Knowledge Management Research & Practice

ISSN: (Print) (Online) Journal homepage: https://www.tandfonline.com/loi/tkmr20

The effect of job crafting to job performance

Chunyu Zhang & Liping Liu

To cite this article: Chunyu Zhang & Liping Liu (2021) The effect of job crafting to
job performance, Knowledge Management Research & Practice, 19:2, 253-262, DOI:
10.1080/14778238.2020.1762517

To link to this article: https://doi.org/10.1080/14778238.2020.1762517

Published online: 13 May 2020.

Submit your article to this journal

Article views: 670

View related articles

View Crossmark data

Citing articles: 1 View citing articles

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=tkmr20
KNOWLEDGE MANAGEMENT RESEARCH & PRACTICE
2021, VOL. 19, NO. 2, 253–262
https://doi.org/10.1080/14778238.2020.1762517

ORIGINAL ARTICLE

The effect of job crafting to job performance

Chunyu Zhang and Liping Liu
National Institute of Development Administration, International College, Bangkok, Thailand

ABSTRACT ARTICLE HISTORY

This research studies the mediating effect of employee’s social capital on job crafting and job Received 20 April 2019
performance, and the moderating effect of core self-evaluation on social capital and job perfor- Accepted 26 April 2020
mance. This research will help Guangxi enterprises, against the background of “One Belt One KEYWORDS
Road”, to improve their competitiveness through rational human resource management. Job Crafting; social Capital;
A questionnaire survey was conducted, using the convenience sampling method, on 438 core Self-evaluation; job
employees in Guangxi enterprises, and data analysis was performed using Amos, Smart-PLS Performance; social
and SPSS. The results of the hierarchical linear regression analysis show that social capital partly Exchange Theory; social
mediates the effect of job crafting on job performance, and core self-evaluation moderates the Cognition Theory
effect of social capital on job performance. However, the impact of social capital on job
performance cannot be tested by SEM.

1. Introduction exchange increases, employees’ social capital will

increase. The impact of core self-evaluation on job per-
In the context of China’s “One Belt One Road” initiative,
formance has been confirmed (Bhargava & Pradhan,
Guangxi, an important radiating node for southeast Asia
2019). According to social cognition theory, employees’
(Yu, 2017), provides important support for local enter-
behavioural factors, environmental factors, and personal
prises, but, at the same time, competition among enter-
factors are closely related (Bandura & Walters, 1977);
prises has intensified. Knowledge sharing can improve
environmental factors include the interaction between
the competitiveness of enterprises by improving custo-
people, and personal factors include human characteris-
mer satisfaction, financial performance (Chen et al.,
tics (Lu, 2012). Therefore, this study will explore the
2018) and innovation. In Chinese society, internal social
relationship between job crafting, social capital, core self-
capital increases the knowledge sharing of corporate
evaluation, and job performance.
employees (Yen et al., 2015). Mao and Li (2015) pro-
The main contributions of this research are two-
posed that employees’ proactive behaviour for the pur-
fold. This study conducts empirical research by col-
pose of relationship building was helpful in improving
lecting data from Chinese company employees. First,
their social capital. Strength-based job crafting is pro-
this study examines the relationship between the three
social behaviour (Tian & Liu, 2017), and can lead to the
dimensions of job crafting (i.e., cognition, relationship
accumulation of social capital (Bolino et al., 2002). The
and task) and social capital. Second, it examines the
notion of a social network is taken from job design, and
relationship between core self-assessment, social capi-
the relationship between job crafting and social capital is
tal and job performance.
worth exploring (Qi et al., 2014). However, there has not
yet been any empirical research on whether the reshap-
ing of work has an impact on social capital. Job crafting 2. Literature review
has an indirect (Tims et al., 2013) and a direct (Tims
2.1. The relationship between job crafting and job
et al., 2015, 2013) impact on job performance, and its
performance
indirect impact is that job crafting affects job perfor-
mance through job involvement. (Tims et al., 2013). Job crafting is defined as an employee making changes
Applying social exchange theory (Blau, 1960), this in his or her cognition, relationships and tasks of
shows that job crafting employees have the advantage work. Cognition refers to how the individual views
of exchanging high-value resources with their colleagues work, while changes in relationships and tasks refer
and that they are good at building mutually beneficial to changes in form, scope or quantity (Bakker et al.,
network relationships with other members of the team 2012). Employees actively improve their work content
(L. Zhang et al., 2016). Job crafting employees actively according to their job needs, skills and preferences,
help their colleagues, and their colleagues will often gen- and this is what constitutes job crafting (Tims et al.,
erate returns. As the degree of perception of the social 2013). Wrzesniewski and Dutton (2001) argue that

CONTACT Chunyu Zhang [email protected] National Institute of Development Administration, International College, Bangkok, Thailand
© Operational Research Society 2019.
254 C. ZHANG AND L. LIU

there are three types of job crafting, namely, task 2.2. The mediating role of social capital
crafting (changing the type or amount of work
At the individual level, social capital is the sum of the
done), relational crafting (being cautious about
actual resources and the potential resources contained
whom you work with at work), and cognitive crafting
in the network of personal relationships owned by an
(changing your views on work tasks in order to make
individual (Nahapiet & Ghoshal, 1998). The social
them more meaningful to you). The job crafting scale
capital scale of employees in a virtual organisation, as
developed by Slemp and Vella-Brodrick (2013) made
compiled by Chumg et al. (2015) in the context of
up for the deficiencies of the previous scholars and has
Chinese culture, examined the social network connec-
good reliability and validity. On the basis of empirical
tions, trust and goals of employees. This scale has
research, it has been concluded that job crafting has
a particular applicability to Chinese employees.
a positive correlation with the use of strengths, intrin-
Strength-based job crafting is pro-social behaviour
sic goal-setting, organisational citizenship behaviour,
(Tian & Liu, 2017), and can lead to the accumulation
job satisfaction, work motivation and work enthu-
of social capital (Bolino et al., 2002). Job crafting
siasm (Slemp & Vella-Brodrick, 2013). The job per-
employees have the advantage of exchanging high-
formance scale developed by Williams and Anderson
value resources with colleagues, and they are good at
(1991) is composed of an in-role behaviour and an
building mutually beneficial network relationships
organisational citizenship behaviour subscale, and is
with other members of the team (L. Zhang et al.,
evaluated by supervisors; however, self-assessment has
2016). Applying this in combination with social
also been adopted in later studies (Staufenbiel &
exchange theory (Blau, 1960), job crafting employees
König, 2010).
actively help their colleagues, and their colleagues will
The impact of job crafting on job performance has
often generate returns. As the degree of perception of
been demonstrated many times. However, the
the social exchange increases, employees’ social capital
research on job crafting adopted four dimensions in
increases.
order to explore its impact on job performance:
The positive effect of social capital on employee
increasing structural work resources, increasing social
knowledge sharing has been confirmed (Hau et al.,
employment resources, increasing specific challenging
2013; Zimmermann et al., 2018). Similarly, social
work requirements and reducing obstructive work
capital (i.e., social interaction in the cognitive dimen-
demands (Tims et al., 2015). Crafting job resources
sion) has been shown to have a positive and signifi-
and challenging job demands of job crafting have
cant impact on knowledge sharing performance
a mediating influence on in-role performance through
(Lefebvre et al., 2016). Relational and cognitive social
work engagement (Bakker et al., 2012). Thus, accord-
capital has a positive and significant impact on rou-
ing to social exchange theory, when employees take
tine job performance (Ali-Hassan et al., 2015).
the initiative to increase the number of their tasks,
Moreover, Zimmermann et al. (2018) have shown
they will perform better in return. Employees select
that the intensity of social capital, shared understand-
people to be their friends if those people have the same
ing and trust mediates the impact of exclusive pro-
skills, and they are supported by their friends when
curement arrangements on the degree of knowledge
their own skills are insufficient, so that they complete
sharing. According to Zhang and Chen (2017), social
their work tasks better. When employees change their
capital plays a partial mediating role in the impact of
views on tasks, a positive attitude will enable them to
the three dimensions of Zhong-yong thinking on the
enjoy their work more, thus achieving better perfor-
survival ability of employees. Zhong-yong thinking is
mance. Job crafting affects in-role performance indir-
an expression of Chinese Confucianism. Job crafting
ectly through job engagement, but job engagement is
indirectly affects in-role performance through job
not the only intermediary for the impact of job craft-
engagement, but job engagement is not the only med-
ing on performance (Bakker et al., 2012).
iator of the impact of job crafting on performance
This research suggests the following hypotheses:
(Bakker et al., 2012).
H1: Job crafting positively affects job performance.
H2: Job crafting positively affects social capital.
H1a: Cognitive crafting positively affects job
performance. H2a: Cognitive crafting positively affects social capital.

H1b: Relational crafting positively affects job H2b: Relational crafting positively affects social
performance. capital.

H1 c: Task crafting positively affects job performance. H2 c: Task crafting positively affects social capital.
 KNOWLEDGE MANAGEMENT RESEARCH & PRACTICE 255

H3: Social capital positively affects job performance. by individualistic cultures, but the Chinese collectivist
culture values working together (Liu et al., 2015).
H4: Social capital has a mediating role in the impact of Based on social cognition theory (Bandura &
job crafting on job performance. Walters, 1977), employees with high core self-
evaluation will slow down the impact of social capital
H4a: Social capital has a mediating role in the impact on job performance. It is worth noting that the basic
of cognitive crafting on job performance. characteristics of core self-evaluation include four
aspects, namely, self-esteem, generalised self-efficacy,
H4b: Social capital has a mediating role in the impact emotional stability and control points. However, the
of relational crafting on job performance. above scholars have adopted the single-dimensional
core self-evaluation scale of Judge et al. (2003). This
H4 c: Social capital has a mediating role in the impact study proposes the following hypothesis:
of task crafting on job performance.
H5: Core self-evaluation negatively moderates the
effect of social capital on job performance.
2.3. The moderating role of core self-evaluation
Core self-evaluation is defined as an individual’s eva-
luation of himself or herself, as well as the evaluation 3. Method
of the relationship between himself or herself and the 3.1. Research framework
environment (Judge et al., 2003). Through a literature
review, Wang et al. (2018) proposed that core self- Based on social exchange theory and social cognition
evaluation has a direct effect on individual structural theory, the relationships between job crafting, social
capital and relational capital. Core self-evaluation capital, core self-evaluation and job performance are
positively affects both task performance and contex- established. The research framework of this study is
tual performance (Chen et al., 2016). Nevertheless, shown in Figure 1.
Kacmar et al. (2009) revealed that core self-
evaluation was not a significant predictor of job per-
3.2. Measures
formance. According to hypothesis H3, social capital
is a significant predictor of job performance. Both core The job crafting scale utilised in this study is that of
self-evaluation and social capital are important indivi- (Slemp & Vella-Brodrick, 2013). In this scale, job
dual resources (Creed & Gagliardi, 2015). Core self- crafting is divided into three dimensions, namely,
evaluation positively moderates the impact of work cognitive crafting, relational crafting, and task craft-
value consistency on job performance. The mechan- ing, with a total of 15 items. An example of these items
ism here is that work value consistency has a positive is: “Introduce new approaches to improve your work”.
impact on job performance through organisational The social capital scale utilised in this study is that of
identity without violating personal values (Bhargava Chumg et al. (2015). In this scale, social capital is
& Pradhan, 2017). Core self-evaluation positively divided into three dimensions, namely, the cognitive
affects the influence of family support for instrumental dimension, the relational dimension, and the structural
organisation on job burnout, and negatively affects the dimension, with a total of nine items, and the scale is
influence of family support for emotional organisation modified according to the needs of this research. An
on job burnout (P. Wang et al., 2016). In addition, example of these items is: “I have a good relationship
core self-evaluation negatively moderates the relation- with my colleagues in the company”.
ship between family incivility and psychological dis- The core self-evaluation scale utilised in this study
tress (Lim & Tai, 2014). Core self-evaluation is is that of Judge et al. (2003). In this scale, core self-
considered to be consistent with the values recognised evaluation is a single dimension, measured using

Core Self-evaluation

H5
Job Crafting H4
H2
Cognitive Crafting Social Capital Job Performance
Relational Crafting H3
Task Crafting
H1

Figure 1. Research framework.

256 C. ZHANG AND L. LIU

a total of 12 items. An example of these items is: 4. Results

“When I try, I generally succeed”.
4.1. Common method variance
The job performance scale utilised in this study is
the in-role behaviour scale of Williams and Anderson In this study, psychological isolation was adopted to
(1991), and a self-evaluation method is adopted prevent common method variance (CMV). Each scale
(Staufenbiel & König, 2010). The scale has five items, in the questionnaire was an independent part
and an example of these items is: “I adequately com- (Podsakoff et al., 2003).
plete assigned duties”. Harman’s single factor test for CMV was adopted
Referring to previous research (Li & Lin, 2018), the to extract four factors. The largest factor explained
control variables include gender, age, education level, 48.32%. Hence the results do not give cause for con-
monthly salary, position and business category. cern (Podsakoff et al., 2003).

3.3. Participants and procedure 4.2. Reliability analysis

Convenience sampling is adopted in this study, and the Reliability analysis showed that all the scales had good
research objects are employees of enterprises in reliability. For job crafting, Cronbach’s alpha was 0.94,
Guangxi. Guangxi is an important radiating node for for social capital it was 0.92, for social capital it was
“One Belt One Road” in southeast Asia (Yu, 2017). The 0.92, for core self-evaluation it was 0.94, and for job
process for distributing the questionnaire was through performance it was 0.82.
the researcher’s own social relations, who entrusted local
friends to coordinate the answers of their colleagues.
4.3. Confirmatory factor analysis
The questionnaire was compiled on “question-
naire star (www.wjx.cn)”, through the “WeChat” for- Amos 23 was utilised in this study to conduct confirma-
ward link communication software. A total of 500 tory factor analysis on the model to verify whether the
questionnaires were sent out, with the largest number model fit was good. The results show that
of responses to the questionnaire being received on CMIN = 1492.020, DF = 764, CMIN/DF = 1.953,
the first day, with a diminishing number after that. In p < 0.001 (i.e., CMIN/DF < 3) (Medsker et al., 1994).
order to save time, no more responses were accepted However, because of the sample size, it cannot be directly
after the seventh day, when a total of 438 responses used to reflect the degree of a theoretical model and
had been received, giving an effective response rate of observation data. Therefore, the model was evaluated
87.6%. In the sample, 68.7% were male, with the with a composite index of overall fitness, namely, the
highest proportion falling in the age range of comparative fit index (CFI) being greater than 0.9,
25–34 years (47.3%); in terms of education level, the Tucker-Lewis Index (TLI) being greater than 0.9,
the highest proportion had a bachelor’s degree and the root mean square error of approximation
(46.8%); in terms of working tenure, the highest (RMSEA) and standardised root mean squared residual
proportion had a tenure of 3–6 years (34.5%); in (SRMR) being less than 0.08 (Kline, 2005; Marsh et al.,
terms of monthly salary, the highest proportion 2004). The results show that CFI = 0.936, TLI = 0.931,
earned CNY 5,001–10,000 (45.4%); in terms of job RMSEA = 0.047, SRMR = 0.0420, indicating a good fit.
position, the largest proportion were ordinary
employees (34%); and in terms of business category,
4.4. Convergent validity
the highest proportion worked in education (36.8%).
A structural equation model (SEM) was used to ana- In this study, comprehensive reliability (CR) and aver-
lyse the data. The ratio of the number in the sample age variance extraction (AVE) were used to test the
and the number of formal questions in the question- convergent validity of the variables.
naire was 10:1 (Hair et al., 1998). There were 41 From Table 1, the factor loading for all the items
formal questions in the study, which meets the ranges from 0.631 to 0.813, which is above the bench-
requirements for SEM analysis. mark value of 0.5 (Bagozzi & Yi, 1988). The CR value
The respondents in this study were employees of of each variable is higher than the criterion value of 0.6
Chinese enterprises, and the scales used were all in (Bagozzi & Yi, 1988; Fornell & Larcker, 1981). The
English. Therefore, all the questionnaires were trans-
lated into Chinese so that the respondents could answer
Table 1. Convergent validity index.
them. Doctoral students in management (English pro- variables Items Factor Loading CR AVE
gramme) whose mother tongue was Chinese were Cognitive crafting 5 0.631–0.752 0.850 0.514
invited to check the questionnaire to ensure the accu- Relational crafting 5 0.716–0.778 0.854 0.539
Task crafting 5 0.74–0.813 0.880 0.596
racy of the translation. The possible answers to the Social capital 9 0.638–0.779 0.908 0.524
questions ranged from “entirely disagree” to “entirely Core self-evaluation 12 0.715–0.79 0.938 0.558
agree” (i.e., 1 to 5) on a 5-point Likert scale. Job performance 5 0.64–0.752 0.837 0.507
 KNOWLEDGE MANAGEMENT RESEARCH & PRACTICE 257

AVE scores are all greater than 0.5 (Fornell & Larcker, Table 3. Structural model results.
1981). To sum up, all the variables in this study have Proposal
Hypotheses effect Beta t value Results
convergent validity.
H1a Cognitive crafting → + 0.282*** 4.133 Supported
Job performance
H1b Relational crafting → + 0.167* 2.559 Supported
4.5. Correlation and discriminant validity Job performance
H1 c Task crafting → Job + 0.308*** 4.557 Supported
This research adopts the structural equation model, performance
H2a Cognitive crafting → + 0.208*** 3.975 Supported
the comparison of chi-square values and the potential Social capital
variable pairing associated confidence interval detec- H2b Relational crafting → + 0.273*** 4.451 Supported
Social capital
tion method to test potential variable discriminant H2 c Task crafting → Social + 0.342*** 4.831 Supported
validity (Anderson & Gerbing, 1988). capital
H3 Social capital → Job + 0.098 1.461 Not
Based on Table 2, the chi-square value compari- performance
sons show significant differences (p < 0.001), indicat- Supported
*p < 0.05; **p < 0.01; *** p < 0.001.
ing that there are differences between the models
with the correlation between the two factors set at
1.00 and the models with the free estimation of the (p < 0.001), respectively. These results show that task
correlation between all potential variables. The crafting has the greatest impact on job performance,
model with a correlation of 1.00 and the model with and relationship crafting has the smallest impact on
the free estimation of the correlation cannot be job performance.
regarded as equivalent, so the correlation between Hypotheses H2a, H2b and H2 c propose that job
the potential variables is distinguishable, and discri- crafting (cognitive, relational, task) has a positive
minant validity is supported. Secondly, none of the effect on social capital. As shown in Table 3, the
confidence intervals in the brackets covers 1.00. The effects of these factors on social capital have values
results show that the correlation between potential of 0.208 (p < 0.001), 0.273(p < 0.001) and 0.342
variables is distinguishable and the variables have (p < 0.001), respectively. These results show that
discriminant validity. task crafting has the greatest impact on job perfor-
mance and cognitive crafting has the smallest impact
4.6. Testing of hypotheses on job performance.
Social capital does not have a positive impact on job
To test whether hypotheses H1 to H5 are supported, performance (β = 0.098, p > 0.05), so hypothesis H3 in
this study used Smart PLS 3 and SPSS 25 to examine this study is not supported.
the relationship between job crafting, social capital, To assess the mediating effect of social capital in the
core self-evaluation and job performance. relationship between job crafting and job perfor-
mance, further analyses were performed to confirm
4.6.1. Structural model the magnitude and statistical significance of the indir-
Hypotheses H1a, H1b and H1 c propose that job ect effects. Specifically, the bootstrap confidence inter-
crafting (cognitive, relational, task) has a positive vals method with 2,000 iterations, according to
effect on job performance. As shown in Table 3, the Preacher and Hayes (2008), was implemented to test
effects of these factors on job performance have values the significance of the indirect effects. The results are
of 0.282 (p < 0.001), 0.167 (p < 0.05) and 0.308 presented in Table 4.

Table 2. Correlation and discriminant validity.

Cognitive crafting Relational crafting Task crafting Social capital Core self-evaluation Job performance
Cognitive crafting 1
Relational crafting 0.931*** 1
(132.262***)
[0.994, 0.868]
Task crafting 0.928*** 0.925*** 1
(142.418***) (119.331***)
[0.987, 0.869] [0.990, 0.860]
Social capital 0.793*** 0.804*** 0.814*** 1
(130.495***) (106.510***) (113.121 ***)
[0.856, 0.730] [0.875, 0.733] [0.883, 0.745]
Core self-evaluation 0.822*** 0.851*** 0.887*** 0.776*** 1
(150.767 ***) (122.017 ***) (124.686 ***) (110.812***)
[0.879, 0.765] [0.916, 0.786] [0.950, 0.824] [0.845, 0.707]
Job performance 0.872*** 0.833*** 0.852*** 0.722*** 0.825*** 1
(154.944 ***) (137.365 ***) (143.372 ***) (131.504***) (138.863***)
[0.929, 0.815] [0.894, 0.772] [0.911, 0.793] [0.785, 0.659] [0.885, 0.767]
*p <.05, **p < .01, ***p < .001
Note: () chi-square comparisons; [] confidence interval
258 C. ZHANG AND L. LIU

Table 4. Indirect and total effects analysis. performance (β = −0.249, p < 0.001). This research
Confidence intervals bias result is different from Ng and Feldman (2009). The
corrected
research results are different from Ng and Feldman
Lower confi- Upper confi-
Path Beta p value dence level dence level (2009). However, consistent with the research results
Indirect CC → SC 0.020 0.174 −0.004 0.056 of Kahya (2007). As education improves, job perfor-
effects → JP mance may decline. Higher education levels may not
RC→ SC 0.027 0.197 −0.005 0.078
→ JP guarantee higher job performance.
TC → SC 0.034 0.138 −0.006 0.086 Step 2: It can be seen from Table 6 that social capital
→ JP
Total CC → JP 0.303 0.000 0.182 0.421
is positively and significantly affected by job crafting
effects RC → JP 0.194 0.002 0.076 0.321 (β = 0.692, p < 0.001), cognitive crafting (β = 0.607,
TC → JP 0.342 0.000 0.220 0.462 p < 0.001), relational crafting (β = 0.629, p < 0.001) and
CC = Cognitive crafting;
RC = Relational crafting;
task crafting (β = 0.647, p < 0.001), so that H2, H2a,
TC = Task crafting; H2b and H2 c are all supported.
SC = Social capital; Step 3: It can be seen from Table 7 that the signifi-
JP = Job performance
cant positive impact of social capital on job perfor-
mance still exists after introducing social capital. The
The results show that the total effects of the three effects of job crafting, cognitive crafting, relational
facets of job crafting on job performance are signifi- crafting, and task crafting on job performance weaken,
cant. However, the indirect effects of cognitive crafting p < 0.001, with statistical significance. Moreover, the β
(β = 0.020; p = 0.174), relational crafting (β = 0.027; coefficients drop to 0.722, 0.250, and 0.199, indicating
p = 0.197) and task crafting (β = 0.034; p = 0.138) on that social capital plays a positive mediating role.
job performance are not significant within the range of Hypotheses H4, H4a, H4b and H4 c are all supported.
confidence intervals. Thus, none of hypotheses H4a, Furthermore, the results from model 6 in Table 7
H4b and H4 c are supported. show that social capital has a significant positive effect
on job performance (β = 0.652, p < 0.001), and
4.6.2. Hierarchical linear regression hypothesis H3 is supported.
Based on the arguments of Baron and Kenny (1986), This research measured core self-evaluation and
this study utilised three regression analyses to test social capital using a standard score to reduce multi-
whether social capital has a mediating effect. collinearity (Dalal & Zickar, 2012). Based on the argu-
Step 1: It can be seen from Table 5 that job perfor- ments of Baron and Kenny (1986), core self-evaluation
mance is positively and significantly affected by job negatively moderates the impact of social capital on
crafting (β = 0.804, p < 0.001), cognitive crafting work performance (β = −0.180, p < 0.001), and
(β = 0.729, p < 0.001), relational crafting (β = 0.705, hypothesis H5 is supported, as shown in Table 8.
p < 0.001) and task crafting (β = 0.757, p < 0.001), so The simple slope method proposed by (Aiken &
that H1, H1a, H1b and H1 c are all supported. West, 1991) was used to analyse the moderating role of
In addition, among the control variables, only edu- core self-evaluation in the prediction of job perfor-
cation level has a significant negative impact on job mance by social capital. Based on the average, the core

Table 5. Summary of hierarchical regression analysis of job crafting on job performance.

Job performance
Model 1 Model 2 Model 3 Model 4 Model 5
Variables
Gender .043 −.056 −.051 −.033 −.039
Age .028 −.011 .006 −.025 .008
Educational level −.249*** .011 −.047 −.029 −.015
Working time −.054 .030 −.004 .037 .014
Monthly salary .009 −.105* −.087 −.073 −.103*
Position .108 .033 .070 .038 .029
Business category −.073 −.075* −.083* −.054 −.090*
Job crafting .804***
Cognitive crafting .729***
Relational crafting .705***
Task crafting .757***
R square .056 .613 .529 .495 .551
Adj R square .041 .605 .520 .486 .543
F values 3.645** 84.790*** 60.109*** 52.622*** 65.906***
Δ R square .056 .557 .472 .439 .495
ΔF 3.645 616.300 429.911 373.370 473.686
Mix VIF 2.910 2.923 2.915 2.929 2.920
D-W 1.963 2.041 1.916 1.887
Note: *p < .05, *p < .01, *p < .001
 KNOWLEDGE MANAGEMENT RESEARCH & PRACTICE 259

Table 6. Summary of hierarchical regression analysis of job crafting on social capital.

Social capital
Model 1 Model 2 Model 3 Model 4 Model 5
Variables
Gender .111 .026 .033 .044 .041
Age .005 −.029 −.014 −.043 −.012
Educational level −.394*** −.170*** −.226*** −.198*** −.194***
Working time −.081 −.010 −.040 −.001 −.023
Monthly salary .191** .093* .111* .118** .096*
Position .158** .094* .127** .096* .091*
Business category −.049 −.050 −.056 −.031 −.063
Job crafting .692***
Cognitive crafting .607***
Relational crafting .629***
Task crafting .647***
R square .193 .605 .521 .542 .555
Adj R square .180 .597 .512 .534 .546
F values 14.659*** 82.022*** 58.255*** 63.560*** 66.815***
Δ R square .193 .412 .328 .350 .362
ΔF 14.659 447.108 293.603 327.868 348.891
Mix VIF 2.910 2.923 2.915 2.929 2.920
D-W 1.719 1.556 1.745 1.792
*p < .05, *p < .01, *p < .001

Table 7. Hierarchical regression analysis of mediation.

Job performance
Model 1 Model 2 Model 3 Model 4 Model 5 Model 6
Variables
Gender .043 −.059 −.060 −.045 −.049 −.030
Age .028 −.007 .010 −.013 .011 .025
Educational level −.249*** .032 .015 .029 .030 .008
Working time −.054 .031 .007 .037 .020 −.001
Monthly salary .009 .116** −.118** −.107* −.125** −.116*
Position .108 .022 .035 .010 .008 .005
Business category −.073 −.069* −.067 −.045 −.076* −.042
Job crafting .722***
Cognitive crafting .561***
Relational crafting .521***
Task crafting .608***
Social capital .120* .276*** .293*** .230*** .652***
R square .056 .618 .565 .535 .575 0.399
Adj R square .041 .610 .556 .525 .566 0.388
F values 3.645** 77.009*** 61.791*** 54.630*** 64.343*** 35.581***
Δ R square .056 .562 .509 .479 .519 .343
ΔF 3.645 315.145 250.499 220.080 261.340 244.679
Mix VIF 2.910 2.923 2.919 2.929 2.898 2.918
D-W 1.903 1.951 1.799 1.778 1.670
*p < .05, *p < .01, *p < .001

Table 8. Summary of hierarchical regression analysis.

Job performance self-evaluation was divided into two groups and used
Model 1 Model 2 Model 3 to give the simple slopes for high core self-evaluation
Variables β = 0.072 (t = 1.345, p = 0.179) and low core self-
Gender .043 −.046 −.053 evaluation β = 0.247 (t = 6.882, p = 0.000). The results
Age .028 .010 .013
Educational level −.249*** .027 −.010 show that there is no significant correlation between
Working time −.054 .025 .013 social capital and job performance for high core self-
Monthly salary .009 −.115** −.081
Position .108 .026 .041 evaluation, but there is positive correlation between
Business category .043 −.055 −.048 social capital and job performance for low core self-
Z score: Social capital .245*** .246***
Z score: Core self-evaluation . .582*** 461***
evaluation (Figure 2).
Social capital* Core self-evaluation −.180***
R square .056 .563 .582
Adj R square .041 .554 .572
F values 3.645** 61.355*** 59.521*** 5. Discussion and conclusion
Δ R square .056 .507 .019
ΔF 3.645 248.645 19.346 5.1. Discussion
Mix VIF 2.910 2.922 2.899
D-W 1.797 The questionnaire collected in this study produced
*p < .05, *p < .01, *p < .001 one-time cross-sectional data. The results of the
260 C. ZHANG AND L. LIU

5.2
5.1
5
4.9

Job performance
Low core self-
4.8 evaluation
4.7 High core self-
evaluation
4.6
4.5
4.4
Low social capital High social capital

Figure 2. Simple slopes.

empirical analysis show that all the hypotheses pro- performance is a behaviour factor. This paper explains
posed in this study are valid. Firstly, job crafting has that social capital has a strong influence on job per-
a positive impact on job performance, which is the formance when there is low core self-evaluation.
same result as found by Tims et al. (2015). Secondly, Using the framework of job crafting, social capital,
job crafting has a positive impact on social capital. core self-evaluation and job performance, this study
Based on the theory of social exchange, this study enriches social exchange theory and social cognition
adopted the definition of job crafting that includes theory. Taking employees in Guangxi as the research
three dimensions (the cognitive dimension, the rela- object, this paper provides human resource manage-
tional dimension and the task dimension) given by ment suggestions for the development of Guangxi
Slemp and Vella-Brodrick (2013), and the definition enterprises against the background of “One Belt One
of social capital including three dimensions (the cog- Road”. First, job crafting and social capital are factors
nitive dimension, the relational dimension and the that influence employee performance. When employ-
structural dimension) given by Chumg et al. (2015). ees with higher job crafting skill and greater social
Although the two variables both have a cognitive and capital are recruited by human resources management
a relational dimension, they have different perspec- departments of enterprises, this can reduce post-
tives. The perspective of job crafting is self-oriented, training costs. At the same time, for employees who
while the perspective of social capital is oriented have already been employed, an improvement in their
towards the individual’s colleagues. According to job crafting skills and social capital will help to
social exchange theory, my efforts are rewarded by improve their job performance. Second, it is not
my colleagues. Next, in the hierarchical linear regres- appropriate to recruit employees with high core self-
sion analysis, social capital has a positive impact on evaluation, as this will slow down the impact of social
job performance, which is the same result as that capital on job performance.
found by Ali-Hassan et al. (2015). However, in SEM,
social capital has no significant effect on job perfor-
5.2. Limitations and future research
mance. In the model, because of the existence of job
crafting, the impact of social capital on job perfor- This study was limited by objective resources such as
mance is not statistically significant. Although Basu manpower, materials, financial resources and time.
et al. (2017) confirmed the positive impact of social Therefore, convenience sampling and cross-sectional
capital on job performance, they used organisational data were used in the data collection. Secondly, in the
citizenship behaviour, social capital, and job perfor- theoretical test, only the relationships between job
mance as observed variables rather than latent vari- crafting, social capital, core self-evaluation and job
ables in the Amos analysis. This means that when performance were confirmed. There was no further
individuals have a positive attitude, such as job craft- examination of the dimensions of the variables other
ing, their social capital will not have an impact on job than job crafting. Finally, taking employees of enter-
performance. In hierarchical linear regression analy- prises in Guangxi as the research object means that the
sis, social capital then partially mediates the impact of results cannot be generalised to other regions of
job crafting on job performance. Finally, core self- China.
evaluation negatively regulates the impact of social In follow-up studies, other sampling methods, such
capital on job performance. According to social cogni- as quota sampling or random sampling, and longitu-
tion theory, social capital is an environmental factor, dinal research could be adopted. Secondly, the intrin-
core self-evaluation is an individual factor, and job sic motivation of the impact of job crafting and social
 KNOWLEDGE MANAGEMENT RESEARCH & PRACTICE 261

capital on job performance could be studied. Thirdly, Bhargava, S., & Pradhan, H. (2017). Work value congruence
in terms of measurement tools, other scholars could and job performance: Buffering role of leader member
choose to compile different questionnaires and could, exchange and core self-evaluation. Asian Social Science,
13(1), 98–105. https://doi.org/10.5539/ass.v13n1p98
for example, measure organisational citizenship beha- Bhargava, S., & Pradhan, H. (2019). Moderating effect of
viour and task performance for the job performance situational factors on core self-evaluation and perfor-
scale. Other environmental, individual and beha- mance relationship. Global Business Review, 20(1),
vioural variables in social cognition theory could be 238–252. https://doi.org/10.1177/0972150918803833
tried in the future; for example, job satisfaction could Blau, P. M. (1960). A theory of social integration. American
Journal of Sociology, 65(6), 545–556. https://doi.org/10.
be used as an individual variable. In addition, in the
1086/222785
SEM analysis, the mediation pathway proposed in this Bolino, M. C., Turnley, W. H., & Bloodgood, J. M. (2002).
study may be wrong or may be limited by the sample, Citizenship behavior and the creation of social capital in
so subsequent studies could try other samples or find organizations. Academy of Management Review, 27(4),
other mediation variables. Some scholars have con- 505–522. https://doi.org/10.5465/amr.2002.7566023
firmed the impact of social capital on knowledge shar- Chen, C.-H. V., Yuan, M.-L., Cheng, J.-W., & Seifert, R.
(2016). Linking transformational leadership and core
ing (Yen et al., 2015), and subsequent attempts could self-evaluation to job performance: The mediating role
be made to study the path from job crafting → social of felt accountability. The North American Journal of
capital → knowledge. Economics and Finance, 35, 234–246. https://doi.org/10.
1016/j.najef.2015.10.012
Chen, M.-H., Wang, H.-Y., & Wang, M.-C. (2018).
Disclosure statement Knowledge sharing, social capital, and financial perfor-
mance: The perspectives of innovation strategy in tech-
No potential conflict of interest was reported by the authors. nological clusters. Knowledge Management Research &
Practice, 16(1), 89–104. https://doi.org/10.1080/
14778238.2017.1415119
ORCID Chumg, H.-F., Cooke, L., Fry, J., & Hung, I.-H. (2015).
Factors affecting knowledge sharing in the virtual orga-
Chunyu Zhang http://orcid.org/0000-0002-7008-0742 nisation: Employees’ sense of well-being as a mediating
Liping Liu http://orcid.org/0000-0002-7422-2277 effect. Computers in Human Behavior, 44, 70–80. https://
doi.org/10.1016/j.chb.2014.11.040
Creed, P. A., & Gagliardi, R.-E. (2015). Career compromise,
References
career distress, and perceptions of employability: The
Aiken, L. S., & West, S. G. (1991). Multiple regression: moderating roles of social capital and core
Testing and interpreting interactions. Sage. self-evaluations. Journal of Career Assessment, 23(1),
Ali-Hassan, H., Nevo, D., & Wade, M. (2015). Linking 20–34. https://doi.org/10.1177/1069072714523082
dimensions of social media use to job performance: The Dalal, D. K., & Zickar, M. J. (2012). Some common myths
role of social capital. The Journal of Strategic Information about centering predictor variables in moderated multi-
Systems, 24(2), 65–89. https://doi.org/10.1016/j.jsis.2015. ple regression and polynomial regression. Organizational
03.001 Research Methods, 15(3), 339–362. https://doi.org/10.
Anderson, J. C., & Gerbing, D. W. (1988). Structural equa- 1177/1094428111430540
tion modeling in practice: A review and recommended Fornell, C., & Larcker, D. F. (1981). Evaluating structural
two-step approach. Psychological Bulletin, 103(3), equation models with unobservable variables and mea-
411–423. https://doi.org/10.1037/0033-2909.103.3.411 surement error. Journal of Marketing Research, 18(1),
Bagozzi, R. P., & Yi, Y. (1988). On the evaluation of struc- 39–50. https://doi.org/10.1177/002224378101800104
tural equation models. Journal of the Academy of Hair, J. F., Anderson, R., Tatham, R., & Black, W. (1998).
Marketing Science, 16(1), 74–94. https://doi.org/10.1007/ Multivariate data analysis. Prentice-Hall.
BF02723327 Hau, Y. S., Kim, B., Lee, H., & Kim, Y.-G. (2013). The effects
Bakker, A. B., Tims, M., & Derks, D. (2012). Proactive of individual motivations and social capital on employees’
personality and job performance: The role of job crafting tacit and explicit knowledge sharing intentions.
and work engagement. Human Relations, 65(10), International Journal of Information Management, 33
1359–1378. https://doi.org/10.1177/0018726712453471 (2), 356–366. https://doi.org/10.1016/j.ijinfomgt.2012.10.
Bandura, A., & Walters, R. H. (1977). Social learning theory. 009
Prentice-hall. Judge, T. A., Erez, A., Bono, J. E., & Thoresen, C. J. (2003).
Baron, R. M., & Kenny, D. A. (1986). The The core self-evaluations scale: Development of
moderator-mediator variable distinction in social psy- a measure. Personnel Psychology, 56(2), 303–331. https://
chological research: Conceptual, strategic, and statistical doi.org/10.1111/j.1744-6570.2003.tb00152.x
considerations. Journal of Personality and Social Kacmar, K. M., Collins, B. J., Harris, K. J., & Judge, T. A.
Psychology, 51(6), 1173–1182. https://doi.org/10.1037/ (2009). Core self-evaluations and job performance: The
0022-3514.51.6.1173 role of the perceived work environment. Journal of
Basu, E., Pradhan, R. K., & Tewari, H. R. (2017). Impact of Applied Psychology, 94(6), 1572–1580. https://doi.org/10.
organizational citizenship behavior on job performance 1037/a0017498
in Indian healthcare industries: The mediating role of Kahya, E. (2007). The effects of job characteristics and work-
social capital. International Journal of Productivity and ing conditions on job performance. International Journal
Performance Management, 66(6), 780–796. https://doi. of Industrial Ergonomics, 37(6), 515–523. https://doi.org/
org/10.1108/IJPPM-02-2016-0048 10.1016/j.ergon.2007.02.006
262 C. ZHANG AND L. LIU

Kline, R. B. (2005). Principles and practice of structural Slemp, G. R., & Vella-Brodrick, D. A. (2013). The job crafting
equation modeling 2nd. Guilford. questionnaire: A new scale to measure the extent to which
Lefebvre, V. M., Sorenson, D., Henchion, M., & Gellynck, X. employees engage in job crafting. International Journal of
(2016). Social capital and knowledge sharing perfor- Wellbeing, 3(2), 126–146. doi:10.5502/ijw.v3i2.1
mance of learning networks. International Journal of Staufenbiel, T., & König, C. J. (2010). A model for the effects
Information Management, 36(4), 570–579. https://doi. of job insecurity on performance, turnover intention, and
org/10.1016/j.ijinfomgt.2015.11.008 absenteeism. Journal of Occupational and Organizational
Li, L., & Lin, T. T. (2018). Examining how dependence on Psychology, 83(1), 101–117. https://doi.org/10.1348/
smartphones at work relates to Chinese employees’ work- 096317908X401912
place social capital, job performance, and smartphone Tian, X., & Liu, M. (2017). Individual strengths-based job
addiction. Information Development, 34(5), 489–503. crafting. Advances in Psychological Science, 25(9),
https://doi.org/10.1177/0266666917721735 1579–1596. https://doi.org/10.3724/sp.J.1042.2017.01579
Lim, S., & Tai, K. (2014). Family incivility and job performance: Tims, M., Bakker, A. B., & Derks, D. (2015). Job crafting
A moderated mediation model of psychological distress and and job performance: A longitudinal study. European
core self-evaluation. Journal of Applied Psychology, 99(2), Journal of Work and Organizational Psychology, 24(6),
351–359. https://doi.org/10.1037/a0034486 914–928. https://doi.org/10.1080/1359432X.2014.
Liu, C., Li, C., Fan, J., & Nauta, M. M. (2015). Workplace 969245
conflict and absence/lateness: The moderating effect of Tims, M., Bakker, A. B., Derks, D., & Van Rhenen, W.
core self-evaluation in China and the United States. (2013). Job crafting at the team and individual level:
International Journal of Stress Management, 22(3), 243. Implications for work engagement and performance.
https://doi.org/10.1037/a0039163 Group & Organization Management, 38(4), 427–454.
Lu, -H.-H. (2012). An empirical study of the relationship https://doi.org/10.1177/1059601113492421
among social capital, knowledge integration capability Wang, P., Wagner, T. A., Boyar, S. L., Corman, S. A., &
and job performance: From the perspective of social cogni- McKinley, R. B. (2016). The relationship between organi-
tive theory. [Master’s thesis]. National Cheng Kung zational family support and burnout among women in the
University. healthcare industry: Core self-evaluation as moderator
Mao, K., & Li, C. (2015). Proactive behaviors of newcomers handbook on well-being of working women. Springer.
and their influences in the organizational socialization. Wang, Z., Bu, X., & Cai, S. (2018). Core self-evaluation,
Advances in Psychological Science, 23(12), 2167–2176. individual intellectual capital and employee creativity.
https://doi.org/10.3724/SP.J.1042.2015.02167 Current Psychology, 1–15. https://doi.org/10.1007/
Marsh, H. W., Hau, K.-T., & Wen, Z. (2004). In search of s12144-018-0046-x
golden rules: Comment on hypothesis-testing approaches Williams, L. J., & Anderson, S. E. (1991). Job satisfaction and
to setting cutoff values for fit indexes and dangers in organizational commitment as predictors of organiza-
overgeneralizing Hu and Bentler’s (1999) findings. tional citizenship and in-role behaviors. Journal of
Structural Equation Modeling, 11(3), 320–341. https:// Management, 17(3), 601–617. https://doi.org/10.1177/
doi.org/10.1207/s15328007sem1103_2 014920639101700305
Medsker, G. J., Williams, L. J., & Holahan, P. J. (1994). Wrzesniewski, A., & Dutton, J. E. (2001). Crafting a job:
A review of current practices for evaluating causal Revisioning employees as active crafters of their work.
models in organizational behavior and human Academy of Management Review, 26(2), 179–201. https://
resources management research. Journal of doi.org/10.5465/amr.2001.4378011
Management, 20(2), 439–464. https://doi.org/10.1177/ Yen, Y.-F., Tseng, J.-F., & Wang, H.-K. (2015). The effect of
014920639402000207 internal social capital on knowledge sharing. Knowledge
Nahapiet, J., & Ghoshal, S. (1998). Social capital, intellectual Management Research & Practice, 13(2), 214–224. https://
capital, and the organizational advantage. Academy of doi.org/10.1057/kmrp.2013.43
Management Review, 23(2), 242–266. https://doi.org/10. Yu, H. (2017). Motivation behind China’s ‘one belt, one
5465/amr.1998.533225 road’ initiatives and establishment of the Asian infra-
Ng, T. W., & Feldman, D. C. (2009). How broadly does structure investment bank. Journal of Contemporary
education contribute to job performance? Personnel China, 26(105), 353–368. https://doi.org/10.1080/
Psychology, 62(1), 89–134. https://doi.org/10.1111/j. 10670564.2016.1245894
1744-6570.2008.01130.x Zhang, C.-Y., & Chen, C.-S. (2017) The effect of Zhong-yong
Podsakoff, P. M., MacKenzie, S. B., Lee, J.-Y., & thinking to employee survival ability-taking social capital
Podsakoff, N. P. (2003). Common method biases in beha- as a mediating variable and knowledge sharing behaviour
vioral research: A critical review of the literature and recom- as a moderating variable. Paper presented at the 2017 2nd
mended remedies. Journal of Applied Psychology, 88(5), international conference on economics and management
879–903. https://doi.org/10.1037/0021-9010.88.5.879 innovations, Bangkok,Thailand.
Preacher, K. J., & Hayes, A. F. (2008). Asymptotic and Zhang, L., Cai, Y., & Zhou, N. (2016). The formation mechan-
resampling strategies for assessing and comparing indir- ism of knowledge sharing under the boundary condition of
ect effects in multiple mediator models. Behavior team creativity: Multilevel moderated mediation model.
Research Methods, 40(3), 879–891. https://doi.org/10. Science & Technology Progress and Policy, 33(10),
3758/BRM.40.3.879 134–139. https://doi.org/10.6049/kjjbydc.2015120301
Qi, J., Li, J., & Zhang, Q. (2014). How organizational Zimmermann, A., Oshri, I., Lioliou, E., & Gerbasi, A. (2018).
embeddedness and affective commitment influence job Sourcing in or out: Implications for social capital and
crafting. Social Behavior and Personality: An knowledge sharing. The Journal of Strategic Information
International Journal, 42(10), 1629–1638. https://doi. Systems, 27(1), 82–100. https://doi.org/10.1016/j.jsis.2017.
org/10.2224/sbp.2014.42.10.1629 05.001