Journal of Business Ethics (2020) 161:627–639

https://doi.org/10.1007/s10551-018-3948-5

ORIGINAL PAPER

Antecedents of Duty Orientation and Follower Work Behavior: The

Interactive Effects of Perceived Organizational Support and Ethical
Leadership
Nathan Eva1 · Alexander Newman2 · Qing Miao3 · Dan Wang1 · Brian Cooper1

Received: 20 March 2017 / Accepted: 13 June 2018 / Published online: 25 June 2018
© Springer Nature B.V. 2018

Abstract
Drawing on social exchange theory, the present study seeks to understand how ethical leaders channel followers’ responses
to positive treatment from the organization into a dutiful mindset, resulting in in-role and extra-role performance. Specifi-
cally, it examines the influence of perceived organizational support on both followers’ job performance and organizational
citizenship behaviors, and the mediating effects of duty orientation on such relationships. In addition, it examines whether
the mediated effects are contingent on the ethical leadership exhibited by the team leader. Based on multi-source, multi-level
data obtained from 233 employees in 60 teams from the Chinese public sector, we found that ethical leadership moderated
the mediated relationship between perceived organizational support and follower work behaviors through duty orientation,
such that this relationship was stronger in the presence of higher ethical leadership.

Keywords Ethical leadership · Duty orientation · Perceived organizational support · Organizational citizenship behaviors ·
Job performance

Introduction
Editors at the Journal of Business Ethics are recused from all
decisions relating to submissions with which there is any identified A growing body of empirical work has examined the critical
potential conflict of interest. Submissions to the Journal of role played by ethical leadership in limiting the unethical
Business Ethics from editors of the journal are handled by a senior behavior of followers (Mayer et al. 2010). Researchers have
independent editor at the journal and subject to full double blind
also found a positive relationship between ethical leadership
peer review processes.
and both the in-role and extra-role behaviors of followers
* Qing Miao (Mayer et al. 2009; Newman et al. 2014; Walumbwa et al.
mqok@163.com 2011). However, recent calls from scholars (e.g., Den Hartog
Nathan Eva 2015; Mayer et al. 2012) have stressed the need for greater
nathan.eva@monash.edu research into how ethical leaders influence the work behav-
Alexander Newman ior of followers within organizations. In response, this study
a.newman@deakin.edu.au examines how ethical leaders channel followers’ responses
Dan Wang to positive treatment from the organization into a dutiful
dan.wang@monash.edu mindset, resulting in in-role and extra-role performance.
Brian Cooper Specifically, we analyze how ethical leadership moderates
brian.cooper@monash.edu the mediated relationship of perceived organizational sup-
1 port (POS) through duty orientation on job performance and
Monash Business School, Monash University, Caulfield,
Australia organizational citizenship behaviors (OCBs). In doing so,
2 we contribute to the development of the nascent literature
Deakin Business School, Deakin University, Geelong,
Australia on duty orientation, and address the calls of researchers to
3 examine how ethical leadership may interact with employee
Center of Social Welfare & Governance, School
of Public Affairs, Zhejiang University, Hangzhou, perceptions of the organizational context to influence the
People’s Republic of China work outcomes of followers (Ko et al. 2017).

13
Vol.:(0123456789)
628 N. Eva et al.

Duty orientation has been defined as “an individual’s contribution by answering the calls of researchers to examine
volitional orientation to serve and faithfully support other contextual moderators which explain the influence of POS on
members of the group, to strive and sacrifice to accomplish employee work outcomes (Kurtessis et al. 2017). From a more
the tasks and missions of the group, and to honor its codes practical perspective, our study highlights the importance to
and principles” (Hannah et al. 2014, p. 220). Hannah et al. organizations of creating a climate where employees feel sup-
(2014) introduced the concept of duty orientation to the ported, and investing in the selection and development of ethi-
literature to capture how employees make decisions about cal leaders.
how to behavior in the workplace based on their own moral In the proceeding sections, we first review the literature on
perspectives. While Hannah and his colleagues called for POS, duty orientation, and ethical leadership, before develop-
future research to investigate potential antecedents of duty ing hypotheses. We then explain the sampling method and data
orientation and whether duty orientation promotes perfor- collection procedures, measures, and the methods of analysis
mance outcomes (Hannah et al. 2014), this research has yet utilized, before presenting the results of the study. Lastly, we
to materialize. discuss the findings and theoretical and practical implications
With this in mind, the current study examines perceived arising from the study, before recommending avenues for
organizational support as an antecedent of duty orientation future research.
and whether duty orientation influences the job performance
and organizational citizenship behaviors (OCBs) of follow- Literature Review
ers, defined as discretionary behaviors on the part of employ-
ees which enable their team members to deliver work that Perceived Organizational Support and Followers’
assists with the organization’s missions and goals (Graham Work Behaviors
1991). Perceived organizational support (POS) refers to the
extent to which “employees develop global beliefs concern- High levels of perceived organizational support create feel-
ing the extent to which the organization values their con- ings of obligation on the part of employees towards the
tributions and cares about their well-being” (Eisenberger organization (Eisenberger et al. 2001). This in turn leads
et al. 1986, p. 501). Drawing on social exchange theory employees to feel committed to achieving the organization’s
(Blau 1964), we argue that when employees are provided goals (Wayne et al. 1997) and incorporate organizational
with support by the organization, they feel greater obliga- membership into their social identity (Rhoades and Eisen-
tion towards the organization and their team members (duty berger 2002). In other words, when employees feel that their
orientation), and are more likely to reciprocate in kind with organization supports them in their daily work, through pro-
positive in-role and extra-role behaviors directed towards the viding them with appropriate rewards and demonstrating
organization and their team. care for their well-being, employees typically reciprocate
As highlighted earlier, we also argue that the social with improved attitudes towards the organization (Djurkovic
exchange process of POS fostering job performance and et al. 2008; Edwards 2009), and higher levels of in-role and
OCBs through the mediating mechanism of duty orienta- extra-role performance (Eisenberger et al. 1986; Kaufman
tion is amplified by the presence of an ethical leader. Ethi- et al. 2001; Wong et al. 2012).
cal leadership has been defined as “the demonstration of These findings were confirmed in a recent meta-analysis
normatively appropriate conduct through personal actions on POS by Kurtessis et al. (2017). Specifically, they demon-
and interpersonal relationships, and the promotion of such strated that through the social exchange process, POS was
conduct to followers through two-way communication, shown to positively predict felt obligation, affective and
reinforcement, and decision-making” (Brown et al. 2005, normative commitment, organizational identification to the
p. 120). Our study argues that the role modeling and advice organization, and the outcome variables of OCBI, OCBO,
offered by ethical leaders provides followers with guidance and job performance which are the subject of this study.
on the behaviors they should display in response to the sup- OCBs refer to discretionary extra-role work behaviors aimed
port showed to them by the organization. at the individual (OCBI) or the organization (OCBO) which
Overall, our study makes an important contribution to the are not part of the job description (Lee and Allen 2002),
ethics literature by highlighting the social exchange process whereas job performance refers to in-role behavior which is
by which the organization influences the in-role and extra-role undertaken to meet the requirements of the job.
work behaviors of followers, and how the ethical leader can
help shape the extent to which followers engage in duty-ori-
ented behaviors in response to their perceptions of supportive
organizational practices. In examining the cross-level interac-
tion of POS as an individual-level variable and ethical leader-
ship as a team-level variable, we make an important empirical

13
Antecedents of Duty Orientation and Follower Work Behavior: The Interactive Effects of Perceived… 629

Duty Orientation as a Mediator in the Relationship faithfully support other members of the group. Such behav-
Between Perceived Organizational Support iors may include furnishing assistance to team members who
and Followers’ Work Behaviors need help, mentoring and training other team members, and
covering team members who are absent from work. Duty
Duty orientation comprises three core elements of duty; a orientation also lessens group-limiting behaviors such as
duty to loyally serve and faithfully support team members speaking negatively about others in the team (Hannah et al.
(duty to members); a duty to achieve the organization’s mis- 2014). In showing duty to the mission of the organization,
sion and tasks through sacrifice and expending additional employees are likely to increase both their job performance
effort (duty to mission); and a duty to uphold the principles and OCBs directed towards the organization (OCBO), as
and honor the codes of the team, as well as doing what is duty to mission reflects the extent to which the employee
right according to the norms and morals of the group (duty feels a moral obligation to support the mission of the organi-
to codes) (Hannah et al. 2014). Taken together, the three zation by performing their job duties to their maximum abil-
components of duty orientation comprise a normative ori- ity as they are compensated by the organization to do, and
entation towards doing what is right for the team and organi- go above and beyond what is required of them in their job
zation (Hannah et al. 2014). Employees who possess high role. This leads to:
levels of duty orientation are more likely to think about team
and organizational issues through the lens of their duty to H1 Duty orientation mediates the relationship between per-
others in their team and the organizational as a whole. As a ceived organizational support and followers’ (a) OCBI; (b)
result, they are more likely to act in a way that benefits their OCBO; and (c) job performance.
team and the organization compared to those with low levels
of duty orientation (Hannah et al. 2014). Ethical Leadership as a Moderator in the Mediated
Before discussing the role of duty orientation as a mecha- Relationship Between Perceived Organizational
nism linking POS to OCBs and job performance, we must Support and Followers’ Work Behaviors Through
first distinguish of the concept of felt obligation from duty Duty Orientation
orientation, as felt obligation has been previously been
found to mediate the relationship between POS and forms Brown et al. (2005) highlight two key characteristics of ethi-
of in-role and extra-role performance (i.e., organizational cal leaders, being a moral person and a moral manager. The
spontaneity) (Eisenberger et al. 2001). Felt obligation “is a moral person dimension highlights the exhibition of certain
prescriptive belief regarding whether one should care about personal attributes by the leader including honesty, integrity,
the organization’s well-being and should help the organiza- and altruism, while the moral manager dimension highlights
tion reach its goals” (Eisenberger et al. 2001, p. 42). Duty the role played by the leader in influencing followers towards
orientation goes beyond felt obligation as it not only focuses the adoption of good ethical conduct through engaging in
on the obligation to the organization (duty to mission), but ethical role modeling, and communicating the need for fol-
also to the members of the team (duty to members) and the lowers to act in an ethical and dutiful manner. Ethical lead-
morals of the group (duty to codes) (Hannah et al. 2014). ership has been shown to influence followers’ behaviors in
Thus, we argue duty orientation provides a more complete a number of studies. For example, it has been found to pro-
explanation than felt obligation as to why employees per- mote positive work behaviors among employees including
form extra-role behaviors due to their duty to their team followers’ citizenship behaviors (Avey et al. 2011; Bedi et al.
members, and why employees perform in-role behaviors 2016; Kacmar et al. 2011; Kalshoven et al. 2011; Newman
due to their duty to the mission of the organization (Hannah et al. 2014), and task performance (Ng and Feldman 2015;
et al. 2014). Walumbwa et al. 2011), as well as reducing counterproduc-
Therefore, we contend that duty orientation acts as a tive and unethical behavior (Mayer et al. 2010).
mechanism linking POS to OCBs and job performance. This We contend that ethical leadership will enhance fol-
assertion is based on social exchange theory (Blau 1964) lowers’ duty orientation and therefore OCBs and job per-
which suggests that when employees feel supported by the formance, by strengthening the influence of POS on duty
organization they will develop a strong exchange relation- orientation. More specifically, we argue that employees’
ship with the organization. In return for the support provided duty orientation is enhanced by the ethical leader, through
by the organization, employees will feel an obligation to amplifying the social exchange process that results from
reciprocate in kind by being dutiful towards the organiza- the receipt of support by their organization (Blau 1964). In
tion, which is likely to heighten followers’ job performance Brown et al.’s (2005) review of ethical leadership, they argue
and OCBs in a number of ways. Through being dutiful to that the social exchange relationships with followers are built
members, employees will engage in OCBs that benefit oth- on norms of reciprocity (Gouldner 1960). As a result of the
ers in the team (OCBI), as they feel obliged to loyally and fair and caring treatment followers receive from the ethical

13
630 N. Eva et al.

leader, they will feel morally obliged to return in kind with that the positive relationship becomes stronger as ethical
behaviors that go above and beyond the call of duty (Brown leadership increases.
et al. 2005). Drawing from the moral norms of reciproc-
ity (Gouldner 1960), it can be argued that reciprocating the H3 The mediated relationship between perceived organiza-
support shown by the organization is a moral act on behalf tional support and follower (a) OCBI, (b) OCBO, and (c)
of the employee. This is then enhanced by the moral agent job performance through duty orientation, is moderated by
of the organization (i.e., the ethical leader) who emphasizes ethical leadership, such that the mediated relationships are
the importance of delivering on one’s moral obligations stronger when ethical leadership is higher.
(Brown et al. 2005), acts as a role model for this behavior An overview of our research model can be found in Fig 1.
by ‘walking the talk’ (Mayer et al. 2009), and provides clar-
ity over expected social exchange behaviors that followers
should demonstrate in response to the support offered by
the organization. Method
For example, through an ethical leader’s actions and
words, they encourage followers to reciprocate the provision Sample and Procedures
of support by the organization in the form of being loyal to
and faithfully supporting team members (duty to members); In early 2015, a member of the research team ran a lead-
achieving the organization’s mission and tasks through sac- ership development program for senior managers in the
rifice and expending additional effort (duty to mission); and public sector in Zhejiang Province which is located in the
upholding the principles and honoring the codes of the team, Yangtze River Delta Region of China. From this, a database
as well as doing what is right according to the norms and of government bureaus that would be willing to participate
morals of the group (duty to codes) (Hannah et al. 2014). in research was developed. For the purposes of this project
As explained earlier, this should in turn increase employees’ four cities from Zhejiang Province were selected from the
OCBs and job performance. In contrast, we argue that in sit- database, from which 71 senior managers working in pub-
uations where employees work under a less ethical leader, lic sector labor and social security bureaus were contacted
they will be less likely to reciprocate the positive treatment to ascertain their willingness to participate in the present
forwarded by the organization, and therefore less likely to research study (22 from Hangzhou, 16 from Ningbo, 15
develop a sense of loyalty to team members, the mission of from Fenghua, and 18 from Xiaoshan). The senior manag-
the organization, and team codes, resulting in lower levels ers were sent an email outlining the purpose of the study
of OCBs and job performance than when employees work and asking if they were willing to participate in the research.
under a more ethical leader. This leads us to the following A total of 60 senior managers agreed to participate in the
two hypotheses: research (a response rate of 85%). Each of the senior man-
agers provided the names of their immediate subordinates
H2 Ethical leadership moderates the relationship between who were mid- to low-level managers in their bureaus.
perceived organizational support and duty orientation, such Two hundred and forty-eight of their subordinates across
the 60 bureaus were invited to participate in the study, of

Fig. 1  Hypothesized model Team

Ethical
Leadership

Individual
Individual
Performance

Perceived
Duty
Organizational OCB-I
Orientation
Support

OCB-O

13
Antecedents of Duty Orientation and Follower Work Behavior: The Interactive Effects of Perceived… 631

which 233 responded (a response rate of 94%). Examples Previously validated multi-item scales were used to meas-
of the positions the senior managers held are Head of Labor ure each variable.
Law Department, Head of Human Resource Department,
and Head of Social Insurance. Examples of the positions Perceived Organizational Support
subordinates held are Director of Labor Relations, Director
of Human Resource Information Services, and Director of Employees rated the perceived support they receive from
Medical Insurance. A non-response bias test for both sen- the organization using Eisenberger et al.’s (1986) 8-item
ior managers (n = 11) and subordinates (n = 15) was run on perceived organizational support scale (α = 0.88). A sample
the basis of gender, which was the only variable other than item is “Help is available from my organization when I have
email addresses that we had from the non-respondents. Anal- a problem.”
ysis showed there were no significant differences in terms
of response rates according to gender for either group. As Ethical Leadership
there were only 11 non-responding senior managers and 15
non-responding subordinates in the study, and respondents Employees assessed their supervisor’s ethical leadership
and non-respondents are employed in similar occupations, behavior using Brown et al.’s (2005) 10-item ethical lead-
non-response bias is unlikely to be a problem in this study. ership scale (α = 0.92). A sample item is “My supervisor
Data were collected using a survey questionnaire meth- discusses business ethics or values with employees.”
odology from both the senior managers and their immedi-
ate subordinates across three time periods to minimize the
Duty Orientation
effects of common method bias which is often associated
with single-respondent and single-time data (Podsakoff et al.
Employees rated their own duty orientation using the
2012). Prior to distribution, the questionnaire was translated
12-item duty orientation scale (α = 0.93) developed by Han-
from English to Mandarin by bilingual members of the
nah et al. (2014). Sample items included “Am loyal to my
research team using the back-translation procedure (Brislin
leaders and team” (duty to members); “Do whatever it takes
2000). The participants received a cover letter, paper-based
to not let the mission/organization goals fail” (duty to mis-
questionnaire, and an envelope at work. After completing the
sion); and “Do what is right always” (duty to codes).
questionnaire in their own time, employees were informed
to seal the competed questionnaire in the envelope provided
and submit it to the research team directly. In order to main- Job Performance
tain confidentiality of the participants, each of the question-
naires was coded with a unique identification number in The supervisor rated the job performance of each employee
order to match the subordinate’s response with that of their using the 3-item job performance scale developed by Lam
immediate supervisor. et al. (2002), two items of which were adapted from work
The three points of data collection were undertaken in by Heilman et al. (1992) (𝛼 = 0.88). Sample items included
two week intervals. At time 1, data on the control variables “This employee gets his or her work done very effectively”
and the ethical leadership behaviors of the senior managers and “This employee has performed his/her job well.”
were collected from subordinates. At time 2, 2 weeks later,
data on perceived organizational support and duty orienta- Organizational Citizenship Behaviors
tion were collected from subordinates. At time 3, a further
2 weeks later, senior managers rated the job performance, Managers rated the organizational citizenship behaviors of
OCBI, and OCBO of their subordinates. their employees using the Lee and Allen (2002) 16-item
In total, 233 subordinates from 60 teams completed the scale, from which 8 items measure OCBI and 8 items meas-
questionnaire. The average number of subordinates per team ure OCBO. A sample item from the OCBI scale is “He/
was 3.88 (range 3–6). Most respondents were female (60%), she willingly gives his/her time to help others who have
below 35 years of age (94%), and had completed postgradu- work-related problems” (𝛼 = 0.88). A sample item from the
ate study (50%). The average team tenure was 3.90 years. OCBO scale is “He/she takes action to protect the organiza-
tion from potential problems” (α = 0.89).
Measures
Control Variables
All study variables were measured on a 5-point Likert scale
ranging from 1 = strongly disagree to 5 = strongly agree. Age, gender, education, team tenure, and affective com-
mitment were included as control variables. We included
these control variables as each have been shown to have an

13
632 N. Eva et al.

impact on job performance and/or OCBs in previous stud- does not use aggregated group means as proxies for level-2
ies. Specifically, age was included as a number of studies variables (Preacher et al. 2010).
have found that the age of the subordinate has an impact on More specifically, Hypothesis 1 was tested using a lower-
the performance ratings of the supervisor (e.g., Ferris et al. level (1-1-1) mediation model with random intercepts on the
1985), and previous research has found that older workers between level (Preacher et al. 2010). In the 1-1-1 mediation
perform more OCBs than their younger colleagues (Ehigie model, the independent variable, mediator, and the depend-
and Otukoya 2005; Kuehn and Al-Busaidi 2002). In terms of ent variable are all measured at level-1 (e.g., the individual
gender, a meta-analysis by Roth et al. (2012) showed that on level), but nested in level-2 units (e.g., teams). Following
average females were rated as performing better than males. the work of Preacher et al. (2016), Hypotheses 2 and 3 were
For OCBs, research by Beauregard (2012) demonstrated tested using a moderation model with the moderator (ethical
that females are more likely to perform OCBs than males. leadership) and a randomly varying slope at level-1 (i.e., a
Bearegard indicates that this is through traditional gender “slopes as outcomes” model) to test the cross-level inter-
roles playing out in the workplace with women expected to action. To test the mediation hypotheses, we utilized the
assist others in the pursuit of their goals. A meta-analysis Monte Carlo method for testing indirect effects calculated
of educational attainment showed that the higher the level with asymptotic covariance matrix values (Preacher and
of education the higher an employee’s job performance and Selig 2012). We checked that the correlation between vari-
OCBs (Ng and Feldman 2009). Team tenure is a commonly ables was less than 0.70 (Tabachnick and Fidell 2013), the
used control variable as it is regarded as a proxy for work tolerance values were greater than 0.10, and the variance
experience. As employees spend longer in the team, they inflation factor values were less than 10.0 (Kline 2010), to
accumulate relevant knowledge, ability, and skills for their ensure there were no multicollinearity issues in the data.
role which can translate to high levels of job performance
(Carboni and Ehrlich 2013) and OCBs (Ehigie and Otukoya Validation of Team‑Level Constructs
2005). Further, as affective commitment to the organization
(Allen and Meyer 1990) has been shown to be an outcome Given that we conceptualized ethical leadership as a team-
of POS (Kurtessis et al. 2017) and ethical leadership (Neves level variable, as there are likely to be shared perceptions of
and Story 2015), and also a predictor of OCBs (Gong et al. ethical leadership among team members, we assessed the
2010) and performance (Yu et al. 2016), we included it as a team-level properties using the rWGJ, ICC(1), and ICC(2)
control variable as it might be an alternate mediator. Finally, values. We calculated the median rWGJ for ethical leader-
POS has been shown to be related to age, gender, education, ship using a moderately skewed distribution [see LeBreton
and team tenure (Rhoades and Eisenberger 2002). and Senter (2008) for details] as the ethical leadership data
Age was coded in five-year increments ranging from were moderately skewed. The results indicated a high level
1 = 21–25 to 8 = 56–60. Gender was coded as 1 = male and of within-group agreement when calculated with the mod-
2 = female. As employees needed a minimum college degree erately skewed distribution (median rWGJ = 0.85) and also
to have their position, we coded education as 1 = college when calculated with the uniform null distribution (median
degree and 2 = postgraduate degree. Team tenure (how long rWGJ = 0.92) (James et al. 1984). The interclass correlations
they had worked in their current team) was measured in ICC(1) and ICC(2) were 0.14, and 0.39, respectively. While
years. We used the 6-item affective commitment scale (Allen the results showed strong rWGJ values, the ICC(2) value was
and Meyer 1990). A sample item is “I feel a strong sense of small which can be largely attributed to the small team sizes
belonging to my organization” (α = 0.92). that participated in the study (M = 3.88). The discrepancy
between the ICC(2) and the rWGJ values typically occurs
Analysis Strategy with small group sizes where the between-group variance is
not large compared to within-group agreement and therefore
All regression analyses were conducted using multi-level the ICC(2) estimates are diminished. Given the relatively
structural equation modeling (MSEM) and robust maxi- small team sizes, ICC statistics can be attenuated which can
mum likelihood estimation with Mplus software version 7.4 lead to “erroneous decisions” on the part of researcher in
(Muthén and Muthén 2012). MSEM decomposes the vari- their decisions to aggregate (LeBreton and Senter 2008, p.
ance of a variable in multi-level data into its latent within- 840). Thus, the results of the rWGJ value as well as the theo-
group and between-group components. By decomposing retical justification for ethical leadership was appropriate to
variance in this manner with model-based group-mean cen- examine at the team level. Also, as we use MSEM to decom-
tering, MSEM avoids potential problems of existing multi- pose the variance of our observed variables into their latent
level modeling methods that conflate within- and between- within-group and between-group components, we minimize
level effects. MSEM is also effective in addressing bias as it bias commonly caused by using aggregated group means as
level-2 variables (Preacher et al. 2010).

13
Antecedents of Duty Orientation and Follower Work Behavior: The Interactive Effects of Perceived… 633

Common Method Variance Checks Results

In order to address concerns over common method variance, Confirmatory Factor Analysis
we first used procedural remedies such as collecting the data
across multiple time points from multiple sources, ensuring Confirmatory factor analysis was conducted at the individ-
little ambiguity in the questions, and ensuring confidential- ual level to establish the discriminant validity of the study
ity and anonymity through the use of individually sealed variables using weighted-least squares means and variance-
envelopes (Podsakoff et al. 2012). Secondly, we conducted a adjusted (WLSMV) estimation (with a correction for clus-
marker variable test following the suggestions by Lindell and tering) as the item-level data were rated on ordinal response
Whitney (2001), using the 3-item measure of information scales (Muthén and Muthén 2012). The hypothesized six-
availability from Moynihan, Wright, and Pandey (2012). A factor model (i.e., with items loading onto their correspond-
sample item was “My organization has implemented bench- ing factors of POS, ethical leadership, duty orientation, job
marks for measuring program outcomes or results.” The performance, OCBI, and OCBO) yielded a good fit to the
smallest observed correlation between the marker variable data χ2 (df = 1112) = 1782, RMSEA = 0.05, TLI = 0.95,
and the variables used in the study was r = − 0.03, p = 643. CFI = 0.96. We compared the fit of the six-factor model with
As the correlations are non-significant and weak, this sug- that of alternative models under which items were loaded
gests there is not a substantial amount of common method onto five, four, three, two, and single factors (see Table 1).
variance present in the study (Lindell and Whitney 2001; As argued by Hair et al. (2010), RMSEA values between
Podsakoff et al. 2003). Further, we tested the regression 0.03 and 0.08 and TLI and CFI values above 0.90 indicate
analyses with and without the marker variable and there a good fit. The six-factor model used in this study fit the
was no change in the statistical significance of the results, data better than alternative models. These results provide
with only slight changes in the magnitude. It also should evidence for the discriminant and convergent validity of the
be noted that previous research has shown that significant variables used in this research.
interaction effects (like those displayed in this study) can-
not be obtained when there is common method variance. If Hypothesis Testing
anything, the presence of common method variance would
reduce the significance and the magnitude of the interaction Table 2 shows the descriptive statistics, reliabilities, and cor-
(Siemsen et al. 2010). Taken together, common method vari- relations of the study variables. Consistent with our theo-
ance appears not to be an issue in this study. retical predictions, all of the study variables were positively
correlated.
The results of the MSEM analyses is given in Table 3,
with the conditional indirect effects of the moderated media-
tion model presented in Table 4.

Table 1  Results of confirmatory factor analysis

χ2
Models df RMSEA 90% CI RMSEA CFI TLI

Six-factor model 1781.51 1112 0.051 0.046–0.055 0.957 0.954

Five-factor model (POS and duty orientation combined) 2592.82 1117 0.075 0.072–0.079 0.904 0.899
Five-factor model (ethical leadership and duty orientation combined) 3033.33 1117 0.086 0.082–0.089 0.876 0.869
Four-factor model (POS and ethical leadership combined; OCBI and OCBO 2428.59 1121 0.071 0.067–0.075 0.915 0.911
combined)
Three-factor model (POS, ethical leadership, and duty orientation combined; 3643.93 1124 0.098 0.095–0.102 0.837 0.829
OCBI and OCBO combined)
Three-factor model (POS and ethical leadership combined; OCBI, OCBO, and 2748.05 1124 0.079 0.075–0.082 0.895 0.890
individual job performance combined)
Two-factor model (POS, ethical leadership, and duty orientation combined; OCBI, 3887.15 1126 0.103 0.099–0.106 0.821 0.813
OCBO, and individual job performance combined)
One-factor model 4492.92 1127 0.113 0.110–0.117 0.782 0.772

N = 233. All alternative models were compared to the six-factor model

RMSEA root mean square error of approximation, CFI comparative fit index, TLI Tucker–Lewis index

13
 634 N. Eva et al.

Hypothesis 1 predicted that duty orientation would medi-

(0.88)
11 ate the relationship between perceived organizational sup-
port and OCBI, OCBO, and job performance. As shown in
Table 3, duty orientation mediated the relationships between

0.54**
(0.89)
POS and each of the outcome variables. Specifically, the
10

estimated average indirect effects of POS through duty ori-

entation were 0.09 for OCBI (95% CI 0.03–0.17); 0.07 for

0.60**
0.27**
(0.88)
OCBO (95% CI 0.02–0.14); and 0.13 for job performance
(95% CI 0.04–0.22). As zero was not contained in any of
9

the 95% confidence intervals, Hypothesis 1 was supported.

0.33**
0.65**
0.41**
Hypothesis 2 proposed that ethical leadership would
(0.93)

moderate the relationship between POS and duty orienta-

8

tion. Hypothesis 2 involved testing a cross-level interaction

with a random slope (POS and duty orientation) at level-1.
0.47**
0.52**
0.67**
0.66**
(0.92)

The analysis indicated there was a statistically significant,

7

cross-level interaction (β = 0.38, p < 0.05). To aid in interpre-

tation of the cross-level interaction, we plotted the interac-
0.70**
0.51**
0.33**
0.60**
0.32**
(0.88)

tion using the conventional values of one SD below the mean

and one SD above the mean of the moderator (Aiken and
6

West 1991). Figure 2 shows that the average within-team

0.71**
0.69**
0.56**
0.39**
0.75**
0.48**
(0.92)

slope between POS and duty orientation was stronger with

higher (one SD above the mean; β = 0.56, p < 0.01) levels of
5

team-level ethical leadership and weaker with lower (one SD

below the mean; β = 0.20, p = 0.34) levels of ethical leader-
0.15*

0.14*
0.18*

0.14*
0.09

0.04
0.11

ship. Hence, Hypothesis 2 was supported. Strictly speaking,

–
4

a level-2 variable can moderate both the within-group and

between-group effects of level-1 variables. Although our
− 0.03

− 0.02
− 0.08
− 0.02
0.05
0.01

0.01
0.01
–

theoretical interest is in the hypothesized cross-level mod-

3
Table 2  Means, standard deviations, reliabilities, and correlations among the study variables

eration, we found no evidence of a between-group interac-

tion by ethical leadership.
− 0.14*
− 0.03

− 0.09
− 0.07
− 0.07
− 0.07
− 0.09
0.02

− 0.04

Hypothesis 3 predicted that the mediated relationship

–

between perceived organizational support and follower

2

OCBI, OCBO, and job performance through duty orientation

− 0.21**

0.32**

0.21**
0.25**

0.19**

is moderated by ethical leadership, in such a way that the

0.14*
0.13*
− 0.01

0.13
0.12

mediated relationships are stronger when ethical leadership

–

p < 0.05, **p < 0.01. Cronbach’s alpha coefficients on the diagonal

is higher. To test Hypothesis 3, we calculated the indirect

1

effects of POS on each of the outcome variables at higher (1

0.76
0.49
0.50
2.54

0.87
0.60
0.64
0.75
1.00

0.79
0.74

SD above the mean) and lower (1 SD below the mean) lev-

SD

els of the moderator, ethical leadership. These conditional

indirect effects are shown in Table 4. As shown in Table 4,
2.41
1.60
1.50
3.90

3.49
3.98
4.09
3.85
3.29

3.66
3.48

the conditional indirect effects of POS were stronger when

Education coded: 1 = college; 2 = postgraduate
Means based on individual-level data (N = 233)
M

there were higher rather than lower levels of ethical leader-

ship. Overall, these results provide support for Hypothesis 3.
Perceived organizational support (POS)

Gender coded: 1 = male; 2 = female

Age coded: 1 = 21–25 to 8 = 56–60

Discussion
Individual job performance
Affective commitment

Duty orientation (DO)

In the current study, we drew on social exchange theory to

Team tenure (years)

examine the role played by POS and ethical leadership in

Ethical leadership

enhancing employees’ OCBs and job performance through

Educationc

heightening their duty orientation. Our findings demonstrate

Variable

Genderb

that the high levels of duty orientation fostered through

OCBO
OCBI
Agea

POS lead followers to make a positive contribution to the

b

*
a

13
Antecedents of Duty Orientation and Follower Work Behavior: The Interactive Effects of Perceived… 635

Table 3  Results of the MSEM- Lower-level mediation Cross-level

moderated mediation analyses
DO OCBI OCBO Ind Job Per DO

Level-1
Age − 0.03 0.06 0.06 0.09 − 0.03
Gender − 0.03 − 0.02 0.03 − 0.08 − 0.02
Education − 0.04 − 0.07 − 0.04 0.03 − 0.04
Team tenure 0.07 − 0.07 − 0.02 0.01 0.06
Affective commitment 0.41** 0.14 0.43** 0.08 0.38**
Perceived organizational sup- 0.22** 0.01 0.17** 0.05 0.18*
port (POS)
Duty orientation – 0.44** 0.34** 0.57** –
Level-2
Ethical leadership 0.13
Cross-level interaction
POS × ethical leadership 0.38*

N = 233, 60 teams. Standardized coefficients reported

*
p < 0.05, **p < 0.01

Table 4  Conditional indirect effects of moderated mediation analyses in an ethical manner, and influences followers towards the
adoption of good ethical conduct through communicating
Indirect effect 95% (Monte Carlo)
confidence interval the need for employees to abide by the ethical codes of the
team and strengthening this through the use of rewards
OCBI and sanctions (Brown and Trevino 2006; Fehr et al. 2015),
− 1 SD below the mean − 0.09 − 0.29 to 0.10 followers will exhibit higher levels of duty orientation in
+ 1 SD above the mean 0.24* 0.06 to 0.44 response to support provided by the organization. This in
OCBO turn was found to lead them to perform at a higher level
− 1 SD below the mean − 0.07 − 0.22 to 0.08 and engage in extra-role behaviors that benefit both the
+ 1 SD above the mean 0.19* 0.05 to 0.36 organization and co-workers. In other words, our study
Individual job performance suggests that influence of POS on followers’ in-role and
− 1 SD below the mean − 0.11 − 0.36 to 0.13 extra-role performance through the mediating mechanism
+ 1 SD above the mean 0.32* 0.08 to 0.58 of duty orientation is contingent on the levels of ethical
Standardized estimates reported leadership present within the team.
*
p < 0.05 Specifically, in analyzing the slopes and the conditional
indirect effects at 1 standard deviation below the mean for
ethical leadership, there was no significant relationship
present between POS and followers’ in-role and extra-role
organization in the form of higher OCBI, OCBO, and job performance through duty orientation. This indicates that
performance. Furthermore, we found that the mediated rela- it may not be enough for organizations to provide support
tionships between POS and employee work behaviors were to their employees to increase duty orientation. Rather,
also moderated by ethical leadership, in such a way that the organizations must also ensure that the positive effects
positive relationships were stronger when there were high felt by employees are channeled into positive organiza-
rather than low levels of ethical leadership. tional behaviors through an organizational agent, such as
Our study contributes to the organizational behavior lit- an ethical leader. In fact, the investments that an organiza-
erature in the following ways. First, and most importantly, tion makes to support their employees might be wasted if
this study draws on social exchange theory (Blau 1964) employees are working underneath a manager who does
to highlight ethical leadership as a boundary condition not display ethical leadership. In order to better test this
which explains when employees are more likely to respond assertion, future research should examine this relationship
positively to POS in the form of increased duty orienta- in an experimental setting.
tion and in-role and extra-role behaviors at work. More Second, our study contributes to the literature on POS by
specifically, our findings suggest that when the leader acts examining the mediating effects of duty orientation on the

13
636 N. Eva et al.

5 as a moderator. This is surprising considering both the trans-

formational leadership (Shin and Zhou 2007) and servant
4 leadership (Hu and Liden 2011) literature studies have both
Duty Orientation

examined leadership as a contextual variable which may

3 strengthen the effects of follower characteristics or percep-
High Ethical Leadership
tions of the organizational environment on follower atti-
2 Low Ethical Leadership
tudes. We hope that our research spurs more interest in this
underdeveloped area.
1
Low High
Perceived Organizational Support
Practical Implications

The current research has a number of practical implications.

Fig. 2  Cross-level interaction of perceived organizational support and
ethical leadership on duty orientation First, our research reinforces that it is important for organi-
zations to create a climate where employees feel supported
in order to increase employees’ duty orientation and their
relationship between POS and employees’ work behaviors. subsequent in-role and extra-role performance. Organiza-
Drawing on the social exchange theory (Blau 1964), our tions may start by increasing the provision of visible sup-
results indicated that POS acts as an important antecedent port through the provision of flexible working arrangements,
to foster employees’ duty orientation and subsequently their mentoring schemes, childcare provision, and recreation
OCBs and job performance. When employees feel supported facilities. Rather than having a blanket approach, organiza-
by the organization, they develop a strong exchange rela- tions should look at what support their employees need and
tionship with their employer and seek to reciprocate in kind how best to target these programs to ensure that organiza-
the positive behaviors shown towards them. This extends tional resources are not wasted (Maertz et al. 2007).
previous work by Eisenberger et al. (2001) on duty orienta- Second, although research on ethical leadership has
tion which examined felt obligation as a mediator between typically stressed the role played by ethical leadership in
POS and employees’ work outcomes, as duty orientation reducing unethical behavior in organizational contexts, our
encompasses multiple obligations employees feel (mission, research also confirms that ethical leadership can be used to
members, and codes), rather than just their obligation to the increase employees in-role and extra-role behaviors through
organization. heightening the extent to which they reciprocate positive
Third, as duty orientation is a relatively new theoretical treatment from the organization. Specifically, there was evi-
concept, there are still major gaps in our knowledge about dence to indicate that the organizational investments made
its antecedents and outcomes. The present study builds on to support employees might be redundant if managers do not
recent work (Hannah et al. 2014) by examining POS as a display ethical leadership. Therefore, organizations should
new antecedent of duty orientation, and duty orientation as seek to identify individuals who are likely to exhibit ethi-
a mediating mechanism linking POS to the positive work cal leadership in their selection processes, and incorporate
behaviors of followers. Although Hannah et al. (2014) estab- ethical training into leadership development programs. For
lished a link between duty orientation and ethical outcomes, example, in selection processes vignettes could be used to
they did not examine its influence on positive work behaviors assess how potential leaders may respond to different ethical
that contribute to effective functioning of the organization. scenarios. Secondly, training programs within the organiza-
Fourth, our findings offer an insight into how ethical lead- tion can be used to develop ethical leadership amongst their
ership influences managerial-level employees. As the sample managers. This might be through in-house, HR delivered
was drawn from senior managers and their immediate sub- training, utilizing external ethical leadership trainers, or at
ordinates, i.e., mid-to low-level managers in their respective the very least providing online training modules on ethical
bureaus, our research demonstrates that ethical leadership leadership. Regardless of method, topics such as acting as an
from senior level managers fosters OCBs amongst low- to ethical role model, communicating an ethical message, and
mid-level managers directed towards their own followers. rewarding employees’ ethical behavior will be beneficial for
While we did not measure the trickle-down effects of ethical employees to incorporate into their programs (Mayer et al.
leadership (see Mayer et al. 2009), our results do offer some 2009).
support for Mayer et al.’s findings, i.e., that the display of
ethical leadership at the senior level, trickles down through
influencing the behaviors of middle-level managers.
Finally, it should be noted that there is no published
research to our knowledge that examines ethical leadership

13
Antecedents of Duty Orientation and Follower Work Behavior: The Interactive Effects of Perceived… 637

Limitations and Avenues for Future Research Conclusion

As with any research, the present study has a number of The present study builds on growing research on ethical
limitations. First, the present study has methodological limi- leadership by understanding how it channels followers’
tations that may be addressed in future work. As we did not responses to positive treatment from the organization into
collect data on all variables at all points in time, we are una- a dutiful mindset, resulting in in-role and extra-role per-
ble to conclusively determine causal relationships between formance. More specifically, it examined the influence of
the study variables. To strengthen causality, researchers perceived organizational support on both followers’ job per-
should examine the variables in a cross-lagged panel design formance and organizational citizenship behaviors, and the
or consider experimental designs. For example, one team mediating effects of duty orientation on such relationships.
should be exposed to a condition with a leader who displays In addition, it examines whether the mediated effects are
high levels of ethical leadership and the control group should contingent on the ethical leadership exhibited by the team
be exposed to a leader who displays low levels of ethical leader. Using multi-source, multi-level data from manage-
leadership. Panel and experimental designs would also ease rial teams in the Chinese public sector, duty orientation was
concerns with reverse causality where an employee’s duty found to mediate the relationships between POS and work
orientation may influence their perceptions of the support outcomes, namely, OCBI, OCBO, and job performance. Fur-
provided by the organization due to their duty to the mis- ther, the relationship between POS and work outcomes was
sion, and ethical leadership of their manager through their found to be stronger when employees were working under an
duty to members. Further, researchers should also consider ethical leader. Our findings suggest that in order to improve
incorporating more objective measures of job performance employees’ performance and OCBs, it is critical for organi-
and OCBs from company records rather than relying on zations to provide adequate support for employees while
supervisor-rated measures. identifying and developing ethical leaders.
Second, the generalizability of our results across cultures,
organizations, and levels of the organization may be brought Acknowledgements This research was funded by the Natural Science
Foundation of China (No. 71672174 & R17G020002).
into question as we utilized a sample of senior leaders from
government agencies in a single Chinese Province. In order
to overcome this limitation, we encourage future research to
Compliance with Ethical Standards
replicate our study in different cultural contexts, organiza- Ethical Approval All procedures performed in studies involving human
tional settings, and at lower levels of the organization. participants were in accordance with the ethical standards of the insti-
Third, while we controlled for affective organizational tutional and/or national research committee and with the 1964 Helsinki
commitment in our model, we suggest future research con- Declaration and its later amendments or comparable ethical standards.
trol for other possible variables that may mediate the influ- Informed Consent Informed consent was obtained from all individual
ence of POS on work outcomes including felt obligation participants included in the study.
to the organization (Eisenberger et al. 2001), and organiza-
tional identification (Kurtessis et al. 2017). This will allow
us to determine the relative explanatory power of duty ori-
entation vis-à-vis other mechanisms which may explain the References
social exchange process of POS.
Finally, although the present study has started to unpack Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and
interpreting results. Newbury Park, CA: Sage.
potential boundary conditions of the effects of POS on fol-
Allen, N. J., & Meyer, J. P. (1990). The measurement and anteced-
lower work outcomes through duty orientation by examining ents of affective, continuance and normative commitment to the
ethical leadership, future research may examine other pos- organization. Journal of Occupational and Organizational Psy-
sible moderators. For example, empirical work may exam- chology, 63(1), 1–18.
Avey, J., Palanski, M., & Walumbwa, F. (2011). When leadership
ine whether factors such as ethical climate and employees’
goes unnoticed: The moderating role of follower self-esteem on
own personal moral values also strengthen the effects of the relationship between ethical leadership and follower behav-
POS on the duty orientation and positive work behaviors ior. Journal of Business Ethics, 98(4), 573–582. https​://doi.
of followers. org/10.1007/s1055​1-010-0610-2.
Beauregard, T. A. (2012). Perfectionism, self-efficacy and OCB: The
moderating role of gender. Personnel Review, 41(5), 590–608.
Bedi, A., Alpaslan, C. M., & Green, S. (2016). A meta-analytic review
of ethical leadership outcomes and moderators. Journal of Busi-
ness Ethics, 139(3), 517–536.
Blau, P. M. (1964). Exchange and power in social life. New York:
Wiley.

13
638 N. Eva et al.

Brislin, R. (2000). Understanding culture’s influence on behavior (2). Kacmar, K. M., Bachrach, D. G., Harris, K. J., & Zivnuska, S. (2011).
Belmont, CA: Wadsworth Publishing. Fostering good citizenship through ethical leadership: Exploring
Brown, M. E., & Treviño, L. K. (2006). Ethical leadership: A review the moderating role of gender and organizational politics. Journal
and future directions. The Leadership Quarterly, 17(6), 595–616. of Applied Psychology, 96(3), 633–642. https​://doi.org/10.1037/
https​://doi.org/10.1016/j.leaqu​a.2006.10.004. a0021​872.
Brown, M. E., Treviño, L. K., & Harrison, D. A. (2005). Ethical leader- Kalshoven, K., Hartog, D., & De Hoogh, A. H. B. (2011). Ethical
ship: A social learning perspective for construct development and leadership at work questionnaire (ELW): Development and valida-
testing. Organizational Behavior and Human Decision Processes, tion of a multidimensional measure. Leadership Quarterly, 22(1),
97(2), 117–134. https​://doi.org/10.1016/j.obhdp​.2005.03.002. 51–69.
Carboni, I., & Ehrlich, K. (2013). The effect of relational and team Kaufman, J. D., Stamper, C. L., & Tesluk, P. E. (2001). Do supportive
characteristics on individual performance: A social network per- organizations make for good corporate citizens? Journal of Mana-
spective. Human Resource Management, 52(4), 511–535. gerial Issues, 13(4), 436–449.
Den Hartog, D. (2015). Ethical leadership. Annual Review of Organi- Kline, R. B. (2010). Principles and practice of structural equation
zational Psychology and Organizational Behavior, 2(1), 409–434. modeling (3). New York: Guilford Press.
Djurkovic, N., McCormack, D., & Casimir, G. (2008). Workplace bul- Ko, C., Ma, J., Bartnik, R., Haney, M. H., & Kang, M. (2017). Ethical
lying and intention to leave: The moderating effect of perceived leadership: An integrative review and future research agenda, Eth-
organizational support. Human Resource Management Journal, ics & Behavior, https:​ //doi.org/10.1080/105084​ 22.2017.131806​ 9.
18(4), 405–422. https​://doi.org/10.1111/j.1748-8583.2008.00081​ Kuehn, K. W., & Al-Busaidi, Y. (2002). Citizenship behavior in a
.x. non-western context: An examination of the role of satisfaction,
Edwards, M. R. (2009). HR, perceived organizational support and commitment and job characteristics on self-reported OCB. Inter-
organizational identification: An analysis after organizational for- national Journal of Commerce & Management, 12(2), 107–125.
mation. Human Resource Management Journal, 19(1), 91–115. Kurtessis, J. N., Eisenberger, R., Ford, M. T., Buffardi, L. C., Stewart,
https​://doi.org/10.1111/j.1748-8583.2008.00083​.x. K. A., & Adis, C. S. (2017). Perceived organizational support: A
Ehigie, B. O., & Otukoya, O. W. (2005). Antecedents of organizational meta-analytic evaluation of organizational support theory. Journal
citizenship behaviour in a government-owned enterprise in Nige- of Management, 43(6), 1854–1884.
ria. European Journal of Work and Organizational Psychology, Lam, S. S., Chen, X. P., & Schaubroeck, J. (2002). Participative deci-
14(4), 389–399. sion making and employee performance in different cultures:
Eisenberger, R., Armeli, S., Rexwinkel, B., Lynch, P. D., & Rhoades, The moderating effects of allocentrism/idiocentrism and efficacy.
L. (2001). Reciprocation of perceived organizational sup- Academy of Management Journal, 45(5), 905–914.
port. Journal of Applied Psychology, 86(1), 42–51. https​://doi. LeBreton, J. M., & Senter, J. L. (2008). Answers to 20 questions about
org/10.1037//0021-9010.86.1.42. interrater reliability and interrater agreement. Organizational
Eisenberger, R., Huntington, R., Hutchison, S., & Sowa, D. (1986). Research Methods, 11(4), 815–852.
Perceived organizational support. Journal of Applied Psychology, Lee, K., & Allen, N. J. (2002). Organizational citizenship behavior and
71(3), 500–507. workplace deviance: The role of affect and cognitions. Journal of
Fehr, R., Yam, K. C., & Dang, C. (2015). Moralized leadership: The Applied Psychology, 87(1), 131–142.
construction and consequences of ethical leader perceptions. Lindell, M. K., & Whitney, D. J. (2001). Accounting for common
Academy of Management Review, 40(2), 182–209. https​://doi. method variance in cross-sectional research designs. Journal of
org/10.5465/amr.2013.0358. Applied Psychology, 86(1), 114–121.
Ferris, G. R., Yates, V. L., Gilmore, D. C., & Rowland, K. M. (1985). Maertz, C. P., Griffeth, R. W., Campbell, N. S., & Allen, D. G. (2007).
The influence of subordinate age on performance ratings and The effects of perceived organizational support and perceived
causal attributions. Personnel Psychology, 38(3), 545–557. supervisor support on employee turnover. Journal of Organiza-
Gong, Y., Chang, S., & Cheung, S. Y. (2010). High performance work tional Behavior, 28(8), 1059–1075.
system and collective OCB: A collective social exchange per- Mayer, D., Kuenzi, M., & Greenbaum, R. (2010). Examining the link
spective. Human Resource Management Journal, 20(2), 119–137. between ethical leadership and employee misconduct: The mediat-
Gouldner, A. W. (1960). The norm of reciprocity: A preliminary state- ing role of ethical climate. Journal of Business Ethics, 95, 7–16.
ment. American Sociological Review, 25(2), 161–178. https​://doi.org/10.1007/s1055​1-011-0794-0.
Graham, J. W. (1991). An essay on organizational citizenship behavior. Mayer, D. M., Aquino, K., Greenbaum, R., & Kuenzi, M. (2012). Who
Employee Responsibilities and Rights Journal, 4(4), 249–270. displays ethical leadership, and why does it matter? An examina-
Hair, J. F., Black, W. C., Babin, B. J., & Anderson, R. E. (2010). Mul- tion of antecedents and consequences of ethical leadership. Acad-
tivariate data analysis (7). Upper Saddle River, NJ: Pearson Pren- emy of Management Journal, 55(1), 151–171.
tice Hall. Mayer, D. M., Kuenzi, M., Greenbaum, R., Bardes, M., & Salva-
Hannah, S. T., Jennings, P. L., Bluhm, D., Peng, A. C., & Schaubroeck, dor, R. (2009). How low does ethical leadership flow? Test of a
J. M. (2014). Duty orientation: Theoretical development and pre- trickle-down model. Organizational Behavior & Human Deci-
liminary construct testing. Organizational Behavior and Human sion Processes, 108(1), 1–13. https​://doi.org/10.1016/j.obhdp​
Decision Processes, 123(2), 220–238. https​://doi.org/10.1016/j. .2008.04.002.
obhdp​.2013.10.007. Moynihan, D. P., Wright, B., & Pandey, S. (2012). Setting the table:
Heilman, M. E., Block, C. J., & Lucas, J. A. (1992). Presumed incom- How transformational leadership fosters performance informa-
petent? Stigmatization and affirmative action efforts. Journal of tion use. Journal of Public Administration Research and Theory,
Applied Psychology, 77(4), 536–544. 22(1), 143–164.
Hu, J., & Liden, R. (2011). Antecedents of team potency and team Muthén, L. K., & Muthén, B. O. (2012). Mplus user’s guide (7 edn).
effectiveness: An examination of goal and process clarity and Los Angeles, CA: Muthén & Muthén.
servant leadership. Journal of Applied Psychology, 96(4), Neves, P., & Story, J. (2015). Ethical leadership and reputation: Com-
851–862. bined indirect effects on organizational deviance. Journal of Busi-
James, L. R., Demaree, R. G., & Wolf, G. (1984). Estimating within- ness Ethics, 127(1), 165–176.
group interrater reliability with and without response bias. Jour- Newman, A., Kiazad, K., Miao, Q., & Cooper, B. (2014). Examining
nal of Applied Psychology, 69(1), 85–98. the cognitive and affective trust-based mechanisms underlying

13
Antecedents of Duty Orientation and Follower Work Behavior: The Interactive Effects of Perceived… 639

the relationship between ethical leadership and organizational Roth, P. L., Purvis, K. L., & Bobko, P. (2012). A meta-analysis of
citizenship: A case of the head leading the heart? Journal of gender group differences for measures of job performance in field
Business Ethics, 123(1), 113–123. https​://doi.org/10.1007/s1055​ studies. Journal of Management, 38(2), 719–739.
1-013-1803-2. Shin, S. J., & Zhou, J. (2007). When is educational specialization
Ng, T. W., & Feldman, D. C. (2015). Ethical leadership: Meta-analytic heterogeneity related to creativity in research and development
evidence of criterion-related and incremental validity. Journal of teams? Transformational leadership as a moderator. Journal of
Applied Psychology, 100(3), 948. Applied Psychology, 92(6), 1709–1721.
Ng, T. W. H., & Feldman, D. C. (2009). How broadly does educa- Siemsen, E., Roth, A., & Oliveira, P. (2010). Common method bias in
tion contribute to job performance? Personnel Psychology, 62(1), regression models with linear, quadratic, and interaction effects.
89–134. Organizational Research Methods, 13(3), 456–476.
Podsakoff, P. M., MacKenzie, S. B., Lee, J. Y., & Podsakoff, N. P. Tabachnick, B. G., & Fidell, L. S. (2013). Using multivariate statistics
(2003). Common method biases in behavioral research: a critical (6 edn). Boston, MA: Pearson.
review of the literature and recommended remedies. Journal of Walumbwa, F. O., Mayer, D. M., Wang, P., Wang, H., Workman, K., &
Applied Psychology, 88(5), 879–903. Christensen, A. L. (2011). Linking ethical leadership to employee
Podsakoff, P. M., MacKenzie, S. B., & Podsakoff, N. P. (2012). Sources performance: The rules of leader–member exchange, self-efficacy,
of method bias in social science research and recommendations on and organizational identification. Organizational Behavior and
how to control it. Annual Review of Psychology, 63(1), 539–569. Human Decision Processes, 115(2), 204–213.
Preacher, K. J., & Selig, J. P. (2012). Advantages of Monte Carlo con- Wayne, S. J., Shore, L. M., & Liden, R. C. (1997). Perceived organiza-
fidence intervals for indirect effects. Communication Methods and tional support and leader–member exchange: A social exchange
Measures, 6(2), 77–98. perspective. The Academy of Management Journal, 40(1), 82–111.
Preacher, K. J., Zhang, Z., & Zyphur, M. J. (2016). Multilevel structural Wong, Y. T., Wong, C. S., & Ngo, H. Y. (2012). The effects of trust
equation models for assessing moderation within and across levels in organization and perceived organizational support on organi-
of analysis. Psychological Methods, 21, 189–205. zational citizenship behavior: A test of three competing models.
Preacher, K. J., Zyphur, M. J., & Zhang, Z. (2010). A general multilevel International Journal of Human Resource Management, 23(2),
SEM framework for assessing multilevel mediation. Psychologi- 278–293. https​://doi.org/10.1080/09585​192.2011.61096​6.
cal Methods, 15, 209–233. Yu, K., Lin, W., Wang, L., Ma, J., Wei, W., Wang, H.,…Shi, J. (2016).
Rhoades, L., & Eisenberger, R. (2002). Perceived organizational sup- The role of affective commitment and future work self salience in
port: A review of the literature. Journal of Applied Psychology, the abusive supervision-job performance relationship. Journal of
87(4), 698–714. Occupational & Organizational Psychology, 89(1), 28–45.

13