al.,

2011; Kong et al., 2018; Wu & Nguyen, 2019) have been limited in their examination of

potential moderators in the relationships between PsyCap and antecedent and outcome

variables. Therefore, to build upon previous meta-analyses (e.g., Avey, Reichard, et al., 2011),

this study will examine the role of sample origin, culture, and industry type as potential

10
moderators of the individual relationships between PsyCap and both antecedent and outcome

variables.

Correlations reported across previous studies suggest that industry type may have a

differential bearing on the relationship between leadership and PsyCap. For example,

Schuckert, Kim, Paek, and Lee (2018) found a correlation of 0.706 between authentic

leadership and PsyCap in a sample of employees working in the service industry. In contrast,

Hystad et al. (2014) reported a notably lower correlation of 0.24 between authentic leadership

and PsyCap among a sample of employees within the manufacturing industry. Similar

correlational variations are noted in the relationship between leadership and PsyCap across

studies in relation to sample origin. For example, Corner (2015) reported a correlation of 0.712

between authentic leadership and PsyCap in a sample from the US. In comparison, Yun and

Kang (2018) reported a much lower correlation of 0.27 for the same relationship among a

sample from South Korea.

Recent research also suggests that the moderation effects of sample origin and industry

type on the relationships between PsyCap and outcome variables may be stronger than

previously reported. For example, Idris and Manganaro (2017) reported a correlation between

PsyCap and job satisfaction of .039 in a study with a sample of employees from Saudi Arabia.

In contrast, Kim, Kim, Newman, Ferris, and Perrewé (2019) reported a correlation of .67

between PsyCap and job satisfaction in a study with a sample of employees from the USA.

Similar disparities are reported in regard to the relationship between PsyCap and turnover

intentions when industry type is considered. For example, Munyaka, Boshoff, Pietersen, and

Snelgar (2017) reported a correlation of -.35 between PsyCap and turnover intentions in a

sample of manufacturing employees. However, Kim et al. (2017) reported a correlation of -.72

between PsyCap and turnover intentions in research with a sample of hospitality employees.

11
 These heterogeneous findings suggest that variables such as sample origin and industry

type play an important moderating role in the relationship between PsyCap and individual

antecedent and outcome variables. Therefore, it is important to investigate the potential

moderators for each individual outcome and antecedent variables, rather than as combined

groups of variables, as moderators may influence each relationship. Further, we go beyond the

US/non-US dichotomy, used in Avey, Reichard, et al. (2011) to code studies for each country.

To comprehensively account for contextual differences in different countries, we also

investigate the moderating role of Hofstede’s six-dimensional model (Hofstede et al., 2010) of

national culture. This includes examining the influence cultural dimensions such as power

distance, individualism, masculinity, uncertainty avoidance, long-term orientation, and

indulgence have on the relationships between PsyCap and its outcomes and antecedents. Given

that the moderating effects of sample origin and industry type have not been investigated

previously in relation to individual outcomes and antecedents, we propose the following

research questions:

Research Question 15: Do industry type, sample origin, and cultural dimensions moderate the

relationship between leadership styles (i.e., authentic, ethical, transactional, transformational,

servant empowering, and abusive leadership) and PsyCap?

Research Question 16: Do industry type, sample origin, and cultural dimensions moderate the

relationship between PsyCap and employee outcomes of employee performance, work

engagement, job satisfaction, turnover intentions, and burnout?

Method

Literature search

An initial search was conducted in May 2018, and to ensure that all the studies were

identified and assessed for inclusion in this study, an updated search was conducted in February

12
2021. The aim of the search was to identify relevant published and unpublished studies that

were undertaken between 2007 (when the first empirical study on PsyCap was published) and

the end of 2020. Relevant databases were searched, including Scopus, Web of Science,

PsycINFO, ProQuest, Ovid Medline, CINAHL, and EBSCO (Business Source Ultimate). The

initial search in May 2018 included the search terms of “psychological capital”, PsyCap,

“authentic leadership”, burnout, engagement, performance, “job satisfaction”, turnover

intentions, intention to quit, and intention to leave, as the keywords individually. Then, the

results of the search for “psychological capital” and PsyCap were combined with the operator

“OR”. The same procedure was followed for other variables. The final search strategy then

combined these two results with the operator “AND”.

Inclusion and exclusion criteria

To be included, studies were required to be written in English, and be quantitative and

empirical in nature. They needed to have used either the 12 or 24 item versions of the PsyCap

questionnaire (Avey, Avolio, & Luthans, 2011; Luthans et al., 2007) to measure PsyCap.

Studies that employed single measures for variables (e.g., job satisfaction) were removed as

single-item measures have been shown to have poor validity and reliability (McIver &

Carmines, 1981). Additionally, only studies that investigated PsyCap in relation to workplace

outcomes were included.

An initial database search in May 2018 identified a total of 1260 published studies.

Reviewing abstracts excluded 1106 studies, and the remaining 154 studies (including eight

conference papers) were further analysed for their inclusion in the meta-analysis. To control

for the possibility of publication bias, the literature search was extended to also include

unpublished studies. The first strategy to identify unpublished studies was to search for relevant

theses in the ProQuest database, which yielded 1906 results. Using this method, 26 potential

13
unpublished theses were identified. An email was also sent via the Academy of Management

(AOM) ListServ (Organisational Behavior Division List) to 5380 researchers and academics

requesting relevant unpublished PsyCap studies. From this, two additional studies were

received (one unpublished thesis and one in-press journal article). One further in-press study

was identified via the journal indexing databases. Thus, in total, 29 unpublished studies were

identified for potential inclusion in the meta-analysis.

In the next step, the full papers/theses for the 183 identified studies (154 published

studies and 29 unpublished studies) were assessed against inclusion/exclusion criteria. Three

studies were removed as they did not use the 12 or 24 item measures of PsyCap (Madrid, Diaz,

Leka, Leiva, & Barros, 2017; Nguyen & Nguyen, 2012; Van Steenbergen, van der Ven,

Peeters, & Taris, 2017), and three studies were removed due to using a single-item measure

(Cassidy, McLaughlin, & McDowell, 2014; Cenciotti, Alessandri, & Borgogni, 2017; Hite,

2015). One study was removed as it employed a student rather than a workplace sample (Gooty

et al., 2009), and four studies were removed due to measuring similar but different constructs

from the focus of this study (e.g., measuring career satisfaction) (Ganotice, Yeung, Beguina,

& Villarosa, 2016; Guo, Xiyuan, & Qin, 2012; Polatci & Akdogan, 2014; Zhang, Li, Ma, Hu,

& Jiang, 2014). One further study was also removed due to inconsistent reporting of the

correlation coefficients for the relationship between PsyCap and burnout (Malekitabar, Riahi,

& Malekitabar, 2017). Finally, 16 theses were removed due to not having access to the full-

text and a further 12 studies were excluded as these studies did not provide the results of the

correlation coefficients, despite follow up email requests to the corresponding authors of these

studies. Therefore, the total number of included studies in the meta-analysis from this phase of

the literature search was 143 studies (151 samples, N = 56608).

A similar procedure was used in February 2021 to identify relevant literature published

between May 2018 until the end of December 2020 regarding the abovementioned variables

14
(i.e., authentic leadership, job satisfaction, job performance, work engagement, burnout, and

turnover intentions). To expand the scope regarding the antecedents of PsyCap, the search

terms of “psychological capital”, PsyCap, “ethical leadership”, “transactional leadership”,

“transformational leadership”, “servant leadership”, “abusive leadership”, “abusive

supervision” were also utilised. The aim was to identify studies (published from 2007 until the

end of December 2020) that have investigated the relationship between PsyCap and leadership

styles beyond only authentic leadership. The same procedure for study inclusion was used as

described earlier. The search for published studies in February 2021 yielded 1600 published

studies. An additional 183 studies were identified as potential studies to be included in the

meta-analysis. After a closer examination, further studies were removed due to using a measure

other than 12-items or 24-items PCQ to measure PsyCap (45 studies), using other levels of

analysis beyond individual level (e.g., team PsyCap) (4 studies), not using an employee sample

(6 studies), and not providing correlation for the identified relationships (30 studies). Therefore,

98 usable published studies were identified to be included beyond those identified in the first

search.

In terms of the unpublished studies, the results of the search in ProQuest yielded 9635

results, of which 30 potential theses were identified. However, all 30 theses were removed due

to not having access to them (29 theses) and not providing the correlation coefficient for the

identified studies (1 thesis). Only one new thesis was added to the database at this stage, which

was the unpublished Ph.D. thesis of one of the authors of this paper (the thesis has not been

indexed in the ProQuest yet). Furthermore, seven potential conference papers were identified

via searching in the journal indexing databases. Among these, five studies were removed due

to using a measure other than 12-items or 24-items PCQ to measure PsyCap (2 studies) and not

providing correlation for the identified relationships (3 studies). Therefore, a total of two

15
conference papers were added to the analysis. Overall, in the second stage of literature research,

101 new studies (103 independent samples, N=39908) were added to the database.

Therefore, as shown in Figure 1, the total number of published and unpublished studies

included in the meta-analysis was 244 studies (254 independent samples, N = 96416). A

comparison of the scope of this meta-analysis compared with previous PsyCap meta-analyses

(Avey, Reichard, et al., 2011; Kong et al., 2018; Wu & Nguyen, 2019) has been provided in

the online supplementary material document, which demonstrates the comprehensiveness of

the meta-analysis reported in this paper compared to the extant PsyCap meta-analyses.

INSERT FIGURE 1 ABOUT HERE

Coding procedure

For each sample, coded sample size, sample origin, industry type, effect size, and

reliability scales (Cronbach’s Alpha) were extracted for the variables of interest. In this study,

the correlation coefficient was used as a measure of effect size since this study is interested in

the relationships between PsyCap and other variables.

Studies were also coded according to whether they had reported the correlation and

reliability scales for PsyCap as an aggregate or had reported each of the four dimensions of

PsyCap. In this case, when the correlation was reported individually for the different

dimensions of PsyCap, the composite correlation was calculated according to methods outlined

in Schmidt and Hunter (2015). Additionally, when a study reported separate reliability scores

(Cronbach’s Alpha) for the different dimensions of a construct, the average score was

calculated. If a study did not report exact reliability scale data (Cronbach’s Alpha), nor provide

this information on request, the average score of the reliability scale from the available studies

in the dataset was calculated and entered (see e.g., Kirca, Jayachandran, & Bearden, 2005;

Mackay, Allen, & Landis, 2017). Coding of the studies was conducted in June and August

16
2018, as well as February 2021 to ensure coding reliability. Any inconsistency in coding was

discussed and resolved within the research team.

Meta-analysis procedure

This meta-analysis combined the methodologies outlined in Hunter and Schmidt (2004)

and Borenstein et al. (2009), which acknowledge artefact corrections and provide similar

formulas to correct the effect sizes. As such, this meta-analysis corrected the effect sizes with

regard to the internal reliability (Cronbach’s alpha) for all variables. Moreover, effect sizes

were weighted by sample size within random-effects models. The summary effect sizes were

then calculated using the Comprehensive Meta-Analysis software program (CMA) version 3.

Tests of heterogeneity were used to assess the degree of variability or inconsistency in

study outcome magnitudes. Heterogeneity tests are important to assess the extent to which

observed dispersion among the findings of different studies is real dispersion, rather than

dispersion arising from within-study error (Borenstein et al., 2009). Three measures were used

to assess the heterogeneity: Cochran’s Q statistic, I2, and Tau2. Where significant/substantial

heterogeneity is identified, either through a significant Q, an I2-value > 75%, or large Tau2, it

is suggested that potential moderators of effect size be investigated (Borenstein et al., 2009;

Higgins, Thompson, Deeks, & Altman, 2003). In this meta-analysis study, when heterogeneity

tests were significant, the potential moderation effects of sample origin and industry type were

investigated using the mixed-effects analysis option in the CMA for subgroup analyses (see

Borenstein et al., 2009) as these moderators are categorical variables. Moreover, the

moderating role of culture was investigated using meta-regression.

For moderation analyses, sample origin and industry type were coded using categories;

names of each country, from which the samples were obtained, were coded for sample origin;

and service and manufacturing for industry type. In addition, for the cultural dimensions,

17
numerical values for each dimension and each country were obtained from open access data

via Hofstede Insight website (https://www.hofstede-insights.com/).

In the next step, sensitivity analysis was conducted via outlier analysis (Borenstein et

al., 2009). Studies located in the 5% percent of both ends of the spectrum (regarding the

strength of the effect sizes) were omitted, and then the result of the new analysis was compared

with the original analysis (see Huber, 1980; Tukey, 1960). The main reason for conducting

outlier analysis is that extreme effect sizes can affect the summary effect sizes, which can then

influence the robustness and validity of the meta-analysis results (Viechtbauer & Cheung,

2010).

Finally, publication bias was tested using funnel plots in conjunction with Duval and

Tweedie’s Trim and Fill method (Duval & Tweedie, 2000). The first step in the Trim and Fill

method is to identify and remove (or trim) the studies that have caused funnel plot asymmetry.

Then, the true centre of the funnel plot is estimated. At the next step, the removed studies are

replaced and their missing counterparts around the centre are added (or filled). In the last step,

the number of missing studies that may be a result of publication bias is estimated and a new

summary effect size is calculated with the filled studies included in the meta-analysis (Duval

& Tweedie, 2000). By conducting the publication bias analysis using this method, a form of

sensitivity analysis is also implemented (Duval, 2005).

Analysis and results

The results yielded by the meta-analysis in relation to the proposed research questions

are described below. As such, Tables 1 and 2 depict the results of the random-effects model

meta-analysis, heterogeneity test, and some components of the moderation analyses for the

identified relationships (Due to the high volume of analyses and outputs, the complete

moderation analysis results can be found in the supplementary material document). Forest plots

18
of the investigated relationships, funnel plots of the publication bias analyses have also been

provided in the supplementary material document.

INSERT TABLE 1 ABOUT HERE

INSERT TABLE 2 ABOUT HERE

Antecedents of PsyCap
According to Table 1, the summary correlations for leadership styles (i.e., authentic,

ethical, transformational, transactional, servant, empowering, and abusive) and employee

PsyCap were found to be significant and positive (for positive leaderships styles) and negative

for abusive leadership. These results provided meaningful insight for RQs 1-9. Results of the

subgroup analyses confirmed that the differences in summary correlations for all leadership

styles and PsyCap are statistically significant; Q(6)=133.33, p<.001. Accordingly, it was found

that empowering and transformational leadership styles have the strongest correlation with

PsyCap, while the correlations between transactional and abusive leadership and PsyCap were

the smallest in magnitude. Furthermore, the results of the meta-regression analyses suggested

that the differences between empowering leadership and abusive/transactional leadership, as

well as transformational leadership and abusive/ transactional leadership remained statistically

significant (Table 3 and 4). This further supports the role of empowering and transformational

leadership as the strongest antecedents of PsyCap and abusive and transactional leadership as

the weakest antecedents of PsyCap.

INSERT TABLE 3 ABOUT HERE

INSERT TABLE 4 ABOUT HERE

The significant Cochran’s Q and high I2 showed evidence of significant heterogeneity

in the relationship between the leadership styles (except for transactional leadership) and

PsyCap, and therefore, moderation analyses were conducted (RQ15). The results of the

19
subgroup analyses did not show a significant moderation effect for industry type. Moderation

analysis was not conducted for the effect of industry type on the relationships between ethical,

servant, and empowering leadership and PsyCap, due to insufficient study numbers. However,

the results of the subgroup analyses showed a significant moderation effect of sample origin in

the relationships between PsyCap and the leadership styles, except for servant leadership

(Authentic Leadership, Q(14)=210.38, p<.001; Ethical leadership, Q(5)=11.85, p=.037;

Transformational Leadership, Q(5)=161.89, p<.001; Empowering Leadership, Q(2)=14.80,

p=.001; and Abusive Leadership, Q(2)=26.51, p<.001). For example, for the relationship

between authentic leadership and PsyCap, the summary effect sizes for studies conducted in

Iran and Portugal were the strongest, while for Canada, Namibia, and Norway were the

weakest. For the relationship between ethical leadership and PsyCap, the summary effect sizes

for studies in Jordan and Pakistan were the largest, and for the USA was the lowest. For the

relationship between transformational leadership and PsyCap, the strongest effect size was

from South Korea, and the smallest summary effect size was from the Chinese samples. For

the relationship between empowering leadership and PsyCap, the effect size from the US

sample was the strongest, and for the sample from China was the smallest. Finally, for the

relationship between abusive leadership and PsyCap, the summary effect sizes from samples

in Pakistan were the largest, and those from India were the lowest. The results clearly show the

diversity of the effect sizes among countries (Detailed information can be found in the

supplementary materials).

In terms of the cultural analysis, using meta-regression, power distance and masculinity

were found to moderate the relationship between transformational leadership and PsyCap

(power distance: R2=.42, p=.015; masculinity: R2=.41, p=.009). Overall, the analyses showed

that as power distance increases, the strength of the positive relationship between

transformational leadership and PsyCap decreases. Similarly, the analyses showed that, overall,

20
as masculinity increases, the strength of the positive relationship between transformational

leadership and PsyCap decreases. Additionally, long-term orientation demonstrated to

moderate the relationship between servant leadership and PsyCap (R2=.46, p=.014). The results

suggested that as long-term orientation increases, the positive relationship between servant

leadership and PsyCap becomes stronger. Finally, the results indicated that uncertainty

avoidance moderates the relationship between abusive leadership and PsyCap (R2=.64,

p=.002). The results suggested that as uncertainty avoidance increases, the strength of the

negative relationship between abusive leadership and PsyCap becomes stronger. No

moderation effect was found for either individualism or indulgence in the relationship between

PsyCap and leadership styles (Further information about the moderation analysis can be found

in the supplementary material document).

Outlier analysis. Using the 5% rule, the lowest and highest effect sizes were removed

for each of the identified relationships (see Huber, 1980; Tukey, 1960). The adjusted summary

effect sizes (Authentic Leadership, .487, 95% CI [.442, .528]; Ethical Leadership, .405, 95%

CI [.347, .461]; Transformational Leadership, .520, 95% CI [.371, .642]; Servant

Leadership, .496, 95% CI [.465, .526]; Abusive Leadership, -.349, 95% CI [-.507, -.169]), were

not meaningfully different to the original values suggesting outliers are not strongly impacting

the result of the meta-analyses. Outlier analysis was not conducted for the relationships

between transactional and empowering leadership and PsyCap as there were only three studies

for each of the relationships.

Publication bias. As seen in Table 1, potential unpublished studies were only suggested

for two of the relationships (i.e., authentic and servant leadership and PsyCap). However, for

authentic leadership, the summary effect size estimate following Duval and Tweedie’s trim-

and-fill method was negligibly higher (+.019) than the initial summary effect sizes, with only

three potential unpublished studies suggested. Similarly, for servant leadership, the difference

21
was negligible (-.025), with only one potential unpublished study suggested. Therefore,

publication bias does not seem to be substantially impacting the results.

Outcomes of PsyCap
The summary effect sizes for the relationship between PsyCap and outcome variables

of (self-reported and supervisor-reported) job performance, work engagement, job satisfaction,

turnover intentions, and burnout have been depicted in Table 2, which provide meaningful

insight for RQs 10-14. The results show that PsyCap has a significant positive relationship with

job performance, work engagement, and job satisfaction and a significant negative relationship

with turnover intentions and burnout. In addition, the magnitudes of all the identified

relationships were above medium in strength (see Cohen, 1977), with the relationship between

PsyCap and work engagement to be found the strongest. In terms of the heterogeneity, the

significant Cochran’s Q and high I2 in all relationships suggested considerable heterogeneity;

thus, moderation analyses were conducted (RQ16). The results of the subgroup analysis

suggested that sample origin significantly moderated all the relationships (Self-reported

performance, Q(12)=236.96, p<.001; supervisor-reported Performance, Q(11)=171.94,

p<.001; Work Engagement, Q(21)=266.50, p<.001; Job Satisfaction, Q(21)=3559.41, p<.001;

Turnover Intentions, Q(13)=111.03, p<.001; and Burnout, Q(13)=224.96, p<.001). for

example, for the relationship between PsyCap and Self-reported performance, the largest effect

size was from a Taiwanese sample, and the smallest summary effect size was from Romania.

Similarly, for the relationship between PsyCap and Self-reported performance, the largest

effect size was from Taiwan, but the smallest effect size was from China. For PsyCap – work

engagement relationship, the summary effect size from Indian samples was the strongest, while

the summary effect size from North Cyprus samples was the weakest. For PsyCap – job

satisfaction relationship, the summary effect size for a sample from Ghana was the strongest,

and the summary effect size for a sample from Saudi Arabia was the weakest. For the

22
relationship between PsyCap and turnover intentions, the largest summary effect size was from

Romanian samples, and the smallest effect size was from North Cyprus. Finally, for the

relationship between PsyCap and burnout, the summary effect size from the US samples was

the strongest, and the effect size for a sample from Iran was the weakest (Further information

about the moderation analysis can be found in the supplementary material document).

Moreover, industry type was found to significantly moderate the relationship between

PsyCap and self-reported performance, Q(1)=8.09, p=.004, with the relationship stronger in

the manufacturing industry sample than the service industry. Besides, results of the meta-

regression analyses for cultural analysis demonstrated that masculinity moderates the positive

relationship between PsyCap and self-reported job performance (R2=.05, p=.048) and the

negative relationship between PsyCap and turnover intentions (R2=.10, p=.045). The results

suggested that as the level of masculinity increases, the strength of the positive relationship

between PsyCap and self-reported performance decreases. It was also found that as the level

of masculinity increases, the strength of the negative relationship between PsyCap and turnover

intentions decreases. In addition, uncertainty avoidance was found to moderate the negative

relationship between PsyCap and turnover intentions (R2=.15, p=.012), in which as the level of

uncertainty avoidance increases, the negative relationship becomes stronger. Finally, for the

relationship between PsyCap and job satisfaction, although none of the six cultural dimensions

were identified to be a significant moderator, the total regression model (comprising the six

factors) was statistically significant (p=.037) with R2=.21.

Outlier analysis. For each identified relationship, 5% of effect sizes from both ends

were removed and the adjusted summary effect sizes were calculated. The results showed that

the adjusted summary effect sizes were not meaningfully different from the original summary

effect sizes (Self-reported Performance, .564, 95% CI [.509, .614]; Supervisor-reported

Performance, .389, 95% CI [.314, .458]; Work Engagement, .682, 95% CI [.655, .708]; Job

23
Satisfaction, .631, 95% CI [.596, .664]; Turnover Intentions, -.357, 95% CI [-.402, -.310]; and

Burnout, -.508, 95% CI [-.557, -.456]). Thus, outliers do not appear to be meaningfully

affecting the effect sizes.

Publication bias. Using Duval and Tweedie’s trim-and-fill method (Table 2), no

potential unpublished studies were suggested for the relationships between PsyCap, turnover

intentions, and burnout. Moreover, for the relationship between PsyCap and self-reported job

performance, only three unpublished studies were suggested, and the adjusted summary effect

size was found to be negligibly higher than the original value (+.032). In addition, Outliers

might be responsible for the difference in the calculated effect sizes in the relationship between

PsyCap, supervisor-reported job performance, work engagement, and job satisfaction. After

removing the outliers from the analysis as identified in the outlier analysis, using the Duval

and Tweedie’s trim-and-fill method, the summary effect size estimates were only negligibly

different from the original summary effect sizes (-.004 for supervisor-reported performance,

-.015 for work engagement, and -.026 for job satisfaction) with few potential unpublished

studies suggested (3 suggested unpublished studies for supervisor-reported performance, 6

suggested unpublished studies for work engagement, and 6 suggested unpublished studies for

job satisfaction). Therefore, publication bias does not seem to substantially impact the results.

The potential role of study design

As “cross-sectional studies of attitude-behavior relationships are vulnerable to the

inflation of correlations by common method variance” (Lindell & Whitney, 2001, p. 114), it is

important to investigate the probability that the nature of study designs has inflated effect size

results, especially as most of the studies included in this meta-analyses had cross-sectional

designs. However, no PsyCap meta-analysis study has investigated it to date. As such,

moderation analysis was conducted to compare whether there are any differences between the

effect sizes reported in the cross-sectional studies compared to longitudinal studies (Table 5).

24
Due to the insufficient number of longitudinal studies, we were only able to conduct the

subgroup analysis for the relationships between PsyCap and the employee outcomes of work

engagement, job satisfaction, and supervisor-reported job performance. The results of the

moderation analysis were only significant for the relationship between PsyCap and work

engagement, showing that the association was stronger among cross-sectional studies

compared to longitudinal studies, Q(1)=10.43, p=.001. However, both effects were still large,

and thus while it is plausible that study design has inflated correlations, this does not clearly

affect the interpretation of the effect. The results of the moderation analyses were not

significant for Supervisor-reported Performance, Q(1)=.80, p=.371 or Job Satisfaction,

Q(1)=1.25, p=.264).

INSERT TABLE 5 ABOUT HERE

Discussion

This study undertook a comprehensive and methodologically rigorous meta-analysis to

generate a more in-depth understanding of the relationship between PsyCap, and its outcomes

and antecedents, as well as the moderators of these relationships. Drawing on 254 samples

from different countries and cultural contexts derived from multiple indexing databases, this

study represents the largest PsyCap meta-analysis dataset conducted to date as it includes more

than double the samples included in previous PsyCap meta-analyses (Avey, Reichard, et al.,

2011; Kong et al., 2018; Wu & Nguyen, 2019). Table 6 provides a snapshot of the comparison

between the current study and the previous PsyCap meta-analyses. It clearly demonstrates how

our study has gone above and beyond the previous ones. This adds more weight to the findings

of our study and provides much stronger support for the results of the investigated relationships.

In addition, as the table shows the overlap between these meta-analyses is limited. In particular,

only the relationship between PsyCap and job satisfaction has been investigated across all

25
PsyCap meta-analysis studies. Having said that, the results of our study provided much stronger

support for this relationship (i.e., r = .683) compared to .54 (Avey et al., 2011), .533 (Kong et

al., 2018), and .511 (Wu & Nguyen, 2019), which are quite similar in magnitude.

INSERT TABLE 6 ABOUT HERE

Overall, the findings showed that the correlations between PsyCap and antecedent

(e.g., transformational and empowering leadership) and outcome (e.g., job satisfaction and

work engagement) variables were significant and of strong correlational magnitude (i.e., r >

|.50| for a strong relationship). Thus, PsyCap is an important psychological resource as it

appears to foster desirable employee outcomes (e.g., job performance, work engagement, and

job satisfaction) and buffer against undesirable employee outcomes (e.g., turnover intentions

and burnout). In addition, empowering and transformational leadership were found to have the

strongest link with PsyCap, while transactional and abusive leadership were identified as

having the weakest association. These findings offer further evidence that leadership styles

(and particularly positive leadership styles) are important antecedents of PsyCap. To explore

heterogeneity and to investigate the moderation effects of sample origin and industry type for

each of the relationships, a series of subgroup analyses were conducted. The results from these

analyses showed that the correlation between PsyCap and employee self-reported performance

is higher in samples of employees working in the manufacturing industry than those working

in the service industry. Furthermore, we found that sample origin moderated the relationship

between PsyCap and all variables, except servant and transactional leadership. These findings

significantly expand previous PsyCap meta-analyses (Avey, Reichard, et al., 2011; Kong et al.,

2018; Wu & Nguyen, 2019), which did not investigate the moderating role of sample origin

and industry type in the relationship between PsyCap and each of the individual antecedent and

outcome variables. As discussed previously, to move beyond a simple US/non-US

26
categorisation and to provide a better understanding of the interplay between culture and

PsyCap, we implemented meta-regression to analyse the influence of cultural dimensions on

PsyCap relationships. Our findings provided a substantial addition to the literature and

demonstrated that the cultural dimensions of power distance, masculinity, long-term

orientation, and uncertainty avoidance might be the cultural dimensions that can impact the

relationship between PsyCap and antecedent and outcome variables. The reasons for this are

as yet unconsidered in the research literature, and further substantive research is needed.

However, at current, we suggest that PsyCap and its relationships need to be interpreted

cautiously and with consideration of the context of sample origin and cultural dimensions.

This study has provided further insights into the nomological network of PsyCap by

extending beyond the previous PsyCap meta-analysis studies (Avey, Reichard, et al., 2011;

Kong et al., 2018; Wu & Nguyen, 2019) in several ways. First, it included four leadership styles

that have been overlooked in previous PsyCap meta-analyses (i.e., empowering,

transformational, transactional, and servant leadership). It also investigated and compared the

strengths of leadership styles in predicting PsyCap. Similarly, two additional relationships,

which have not been investigated in previous PsyCap meta-analyses, were analysed; the

relationship between PsyCap and both work engagement and burnout. Second, a larger and

more diverse database was utilised in this study compared to previous PsyCap meta-analyses.

This resulted in a significantly larger total sample size than those used in previous studies,

including far more non-US samples (209 non-US samples in this meta-analysis, compared to

seven in Avey, Reichard, et al., 2011). This has enabled the examination of the moderating

influence of sample origin in the relationship between PsyCap and each of the variables.

Furthermore, the comprehensive moderation analysis of sample origin was conducted by

coding each country and their related cultural profile, which was then accompanied by

subgroup and meta-regression analyses. This level of analysis has not been conducted in any

27
PsyCap meta-analysis to date. Consequently, the results of this meta-analysis have produced

a finer-grained understanding regarding the interplay between PsyCap and cultural context on

work-related outcomes and thereby responded to calls to extend research regarding the cross-

cultural validity of PsyCap (see Luthans & Youssef-Morgan, 2017).

Finally, this meta-analysis has employed more robust and comprehensive

methodologies beyond those used in previous PsyCap meta-analyses. These include

comprehensive search techniques using numerous indexing databases, which led to a larger

database; correcting for artefacts; and conducting moderation, outlier, and publication bias

analyses. Using these methodologies has helped garner results and effect sizes that provide a

more accurate insight into the true nature of the investigated relationships, compared to those

generated in previous PsyCap meta-analyses (Avey, Reichard, et al., 2011; Kong et al., 2018;

Wu & Nguyen, 2019).

Theoretical and practical implications

From a theoretical perspective, this study has responded to calls in the literature for a

more in-depth understanding of the antecedents of PsyCap (see Luthans & Youssef-Morgan,

2017; Newman et al., 2014) by investigating the relationship between authentic, ethical,

transformational, transactional, servant, empowering, and abusive leadership and PsyCap. It

has been argued that there is now ample evidence affirming PsyCap as a positive construct with

useful predictive validity and that what is needed now is a greater understanding of the systems

and structures within persons and organisational life that predict PsyCap itself (Avey, 2014).

This study has also extended findings from previous PsyCap meta-analyses (Kong et al., 2018;

Wu & Nguyen, 2019) by demonstrating that authentic, ethical, and abusive leadership styles

are significant predictors of employee PsyCap. Furthermore, by investigating leadership styles

that have not been included in previous meta-analyses (i.e., empowering, transformational,

transactional, and servant leadership), this study has provided a deeper understanding regarding

28
the relationship between various leadership styles and employee PsyCap. Importantly, this

meta-analysis compared the strength of various leadership styles in predicting PsyCap to

provide insights into which leadership styles are the strongest antecedents of PsyCap.

In relation to outcomes of PsyCap, this study has confirmed findings reported in

previous meta-analyses (Avey, Reichard, et al., 2011; Kong et al., 2018; Wu & Nguyen, 2019)

by demonstrating positive relationships with job performance and job satisfaction and a

negative association with turnover intentions. However, as the current study has undertaken a

more robust meta-analysis methodology and included a much larger number of studies than

previous meta-analyses, it is argued these findings can be interpreted with greater confidence.

Furthermore, this study has provided new meta-analytical insights into the relationships

between PsyCap and burnout and work engagement. Specifically, it was found that PsyCap has

a strong, positive relationship with work engagement and a strong, negative relationship with

job burnout.

This study also represents the first PsyCap meta-analysis to investigate moderators of

the relationship between PsyCap and undesirable employee outcomes, as well as the

relationship between PsyCap and its antecedents. By examining the moderating effects of

sample origin and industry type, this study has responded to several recent calls for research

investigating moderators of PsyCap relationships (e.g., Luthans & Youssef-Morgan, 2017;

Luthans et al., 2015; Newman et al., 2014). Identifying moderators is argued to be the “key

issue for theory development and testing as well as practical applications of a theory” (Aguinis,

Gottfredson, & Wright, 2011, p. 1033). Thus, identifying significant moderators of PsyCap

relationships is important for developing a greater understanding of the conditions which

influence the effects of PsyCap. In particular, the moderation analyses in this study highlight

the importance of sample origin and culture in relation to PsyCap and suggest that PsyCap may

operate differently across various cultural settings. For example, PsyCap might have a stronger

29
impact on employee performance in countries with lower levels of masculinity (e.g., Portugal

and South Korea). As the role of culture has not been investigated to this extent in previous

PsyCap meta-analyses, this study makes an important contribution in responding to calls for

greater investigation into potential cultural differences in relation to PsyCap (Luthans &

Youssef-Morgan, 2017).

In addition, this was the first PsyCap meta-analysis to investigate the moderating role

of study design on the reported findings. The results showed that study design impacted the

strength of the relationships, whereby the relationship between PsyCap and work engagement

was found to be stronger among cross-sectional studies compared to longitudinal studies. It is

likely that this has occurred because the analysed cross-sectional studies might incur common

method variance (Lindell & Whitney, 2001), though it needs to be indicated that the

correlations for both longitudinal and cross-sectional studies were large, and thus the

interpretation of the results is similar. That said, this finding demonstrates the importance of

conducting more longitudinal studies in future research.

This study also makes several important practical contributions. For example, the

findings of this meta-analysis affirm PsyCap as a useful positive construct that has a significant

impact on employee outcomes that are important for organisations, and employees alike,

including job satisfaction, job performance, turnover intentions, and burnout. This, in turn, can

inform organisational practices in terms of fostering resources such as PsyCap, via training

interventions (e.g., PCI; Luthans, Avey, Avolio, Norman, & Combs, 2006), to not only enhance

individual employee PsyCap, but also foster other desirable employee outcomes (e.g., job

performance and job satisfaction) and hinder a wide range of undesirable employee outcomes

(e.g., turnover intentions and burnout).

30
 Our findings provide robust evidence for significant associations between positive

leadership styles (e.g., ethical and servant leadership) and PsyCap, as well as the negative

relationship between abusive leadership and PsyCap. These findings suggest that organisations

could enhance and sustain employee PsyCap by investing in positive leadership development

and training. For example, previous research has found that authentic leadership can be

developed among leaders through training and coaching programs (Baron, 2016; Fusco, 2018).

Our findings also indicate empowering leadership and transformational leadership have the

strongest associations with employee PsyCap. Thus, it is suggested that organisations looking

to maximise the benefits of employee PsyCap focus on attracting and cultivating these

leadership styles within their leaders.

The findings from this study also revealed that to understand the relationship between

PsyCap and antecedent and outcome variables, sample origin, and cultural contexts need to be

taken into account. In regard to antecedents of PsyCap, the results suggest that transformational

leadership might have a stronger influence on employee PsyCap in countries with lower levels

of power distance (e.g., Pakistan) and masculinity (e.g., South Korea). In contrast, the impact

of servant leadership on employee PsyCap might be stronger in the countries with higher levels

of long-term orientation (e.g., Taiwan), while abusive leadership might have a stronger

negative impact on PsyCap in countries with higher levels of uncertainty avoidance (e.g.,

Pakistan).

In regard to outcome variables, our findings demonstrated that PsyCap might be a

stronger predictor of employee self-reported performance in countries with lower levels of

masculinity (e.g., Portugal). Moreover, it was found that the strength of the relationship

between PsyCap and turnover intentions might be lower in countries with higher levels of

masculinity (e.g., USA and Australia). However, the strength of this relationship was found to

be stronger in the countries with higher levels of uncertainty avoidance (e.g., Romania and

31
Turkey). Finally, the relationship between PsyCap and job satisfaction was found to be more

complex than the other identified relationships. The findings suggested that this relationship

may be impacted by a multifactorial combination of variable as indicated by a total regression

model comprising the six factors being statistically significant. Overall, these represent new

findings, as this is the first comprehensive meta-analytical investigation of culture and country

as moderators of the relationship between PsyCap and antecedent and outcome variables. These

findings provide contextual understanding about which (and how) cultural dimensions can

impact the relationship between PsyCap and antecedent/outcome variables. These insights help

identify in which countries PsyCap might have a stronger impact on the investigated employee

outcomes. They also uncover under which circumstances and for which countries, a specific

leadership style might be more effective in fostering employee PsyCap. Overall, the results of

this study demonstrate the cultural differences in relation to PsyCap and PsyCap relationships.

In addition, the results suggested that PsyCap may operate differently across various

industry settings. For example, the results showed that the positive relationship between

PsyCap and performance is stronger in the manufacturing industry in comparison to the service

industry. This finding suggests that PsyCap development (e.g., PsyCap Intervention; Luthans

et al., 2006) may be particularly important in industries such as manufacturing. However, this

needs to be interpreted with some caution, given that there were few studies with manufacturing

samples included in the analysis.

Limitations and future research directions

As with any research, this study has some limitations which should be acknowledged.

First, in this study, only PsyCap studies conducted at the individual level of analysis were

included in the analysis. However, it is important to recognise that in recent years, research has

also begun to explore PsyCap at the team level (e.g., Dawkins, Martin, Scott, Sanderson, &

32
Schüz, 2018; Heled, Somech, & Waters, 2016). Accordingly, there remains an opportunity for

future research to commence investigation across this emerging body of collective PsyCap

(e.g., via meta-analysis or systematic review). This avenue of research would enable the

development of a more in-depth understanding of the factors that influence collective PsyCap

and its relationship with its outcome and antecedent variables.

Moreover, the majority of the included studies had cross-sectional survey-based study

designs. Therefore, the causality of the identified relationships cannot be inferred (Van der

Stede, 2014). As such, conducting future longitudinal studies is necessary to be able to

investigate causality between PsyCap and theorised antecedent and outcome variables. This

would also enable future PsyCap meta-analyses to examine the causation and directionality of

the identified relationships across PsyCap studies with greater certainty. Furthermore, results

of the moderation analysis suggested that cross-sectional studies might yield inflated effect

sizes, which suggests the need for more reliable results using longitudinal study designs.

Besides, it is important to acknowledge that this meta-analysis has only included

studies that have been written in the English language. However, with the emergence of PsyCap

research in many more countries, the PCQ-12 and -24 (Avey, Avolio, et al., 2011; Luthans et

al., 2007) have been translated and validated in other languages, including Spanish, Portuguese,

and Italian (Alessandri, Borgogni, Consiglio, & Mitidieri, 2015; Antunes, Caetano, & Cunha,

2017; León-Pérez, Antino, & León-Rubio, 2017). Therefore, there is an opportunity for future

meta-analyses to include PsyCap studies published in languages other than English so to reflect

a more global understanding of PsyCap.

Lastly, in considering the relationship between negative leadership styles and PsyCap,

it is conceded that recent reviews and meta-analyses of PsyCap (Newman et al., 2014; Nolzen,

2018; Wu & Nguyen, 2019), including the current study, have only focused on abusive

33
leadership. This suggests a paucity of knowledge regarding how other negative leadership

styles (e.g., exploitative, toxic, despotic, and narcissistic leadership), beyond abusive

leadership, influence employee PsyCap, and it is clear that further original empirical research

in this area is required. This is important as the findings of previous studies suggest that

negative leadership styles predict different employee psychological outcomes. For example,

exploitative, despotic, and laissez-faire leadership styles have been found to lead to higher

levels of psychological distress, and lower levels of psychological well-being, and affective

organisational commitment, respectively (Buch, Martinsen, & Kuvaas, 2015; Majeed &

Fatima, 2020; Raja, Haq, De Clercq, & Azeem, 2020). These findings highlight the importance

of investigating different forms of negative leadership styles, beyond only abusive leadership,

as potential unique antecedent variables of PsyCap.

Conclusion

PsyCap research has grown rapidly over the past 15 years, with much of this research

focused on the relationship between PsyCap and work-related outcomes. This paper reported

findings of a comprehensive meta-analytic study of PsyCap that has extended beyond previous

PsyCap meta-analyses, in terms of both methodology and scope. The meta-analysis study

aimed to investigate significant antecedents and outcomes related to PsyCap, along with the

factors that moderate these relationships. The findings of this study demonstrated strong

evidence for the importance of employee PsyCap in positively predicting desirable employee

outcomes (job performance, work engagement, job satisfaction) and overcoming undesirable

employee outcomes (turnover intentions and burnout). The findings also showed the

importance of positive leadership styles (i.e., ethical, empowering, authentic, transformational,

and servant leadership) in enhancing employee PsyCap with the empowering and

transformational leadership to have the strongest relationship with PsyCap. Furthermore,

34
 moderation analyses provided meaningful insight for the impact of sample origin, culture, and

industry type on the relationship between PsyCap and both antecedent and outcome variables.

In addition, the findings demonstrated that study design significantly moderated the

relationship between PsyCap and work engagement. Overall, the results of this meta-analysis

have extended finding from previous PsyCap meta-analyses and provide a greater

understanding of the factors that influence PsyCap and its relationships with leadership styles

and work-related outcomes.

References 2

Abbas, M., & Raja, U. (2015). Impact of psychological capital on innovative performance and job

stress. Canadian Journal of Administrative Sciences / Revue Canadienne des Sciences de

l'Administration, 32(2), 128-138.

Adams, B. G., Meyers, M. C., & Sekaja, L. (2020). Positive leadership: relationships with employee

inclusion, discrimination, and well-being. Applied Psychology, 69(4), 1145-1173.

Agarwal, U. A. (2019). Examining links between abusive supervision, PsyCap, LMX and outcomes.

Management Decision, 57(5), 1304-1334.

Agarwal, U. A., & Avey, J. B. (2020). Abusive supervisors and employees who cyberloaf. Internet

Research, 30(3), 789-809.

Aguinis, H., Gottfredson, R. K., & Wright, T. A. (2011). Best-practice recommendations for estimating

interaction effects using meta-analysis. Journal of Organizational Behavior, 32(8), 1033-1043.

Ahmad, J., Athar, M. R., Azam, R. I., Hamstra, M. R. W., & Hanif, M. (2018). A resource perspective

on abusive supervision and extra-role behaviors: the role of subordinates’ psychological

capital. Journal of Leadership & Organizational Studies, 26(1), 73-86.

2
The studies that have been used in this meta-analysis have been included in the Online Supplementary
Materials document.

35
Alessandri, G., Borgogni, L., Consiglio, C., & Mitidieri, G. (2015). Psychometric properties of the

Italian version of the psychological capital questionnaire. International Journal of Selection

and Assessment, 23(2), 149-159.

Alessandri, G., Consiglio, C., Luthans, F., & Borgogni, L. (2018). Testing a dynamic model of the

impact of psychological capital on work engagement and job performance. Career

Development International, 23(1), 33-47.

Amunkete, S., & Rothmann, S. (2015). Authentic leadership, psychological capital, job satisfaction

and intention to leave in state-owned enterprises. Journal of Psychology in Africa, 25(4), 271-

281.

Antunes, A. C., Caetano, A., & Cunha, M. P. e. (2017). Reliability and construct validity of the

Portuguese version of the psychological capital questionnaire. Psychological Reports, 120(3),

520-536.

Avey, J. B. (2014). The left side of psychological capital: new evidence on the antecedents of PsyCap.

Journal of Leadership & Organizational Studies, 21(2), 141-149.

Avey, J. B., Avolio, B. J., & Luthans, F. (2011). Experimentally analyzing the impact of leader

positivity on follower positivity and performance. The Leadership Quarterly, 22(2), 282-294.

Avey, J. B., Luthans, F., & Jensen, S. M. (2009). Psychological capital: a positive resource for

combating employee stress and turnover. Human Resource Management, 48(5), 677-693.

Avey, J. B., Reichard, R. J., Luthans, F., & Mhatre, K. H. (2011). Meta-analysis of the impact of

positive psychological capital on employee attitudes, behaviors, and performance. Human

Resource Development Quarterly, 22(2), 127-152.

Banks, G. C., McCauley, K. D., Gardner, W. L., & Guler, C. E. (2016). A meta-analytic review of

authentic and transformational leadership: A test for redundancy. The Leadership Quarterly,

27(4), 634-652.

36
Baron, L. (2016). Authentic leadership and mindfulness development through action learning. Journal

of Managerial Psychology, 31(1), 296-311.

Borenstein, M., Hedges, L. V., Higgins, J., & Rothstein, H. R. (2009). Introduction to meta-analysis.

West Sussex, UK: John Wiley & Sons, Ltd

Bouckenooghe, D., Zafar, A., & Raja, U. (2014). How ethical leadership shapes employees’ job

performance: the mediating roles of goal congruence and psychological capital. Journal of

Business Ethics, 129(2), 251-264.

Bouzari, M., & Karatepe, O. M. (2017). Test of a mediation model of psychological capital among

hotel salespeople. International Journal of Contemporary Hospitality Management, 29(8),

2178-2197.

Bown, M. J., & Sutton, A. J. (2010). Quality control in systematic reviews and meta-analyses.

European Journal of Vascular and Endovascular Surgery, 40(5), 669-677.

Buch, R., Martinsen, Ø. L., & Kuvaas, B. (2015). The destructiveness of laissez-faire leadership

behavior: the mediating role of economic leader–member exchange relationships. Journal of

Leadership & Organizational Studies, 22(1), 115-124.

Card, N. A. (2012). Applied meta-analysis for social science research. New York: The Guilford Press.

Cassidy, T., McLaughlin, M., & McDowell, E. (2014). Bullying and health at work: the mediating

roles of psychological capital and social support. Work & Stress, 28(3), 255-269.

Cenciotti, R., Alessandri, G., & Borgogni, L. (2017). Psychological capital and career success over

time: the mediating role of job crafting. Journal of Leadership & Organizational Studies, 24(3),

372-384.

Cleophas, T. J., & Zwinderman, A. H. (2017). Modern meta-analysis: review and update of

methodologies. Switzerland: Springer International Publishing.

Cohen, J. (1977). Statistical power analysis for the behavioural sciences. Hillsdale, NJ, England.:

Lawrence Erlbaum Associates, Inc.

37
Corner, K. J. (2015). Exploring the reliability and validity of research instruments to examine

secondary school principals' authentic leadership behavior and psychological capital: The

University of Arizona.

Cunha, M. P. e., Rego, A., Simpson, A. V., & Clegg, S. (2020). Positive organizational behaviour: a

reflective approach. UK: Routledge.

Dawkins, S., Martin, A., Scott, J., Sanderson, K., & Schüz, B. (2018). A cross-level model of team-

level psychological capital (PsyCap) and individual- and team-level outcomes. Journal of

Management & Organization, 1-20.

Duval, S. (2005). The Trim and Fill method. In H. R. Rothstein, A. J. Sutton, & M. Borenstein (Eds.),

Publication bias in meta-analysis:prevention, assessment and adjustments (pp. 127-144).

England: John Wiley & Sons, Ltd.

Duval, S., & Tweedie, R. (2000). Trim and Fill: a simple funnel-plot–based method of testing and

adjusting for publication bias in meta-analysis. Biometrics, 56(2), 455-463.

Fontes, A., & Dello Russo, S. (2020). An Experimental Field Study on the Effects of Coaching: The

Mediating Role of Psychological Capital. Applied Psychology, n/a(n/a), 1-30.

Fusco, T. (2018). An evidence-based approach to authentic leadership development. London, United

Kingdom: Routledge.

Ganotice, F. A., Yeung, S. S., Beguina, L. A., & Villarosa, J. B. (2016). In search for H.E.R.O among

Filipino teachers: the relationship of positive psychological capital and work-related outcomes.

The Asia-Pacific Education Researcher, 25(3), 407-414.

Gardner, W. L., Avolio, B. J., Luthans, F., May, D. R., & Walumbwa, F. (2005). “Can you see the real

me?” a self-based model of authentic leader and follower development. The Leadership

Quarterly, 16(3), 343-372.

38
Gooty, J., Gavin, M., Johnson, P. D., Frazier, M. L., & Snow, D. B. (2009). In the Eyes of the beholder:

transformational leadership, positive psychological capital, and performance. Journal of

Leadership & Organizational Studies, 15(4), 353-367.

Guo, L., Xiyuan, L., & Qin, C. (2012). Investigating the relationship between MBA education and

professional managers' career satisfaction—mediating function of psychological capital.

Paper presented at the International Conference on Information Management, Innovation

Management and Industrial Engineering (ICIII), pp. 227-231.

Heled, E., Somech, A., & Waters, L. (2016). Psychological capital as a team phenomenon: mediating

the relationship between learning climate and outcomes at the individual and team levels. The

Journal of Positive Psychology, 11(3), 303-314.

Higgins, J. P. T., Thompson, S. G., Deeks, J. J., & Altman, D. G. (2003). Measuring inconsistency in

meta-analyses. BMJ, 327(7414), 557-560.

Hite, B. C. (2015). Positive psychological capital, need satisfaction, performance, and well-being in

actors and stunt people. Walden University,

Hobfoll, S. E. (1989). Conservation of resources: a new attempt at conceptualizing stress. American

Psychologist, 44(3), 513-524.

Hobfoll, S. E. (2011). Conservation of resource caravans and engaged settings. Journal of

Occupational and Organizational Psychology, 84(1), 116-122.

Hobfoll, S. E., Halbesleben, J., Neveu, J.-P., & Westman, M. (2018). Conservation of resources in the

organizational context: the reality of resources and their consequences. Annual Review of

Organizational Psychology and Organizational Behavior, 5(1), 103-128.

Hoch, J. E., Bommer, W. H., Dulebohn, J. H., & Wu, D. (2018). Do ethical, authentic, and servant

leadership explain variance above and beyond transformational leadership? a meta-analysis.

Journal of Management, 44(2), 501-529.

39
Hofstede, G., Hofstede, G. J., & Minkov, M. (2010). Cultures and organizations: software of the Mind

(Rev. 3rd ed. ed.). New York: McGraw-Hill.

Huang, L., & Luthans, F. (2015). Toward better understanding of the learning goal orientation–

creativity relationship: the role of positive psychological capital. Applied Psychology, 64(2),

444-472.

Huber, P. J. (1980). Robust statistics. New York: Wiley.

Hunter, J. E., & Schmidt, F. L. (2004). Methods of meta-analysis: correcting error and bias in research

findings. Thousand Oaks, CA: SAGE.

Hystad, S. W., Bartone, P. T., & Eid, J. (2014). Positive organizational behavior and safety in the

offshore oil industry: exploring the determinants of positive safety climate. The Journal of

Positive Psychology, 9(1), 42-53.

Idris, A. M., & Manganaro, M. (2017). Relationships between psychological capital, job satisfaction,

and organizational commitment in the Saudi oil and petrochemical industries. Journal of

Human Behavior in the Social Environment, 27(4), 251-269.

Karatepe, O. M., & Karadas, G. (2015). Do psychological capital and work engagement foster frontline

employees’ satisfaction?: a study in the hotel industry. International Journal of Contemporary

Hospitality Management, 27(6), 1254-1278.

Karatepe, O. M., & Talebzadeh, N. (2016). An empirical investigation of psychological capital among

flight attendants. Journal of Air Transport Management, 55(Supplement C), 193-202.

Kim, M., Kim, A. C. H., Newman, J. I., Ferris, G. R., & Perrewé, P. L. (2019). The antecedents and

consequences of positive organizational behavior: the role of psychological capital for

promoting employee well-being in sport organizations. Sport Management Review, 22(1), 108-

125.

40
Kim, T., Karatepe, O. M., Lee, G., Lee, S., Hur, K., & Xijing, C. (2017). Does hotel employees’ quality

of work life mediate the effect of psychological capital on job outcomes? International Journal

of Contemporary Hospitality Management, 29(6), 1638-1657.

Kirca, A. H., Jayachandran, S., & Bearden, W. O. (2005). Market orientation: a meta-analytic review

and assessment of its antecedents and impact on performance. Journal of Marketing, 69(2), 24-

41.

Kong, F., Tsai, C.-H., Tsai, F.-S., Huang, W., & de la Cruz, S. (2018). Psychological capital research:

a meta-analysis and implications for management sustainability. Sustainability, 10(10), 3457.

Lee, Y.-C., & Wu, W.-L. (2016). Do employees share knowledge when encountering abusive

supervision? Journal of Managerial Psychology, 31(1), 154-168.

León-Pérez, J. M., Antino, M., & León-Rubio, J. M. (2017). Adaptation of the short version of the

psychological capital questionnaire (PCQ-12) into Spanish / Adaptación al español de la

versión reducida del Cuestionario de Capital Psicológico (PCQ-12). International Journal of

Social Psychology, 32(1), 196-213.

Lindell, M. K., & Whitney, D. J. (2001). Accounting for common method variance in cross-sectional

research designs. Journal of Applied Psychology, 86(1), 114-121.

Lupșa, D., Vîrga, D., Maricuțoiu, L. P., & Rusu, A. (2020). Increasing psychological capital: a pre-

registered meta-analysis of controlled interventions. Applied Psychology, 69(4), 1506-1556.

Luthans, F., Avey, J. B., Avolio, B. J., Norman, S. M., & Combs, G. M. (2006). Psychological capital

development: toward a micro-intervention. Journal of Organizational Behavior, 27(3), 387-

393.

Luthans, F., & Youssef-Morgan, C. M. (2017). Psychological capital: an evidence-based positive

approach. Annual Review of Organizational Psychology and Organizational Behavior, 4(1),

339-366.

41
Luthans, F., Youssef, C. M., & Avolio, B. J. (2007). Psychological capital: developing the human

competitive edge: Oxford University Press, USA.

Luthans, F., Youssef, C. M., & Avolio, B. J. (2015). Psychological capital and beyond: Oxford

University Press, USA.

Mackay, M. M., Allen, J. A., & Landis, R. S. (2017). Investigating the incremental validity of

employee engagement in the prediction of employee effectiveness: a meta-analytic path

analysis. Human Resource Management Review, 27(1), 108-120.

Madrid, H. P., Diaz, M. T., Leka, S., Leiva, P. I., & Barros, E. (2017). A finer grained approach to

psychological capital and work performance. Journal of Business and Psychology.

Majeed, M., & Fatima, T. (2020). Impact of exploitative leadership on psychological distress: a study

of nurses. Journal of Nursing Management, 28(7), 1713-1724.

Malekitabar, M., Riahi, M., & Malekitabar, A. R. (2017). The role of psychological capital in

psychological well-being and job burnout of high schools principals in Saveh, Iran. Iranian

Journal of Psychiatry and Behavioral Sciences, 11(1).

Malik, N., & Dhar, R. L. (2017). Authentic leadership and its impact on extra role behaviour of nurses:

the mediating role of psychological capital and the moderating role of autonomy. Personnel

Review, 46(2), 277-296.

Manzano-García, G., & Ayala, J.-C. (2017). Relationship between psychological capital and

psychological well-being of direct support staff of specialist autism services. the mediator role

of burnout. Frontiers in Psychology, 8(2277).

Marques, J. (2020). Awakened leaders and conscious followers: leading mindful change. In S. Dhiman

& J. Marques (Eds.), New horizons in positive leadership and change: a practical guide for

workplace transformation (pp. 19-32). Switzerland: Springer Nature Switzerland AG.

42
Mazzetti, G., Guglielmi, D., Chiesa, R., & Mariani, M. G. (2016). Happy employees in a resourceful

workplace: just a direct relationship?: a study on the mediational role of psychological capital.

Career Development International, 21(7), 682-696.

McIver, J., & Carmines, E. G. (1981). Unidimensional scaling (Vol. 24). Thousands Oaks, CA: Sage.

Munyaka, S. A., Boshoff, A. B., Pietersen, J., & Snelgar, R. (2017). The relationships between

authentic leadership, psychological capital, psychological climate, team commitment and

intention to quit. SA Journal of Industrial Psychology, 43, 1-11.

Newman, A., Nielsen, I., Smyth, R., Hirst, G., & Kennedy, S. (2018). The effects of diversity climate

on the work attitudes of refugee employees: the mediating role of psychological capital and

moderating role of ethnic identity. Journal of Vocational Behavior, 105, 147-158.

Newman, A., Ucbasaran, D., Zhu, F., & Hirst, G. (2014). Psychological capital: a review and synthesis.

Journal of Organizational Behavior, 35(S1), S120-S138.

Nguyen, T. D., & Nguyen, T. T. M. (2012). Psychological Capital, quality of work life, and quality of

life of marketers: evidence from Vietnam. Journal of Macromarketing, 32(1), 87-95.

Nolzen, N. (2018). The concept of psychological capital: a comprehensive review. Management

Review Quarterly, 68(3), 237-277.

Park, J. G., Kim, J. S., Yoon, S. W., & Joo, B.-K. (2017). The effects of empowering leadership on

psychological well-being and job engagement: the mediating role of psychological capital.

Leadership & Organization Development Journal, 38(3), 350-367.

Polatci, S., & Akdogan, A. (2014). Psychological capital and performance: the mediating role of work

family spillover and psychological well-being. Business and Economics Research Journal,

5(1), 1.

Raja, U., Haq, I. U., De Clercq, D., & Azeem, M. U. (2020). When ethics create misfit: combined

effects of despotic leadership and Islamic work ethic on job performance, job satisfaction, and

psychological well-being. International Journal of Psychology, 55(3), 332-341.

43
Rego, A., Sousa, F., Marques, C., & Cunha, M. P. E. (2012). Authentic leadership promoting

employees' psychological capital and creativity. Journal of Business Research, 65(3), 429-437.

Schmidt, F. L., & Hunter, J. E. (2015). Methods of meta-analysis: correcting error and bias in research

findings (Third Edition ed.). USA: SAGE.

Schuckert, M., Kim, T. T., Paek, S., & Lee, G. (2018). Motivate to innovate: how authentic and

transformational leaders influence employees’ psychological capital and service innovation

behavior. International Journal of Contemporary Hospitality Management, 30(2), 776-796.

Siu, O. L., Cheung, F., & Lui, S. (2015). Linking Positive emotions to work well-being and turnover

intention among Hong Kong police officers: the role of psychological capital. Journal of

Happiness Studies, 16(2), 367-380.

Stander, M. W., & Coxen, L. (2017). A review of the relationship between positive leadership styles

and psychological ownership. In C. Olckers, L. van Zyl, & L. van der Vaart (Eds.), Theoretical

orientations and practical applications of psychological ownership (pp. 37-60). Cham:

Springer International Publishing.

Tukey, J. W. (1960). A survey of sampling from contaminated distributions. In L. Olkin, J. G. Ghurye,

W. Hoeffding, W. G. Madoo, & H. Mann (Eds.), Contributions to probability and statistics

(pp. 448-485). Stanford, California: Stanford University Press.

Tüzün, I. K., Çetin, F., & Basim, H. N. (2018). Improving job performance through identification and

psychological capital. International Journal of Productivity and Performance Management,

67(1), 155-170.

Van der Stede, W. A. (2014). A manipulationist view of causality in cross-sectional survey research.

Accounting, Organizations and Society, 39(7), 567-574.

Van Steenbergen, E. F., van der Ven, C., Peeters, M. C. W., & Taris, T. W. (2017). Transitioning

towards new ways of working: do job demands, job resources, burnout, and engagement

change? Psychological Reports, 0(0), 0033294117740134.

44
Viechtbauer, W., & Cheung, M. W.-L. (2010). Outlier and influence diagnostics for meta-analysis.

Research Synthesis Methods, 1(2), 112-125.

Wang, Y., Chang, Y., Fu, J., & Wang, L. (2012). Work-family conflict and burnout among Chinese

female nurses: the mediating effect of psychological capital. BMC Public Health, 12(1), 915.

Woolley, L., Caza, A., & Levy, L. (2011). Authentic leadership and follower development:

psychological capital, positive work climate, and gender. Journal of Leadership &

Organizational Studies, 18(4), 438-448.

Wu, W.-Y., & Nguyen, K.-V. H. (2019). The antecedents and consequences of psychological capital:

a meta-analytic approach. Leadership & Organization Development Journal, 40(4), 435-456.

Yun, S., & Kang, J. (2018). Influencing factors and consequences of workplace bullying among nurses:

a structural equation modeling. Asian Nursing Research, 12(1), 26-33.

Zhang, X., Li, Y.-L., Ma, S., Hu, J., & Jiang, L. (2014). A structured reading materials-based

intervention program to develop the psychological capital of Chinese employees. Social

Behavior and Personality: An International Journal, 42(3), 503-515.

45
 Figure 1

Flow chart of included/excluded studies in May 2018 and February 2021

Initial search to identify published studies (May 2018) Initial search to identify unpublished studies (May 2018) Initial search to identify published studies (Feb. 2021) Initial search to identify unpublished
studies (Feb. 2021)
(n = 1260) (n = 1909) (n = 1600)
(n =9635)
web of Science (n = 404) ProQuest (n = 1906) (limited access) web of Science (n = 783)
ProQuest (n = 9635) (limited access)
Scopus (n = 324) Academy of Management (AOM) ListServ (n = 2) Scopus (n = 409)
Unpublished thesis (n=1)
EBSCOhost Research Databases (Business Source In-press studies from journal databases (n = 1) EBSCOhost Research Databases (Business Source
Ultimate; CINAHL Complete) (n = 202) Ultimate; CINAHL Complete) (n = 189)

PsycINFO (n = 278) PsycINFO (n = 141)

Ovid MEDLINE (n = 52) Ovid MEDLINE (n = 78)

Removed studies
Total number of the identified studies (n = 3169) Total number of the identified studies (n =11235)
(n = 28) due to:

Not using PCQ-24 or PCQ-

Removed studies
12 questionnaire (n = 3)
(n = 86) due to:
Not using employee
samples (n = 1) Total number of the potential relevant studies (n =221) Not using PCQ-24 or PCQ-
Total number of the potential relevant studies (n = 183)
12 questionnaire (n = 47)
Measuring a similar Published studies (n = 190) and unpublished studies (n =31)
Published studies (n = 154) and unpublished studies (n = 29) construct but different from Not using employee samples
the focus of this study (n = (n = 6)
4)
not having access to the
Inconsistency in reporting theses (n = 29)
the results (n=1)
Total number of the relevant studies (n = 135) Using other levels of
Total number of the relevant studies (n = 155) Using a single-item analysis beyond individual
measure (n = 3) Total number of published studies (n =133) and total number of
Total number of published studies (n = 143) and total number of level (e.g., team PsyCap) (n
unpublished studies (n =2)
unpublished studies (n = 12) = 4)
not having access to the
theses (n = 16)
Removed studies (n = 34)
Removed studies (n = 12) Due to:
Due to:
Not providing the
Total number of included studies in phase 2 (n = 101) correlation coefficients (n =
Not providing the
Total number of included studies in phase 1 (n = 143) (151 samples, N = 56608) correlation coefficients 34)
(103 samples, N = 39908)
even after being contacted
(n = 12)

Total number of included studies in the meta-analysis

46 N=96416)
(n=244) (254 samples,
 Table 1
Results of the random-effects model meta-analysis, heterogeneity test, and moderation analysis for the relationships between Leadership styles and PsyCap
Analysis Summary effect Heterogeneity of effects
95% CI Trim & Fill
Effect k p Trim Q df p I2 T2 Tau
[LL, UL] Adjusted
Authentic Leadership – PsyCap
Analysis .499 34 [.449, .545] <.001 3 .518 486.42 33 <.01 93.22 .03 .18
Moderation – Industry
type
Manufacturing .412 2 [.187, .595] .001
Service .514 21 [.442, .579] <.001
Ethical Leadership – PsyCap
Analysis .421 7 [.333, .501] <.001 0 .421 51.716 6 <.01 88.40 .02 .13
Transformational Leadership – PsyCap
Analysis .557 9 [.388, .690] <.001 0 .557 360.26 8 <.01 97.78 .11 .33
Moderation – Industry type
Manufacturing .428 1 [.355, .496] <.001
Service .605 4 [.384, .760] <.001
Transactional Leadership – PsyCap
Analysis .240 3 [.192, .288] <.001 0 .240 1.712 2 .425 0 0 0
Servant Leadership – PsyCap
Analysis .462 6 [.389, .529] <.001 1 .437 26.05 5 <.01 80.80 .01 .10
Empowering Leadership – PsyCap
Analysis .563 3 [.442, .663] <.001 0 .563 14.80 2 .001 86.49 .02 .13
Abusive Leadership – PsyCap
Analysis -.349 6 [-.507, -.169] <.001 0 -.349 170.15 5 <.01 97.06 .06 .24
Moderation – Industry type
Manufacturing -.386 1 [-.470, -.295] <.001
Service -.547 2 [-.785, -.169] 0.007

Note: k = number of studies, 95%CI = 95% confidence interval, LL = Lower limit of 95% CI, UL = Upper limit of 95% CI, Trim = studies trimmed in Trim &
Fill analysis, Q = Cochran’s Q, T2 = Tau squared, Trim = suggested unpublished studies using Duval & Tweedie Trim & Fill analysis for publication bias.

47
 Table 2
Results of the random-effects model meta-analysis, heterogeneity test, and moderation analysis for the relationships between PsyCap and outcome variables
Analysis Summary effect Heterogeneity of effects
Trim &
95% CI
Effect k p Trim Fill Q df p I2 T2 Tau
[LL, UL]
Adjusted
PsyCap – Self-Reported Job Performance
Analysis .571 28 [.503, .632] <.001 3 .603 570.194 27 <.01 95.27 .06 .25
Manufacturing .751 2 [.703, .793] <.001
Service .580 11 [.441, 0.693] <.001
PsyCap – Supervisor-Reported Job Performance
Analysis .425 27 [.310, .527] <.001 8 (3) .515(.421) 968.25 26 <.01 97.32 .12 .35
Manufacturing .427 2 [.201, .610] <.001
Service .448 14 [.234, .620] <.001
PsyCap – Work Engagement
Analysis .712 85 [.658, .759] <.001 32 (6) .793 (.697) 6027.45 84 <.01 98.61 .22 .47
Manufacturing .706 4 [.518, .829] <.001
Service .716 52 [.639, .779] <.001
PsyCap – Job Satisfaction
Analysis .683 55 [.588, .759] <.001 23 (6) .791 (.657) 7160.79 54 <.01 99.25 .36 .60
Manufacturing .514 5 [.265, .698] <.001
Service .711 30 [.490, .736] <.001
PsyCap – Turnover Intentions
Analysis -.359 42 [-.428, -.287] <.001 0 -.359 973.37 41 <.01 95.79 .07 .26
Manufacturing -.372 2 [-.449, -.291] <.001
Service -.367 26 [-.466, -.259] <.001
PsyCap – Burnout
Analysis -.551 54 [-.644, -.442] <.001 0 -.551 6442.15 53 <.01 99.18 .29 .54
Manufacturing -.266 5 [-.535, .052] .100
Service -.595 44 [-.695, -.472] <.001
Note: k = number of studies, 95%CI = 95% confidence interval, LL = Lower limit of 95% CI, UL = Upper limit of 95% CI, Trim = studies trimmed in Trim &
Fill analysis, Q = Cochran’s Q, T2 = Tau squared, Trim = suggested unpublished studies using Duval & Tweedie Trim & Fill analysis for publication bias.
Wherever necessary, the results of the publication bias analysis after removing the outliers have been reported in parentheses.

48
Table 3
Results of the meta-regression for leadership styles as predictors and PsyCap as outcome (Empowering
Leadership as the reference group)
Coefficien Standard
z p
t Error
Intercept 0.64 0.12 5.40 <.001
Abusive Leadership -1.00 0.14 -6.94 <.001
Authentic Leadership -0.09 0.12 -0.73 .465
Ethical Leadership -0.19 0.14 -1.31 .190
Servant Leadership -0.15 0.15 -1.03 .302
Transactional Leadership -0.39 0.17 -2.34 .019
Transformational Leadership -0.01 0.14 -0.08 .937

Table 4
Results of the meta-regression for leadership styles as predictors and PsyCap as outcome (Transformational
Leadership as the reference group)
Standard
Coefficient z p
Error
Intercept 0.63 0.07 9.21 <.001
Abusive Leadership -0.99 0.11 -9.23 <.001
Authentic Leadership -0.8 0.08 -1.03 .301
Empowering Leadership 0.01 0.14 0.08 .937
Ethical Leadership -0.17 0.10 -1.69 .091
Servant Leadership -0.14 0.11 -1.28 .199
Transactional Leadership -0.38 0.14 -2.80 .005

Table 5
Results of the moderation analysis for the identified relationships
95% CI
Analysis Effect k p
[LL, UL]
PsyCap – Supervisor-Reported Job Performance
Cross-sectional .447 21 [.312, .563] <.001
Longitudinal .341 6 [.132, .522] .002

PsyCap – Work Engagement

Cross-sectional .718 78 [.661, .768] <.001
Longitudinal .610 7 [.583, .636] <.001

PsyCap – Job Satisfaction

Cross-sectional .673 50 [.569, .756] <.001
Longitudinal .764 5 [.617, .859] <.001

49
Table 6
A Summary Comparison of Previous PsyCap Meta-Analyses and the Current Study
Avey et al. (2011) Kong et al. Wu and Current Meta-
(2018) Nguyen Analysis
(2019)
Total number of included studies 45 77 105 244

Included number of unpublished studies N/R (The exact N/A N/A 15

number has not been
provided)

Number of independent samples 51 N/R N/R 254

Total sample size 12567 N/R N/R 96416

Independent samples for AL N/A 5 13 34

Independent samples for ETL N/A N/A 3 7
Independent samples TRFL N/A N/A N/A 9
Independent samples for TRAL N/A N/A N/A 3
Independent samples for SL N/A N/A N/A 6
Independent samples for EMPL N/A N/A N/A 3
Independent samples for ABL N/A N/A 5 6
Independent samples for JS 10 16 14 55

Independent samples for SEP 6 N/R N/A 28

Independent samples for SUP 15 N/R N/A 27

Independent samples for TI 5 N/A N/A 42

Independent samples for burnout N/A N/A N/A 54

Independent samples for WE N/A N/A N/A 85

Samples from the USA 26 N/R N/R 38

Samples from outside of the USA 7 N/R N/R 209
Working adult sample 23 N/R N/R 254

Samples in the service industry 10 N/R N/R 154

Samples in the manufacturing industry 4 N/R N/R 19
Notes: N/A = not assessed, N/R = not reported, AL = Authentic Leadership, ETL = Ethical Leadership, TRFL =
Transformational Leadership, TRAL = Transactional Leadership, SL = Servant Leadership, EMPL = Empowering
Leadership, ABL = Abusive Leadership, JS = Job satisfaction, SEP = Self-reported Performance, SUP = Supervisor-
reported Performance, TI = Turnover Intentions, WE = Work Engagement.

50