p–ISSN: 2723 - 6609 e-ISSN: 2745-5254
Vol. 5, No. 9 September 2024 http://jist.publikasiindonesia.id/
Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 9, September 2024 3491
The Effect of High-Performance Work System on Intention to
Leave and Safety Workarounds: The Role of Burnout as
Mediation and Mentoring & Coping Mechanism as
Moderator
Syavergio Avia Difaputra
1*
, Hunik Sri Runing Sawitri
2
Universitas Sebelas Maret, Indonesia
Email:
[email protected]
1*
, [email protected]
2
*Correspondence
ABSTRACT
Keywords: high-
performance work system,
burnout, intention to
leave, safety
workarounds, mentoring
This study aims to analyze the influence of the High-
Performance Work System (HPWS) on intention to leave
and safety workarounds, with the role of burnout as a
mediation variable and mentoring and coping mechanisms
as a moderator variable. The study was conducted on non-
doctor employees at the Diponegoro National Hospital,
Semarang. The method used is a quantitative approach with
data collection techniques through questionnaires distributed
to respondents. The data was analyzed using Partial Least
Squares Structural Equation Modeling (PLS-SEM). The
results of the study show that HPWS has a negative influence
on burnout, which means that the implementation of HPWS
can reduce the level of emotional fatigue of employees. In
contrast, burnout has a positive effect on the intention to
leave and safety workarounds, indicating that employees
who experience burnout are more likely to want to leave the
company and ignore work safety procedures. In addition,
burnout acts as a mediator between HPWS and the two
variables. Good HPWS implementation can reduce burnout,
thereby reducing employee intentions to move and
increasing compliance with safety procedures. This study
concludes that the effective implementation of HPWS can
reduce employee burnout, reduce the desire to leave work,
and increase compliance with occupational safety.
Mentoring and coping mechanisms play an important role in
strengthening the positive effects of HPWS.
Introduction
Human resources (HR) are valuable assets of an organization. No organization can
exist and grow without the appropriate capabilities and competencies of human resources.
Most organizations have adopted HR practices such as recruitment and selection, training
and development to encourage, motivate, and improve employee morale to achieve
organizational goals. Globalization, privatization, deregulation, competition, and
Syavergio Avia Difaputra, Hunik Sri Runing Sawitri
Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 9, September 2024 3492
technological advances have resulted in dramatic changes in human resource practices.
This change in the environment has forced organizations to adopt a High-Performance
Working System (HPWS) (Jyoti et al., 2015). HPWS is typically used to describe
integrated or combined human resource practice systems, work structures, and processes
designed to produce a high level of employee knowledge, skills, attitudes, motivation,
and flexibility. HPWS has a negative impact and also a positive impact on individuals
and organizations. Apart from the performance benefits resulting from the
implementation of HPWS, the employee's perspective must be considered, especially the
psychological and physiological outcomes towards his or her work, such as attitude,
motivation, work, and life experience (Han et al., 2020).
Currently, it has become a global issue that hospital workers, in this case nurses,
have the intention to leave their jobs or have the intention to leave (ITL) (Rouleau et al.,
2012). Intention to Leave is one of the main difficulties faced by some healthcare
organizations, leading to inadequate nursing staff, increased work stress due to increased
workload, job dissatisfaction, productivity, and the intention to resign and switch to other
healthcare services. (Zaheer et al., 2019). Previous research has stated that higher nurse
turnover encourages longer hospitalizations and an increase in the number of medical
errors. Therefore, it hurts the patient's need for proper high-quality care. Moving
intentions can also incur higher costs caused by the need to replace employees, recruit
and train temporary staff, and ensure the quality of service. (Albougami et al., 2020).
Based on previous research, work-related organizational factors and employee-
related factors influence the decision to leave through a person's job fit and the
individual's response to that fit, which in turn affects the employee's intention to stay or
leave the workplace. These things are likely influenced by physiological responses such
as health problems, cognitive reactions such as negative thoughts towards administration,
and finally emotional reactions stemming from work-related and personal factors.
(Gaudenz et al., 2019). Intention to leave is one of the attitudes of employees whose focus
has developed on negative perceptions of HPWS practices implemented by organizations.
Apart from the results of HPWS that have a positive impact, among others, there is an
increase in financial value, the quality of a product and innovation, as well as the service
and satisfaction of a company's customers.
Burnout from an employee can be seen if the employee shows a slow work
sensation, intention to move, and takes psychological agility. In terms of taking the
psychological agility of employees within the hospital sector, hospital employee
workarounds have also received increased attention over the past 15 years, which has
adjusted to an increased focus on patient safety, evidence-based practices, and increased
use of health information technology. (Seaman & Erlen, 2015). Workarounds in hospital
employees are usually actions taken by nurses or other hospital employees in a health
organization to avoid blocks in the workflow and thus achieve the desired goals, but these
actions deviate from the protocol set by the organization. This is considered a deviation
from practice that puts patients at risk of poor outcomes. (Seaman & Erlen, 2015).
The Effect of High-Performance Work System on Intention to Leave and Safety Workarounds:
The Role of Burnout as Mediation and Mentoring & Coping Mechanism as Moderator
Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 9, September 2024 3493
Coping strategies serve as mitigation techniques or training in intellectual and
behavioral practices that help extend direct action to resolve the situation ("Sunny", Hu
and Cheng, 2010). Many stress management strategies for nurses have different effects
on nurse job completion and managing emotions (Labrague et al., 2018). However,
approach-oriented strategies are helpful and encouraging in managing life stressors, and
difficult and negative behavior attempts. Several studies illustrate that an optimistic
approach to stress management strategies positively affects the quality of life of hospital
employees. (Cruz et al., 2018) And there is a decrease in symptoms of neglect. Proper
planning is necessary for active coping. The most commonly used coping strategies by
nurses are self-control, problem-solving approaches, and seeking social support.
Based on the formulation of the problem that has been made, this study has the
following objectives:
1. Testing and analyzing the effect of HPWS on burnout in employees in hospitals.
2. Testing and analyzing the effect of burnout on ITL on employees in hospitals.
3. Testing and analyzing burnout on safety workaround in employees in hospitals.
4. Testing and analyzing burnout as a mediator of the influence between HPWS and ITL
on employees in hospitals.
5. Testing and analyzing the effect of burnout as a mediator of the relationship between
HPWS and safety workaround in employees in hospitals
6. Testing and analyzing mentoring as a moderator of the effect between HPWS and
burnout on employees in hospitals.
7. Testing and analyzing mentoring as a moderator of the influence between burnout and
ITL.
8. Testing and analyzing the coping mechanism as a moderation of the effect between
burnout and safety workaround on hospital employees.
Method
Research Design
According to (Sekaran & Bougie, 2016) The definition of research design is a
blueprint plan that is carried out in the form of data collection, measurement, and analysis
to answer the questions that exist in the research. In the research conducted, a quantitative
approach was carried out where this approach was aimed at looking at the influence of
the high-performance work system variable on the variable of intention to leave and safety
workarounds which were mediated by the burnout mediation variable and there was also
a mentoring and coping mechanism as a moderator variable.
The data collection process carried out in this study uses a survey technique whose
respondents are employees of medical personnel (general nurses, dental nurses,
midwives, pharmacists, medical records, physiotherapists, CSSD, laboratories, and
radiology) at the Diponegoro National Hospital Semarang. A survey is a research design
that can be classified based on the approach used to collect data. It can be by observing
conditions, behaviors, events, people, or a process. In addition, the research conducted
uses a cross-sectional approach. This approach is based on (Sekaran & Bougie, 2016) is
Syavergio Avia Difaputra, Hunik Sri Runing Sawitri
Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 9, September 2024 3494
an approach in studies where data is only collected once, and may be done over several
days, weeks, or months to answer various questions from existing research. The data
collection process in this study will be carried out from December 2023 to January 2024.
The unit of analysis in this study is individuals, in this case, employees of the Diponegoro
National Hospital Semarang, a State Teaching Hospital in Semarang.
Population, Samples, and Sampling
The following are the population, samples, and sampling in this study.
1. Population
A population is something that refers to an entire group, event, or a particular
interesting thing that the researcher wants to explore further. Population is a group of
people, phenomena, or interesting things that researchers want to make conclusions based
on statistical samples. Based on this definition, the population of the study was 186
employees (general nurses, dental nurses, midwives, pharmacists, medical records,
physiotherapists, CSSD, laboratories, radiology) at Diponegoro National Hospital
Semarang.
2. Sample
A sample is a representation of data obtained by selecting several elements in a
research population, from which conclusions can be drawn about the entire population.
The target sample in this study is employees at the Diponegoro National Hospital
Semarang. This study took 100% of the population at the Diponegoro National Hospital
Semarang consisting of general nurses, dental nurses, midwives, pharmacists, medical
records, physiotherapists, CSSD, laboratories, and radiology employees who were used
as observation units, which are also known as census techniques. According to Cooper
and Schlinder, the 2014 census is a calculation of all elements in a population, where a
list of all elements of a population is used as a sample framework. In small, accessible,
and varied populations, the accuracy of census use tends to be greater compared to
samples.
Research Instruments
In the data collection process carried out, this study uses a questionnaire as a
research instrument. A questionnaire is a collection of written questions formulated by
the researcher to then be answered by the respondents in a study. The questionnaire will
be made online using Google Forms and then will be distributed to employees of the
Diponegoro National Hospital Semarang online.
Data Source
There are two types of data sources used in this study, the first is primary and the
second is secondary. The following is an explanation of the two types of data sources.
Data Primer
Primary data is data that researchers collect directly for the specific purpose of a
study. In this study, primary data was obtained from the results of a survey conducted
online and the respondents were employees of the Diponegoro National Hospital
Semarang.
The Effect of High-Performance Work System on Intention to Leave and Safety Workarounds:
The Role of Burnout as Mediation and Mentoring & Coping Mechanism as Moderator
Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 9, September 2024 3495
Data Seconds
Secondary data is data that has been collected by other parties for research
conducted at the moment. In the writing of this study, secondary data was obtained
through actual and factual news articles, books, and articles from research journals related
to the topic of human resources, management, and other literacy sources.
Data Collection Methods
The method or data collection technique carried out in this study is using a survey
distributed through an online questionnaire. The advantages of distributing questionnaires
conducted online are that data can be obtained more easily and quickly, gain a better
understanding of consumer opinions and preferences, provide access to groups and
individuals that are difficult to reach through other channels, and the latter can also cover
a large geographical area in the survey.
Descriptive Analysis
This descriptive analysis is used to obtain data that presents the topic or problem
being researched so that the characteristics of a group can be understood, as for the
characteristics of the respondents to be researched in this analysis, starting from the last
education, work department, and length of service.
Validity Test
Validity indicates the extent to which an instrument accurately measures what is
intended to be measured, which in this case is the observed behavior. The validity test in
this study was carried out Evaluation of Measurement, where an evaluation of the data
that had been collected was carried out to be used as a measuring tool of the research
variables. In evaluating measurement, it is necessary to carry out convergent validity and
discriminant validity. (Sohu et al., 2023). Convergent validity is analyzed by the loading
factor which is the value produced by each indicator to measure the variable, where values
above 0.4 are maintained, while values below 0.4 are eliminated. Discriminant validity is
assessed from the Fornell Larcker criterion which is the correlation value between the
variable and the variable itself, and the variable with other variables, to see the correlation
of the variable with the variable itself, should not be smaller than the correlation of the
variable with other variables.
Reliability Test
The feasibility test is an analysis conducted to show whether the measurement is
free of errors and evidence that the measurements in the study are consistent if the
measurements are repeated and consistent if they cross in various indicators. In this study,
reliability analysis using Cronbach's Alpha and Composite Reliability was used. The
values of Cronbach's Alpha and Composite Reliability will be categorized as good
reliability if the value is between 0.80 – 1.0, categorized as reliability can be received if
the value is between 0.60 – 0.79, and categorized as poor reliability if the value < 0.60.
Results and Discussion
Respondents by Gender
The following is an overview of respondents by gender presented in Table 1.
Syavergio Avia Difaputra, Hunik Sri Runing Sawitri
Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 9, September 2024 3496
Table 1
Respondents by Gender Group
Gender
Sum
Percentage
Man
24 people
19,9 %
Woman
97 people
80,1 %
Sum
121 people
100 %
Source: Primary data processed, 2024
The demographics of the respondents who filled in this study showed that non-
doctor employees at RSND were dominated by 97 women or 80.1%, while male non-
doctor employees amounted to 24 people or 19.9%. Based on data from the HR
department of RSND, of the 130 non-doctor employees, female employees are indeed
dominated by female employees rather than male employees.
Respondents By Position
The following is an overview of the number of respondents based on their positions
presented in Table 2.
Table 2
Respondents Based on Position Groups at RSND
Position
Sum
Percentage
General Nurse
62 orang
51.2 %
Dental Nurse
4 orang
3,3 %
Pharmacy
22 orang
18,2 %
Laborat
11 orang
9,1 %
Midwife
3 orang
2,5 %
Managing Midwife
1 orang
0,8 %
Medical Registration
and Records
6 orang
4,9 %
Physiotherapist
3 orang
2,5 %
CSSD
3 orang
2,5 %
Radiographer
3 orang
2,5 %
Nutritionist
1 orang
0,8 %
Medical Physicist
1 orang
0,8 %
Sanitarian
1 orang
0,8 %
Sum
121 orang
100%
Source: Primary data processed, 2024
Respondents' Responses to Research Variables
The answers from the respondents contained an analysis of the frequency of
responses of non-doctor employees of Diponegoro National Hospital as research
respondents. The results obtained from the respondents' responses based on the questions
asked by the researcher are by the variables that the researcher has. There were 28
questions from 6 research variables, namely high-performance work system, intention to
leave, safety workarounds, burnout, mentoring, and coping mechanism. Each of these
variables uses a 5-point Likert scale.
Evaluation of Measurement Model (Outer Model)
The Effect of High-Performance Work System on Intention to Leave and Safety Workarounds:
The Role of Burnout as Mediation and Mentoring & Coping Mechanism as Moderator
Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 9, September 2024 3497
The data analysis in this study uses the Partial Least Square-Structural Equation
Model (SEM-PLS). The analysis test using SEM-PLS begins by analyzing the loading
factor or outer loading, which is the value produced by each indicator to measure the
variable. The measurement model or outer model is used as a tool to measure convergent
validity and also discriminant validity to measure the value of each indicator. After that,
a reality test was carried out with composite reality measurements and also Cronbach's
alpha. This is to test whether the statement of each indicator is by the conditions in the
field.
Convergent Validity
In analyzing the data, the first thing to do is convergent validity, where the first
thing to do is to analyze the loading factor or outer loading, which is the value produced
by each indicator to measure the variable, which within the limit is 0.7. So it can be said
that the validity test in this study is carried out by a factor analysis test where the test will
be said to pass if each question instrument in the existing variable has a loading factor
value of more than equal to 0.7. Based on the test results, the calculation of the loading
factor/outer loading value listed in Table 3 was obtained.
Table 3
Outer Loading
BO
CM
HPWS
ITL
MEN
SW
Informa
tion
0.82
7
Valid
0.80
9
Valid
0.85
4
Valid
0.87
0
Valid
0.75
7
Valid
0.710
Valid
0.835
Valid
0.715
Valid
0.836
Valid
0.742
Valid
0.873
Valid
0.874
Valid
0.813
Valid
0.822
Valid
0.834
Valid
0.720
Valid
0.878
Valid
0.866
Valid
0.877
Valid
0.915
Valid
0.801
Valid
Syavergio Avia Difaputra, Hunik Sri Runing Sawitri
Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 9, September 2024 3498
0.809
Valid
0.868
Valid
0.827
Valid
0.879
Valid
0.83
4
Valid
0.73
7
Valid
0.81
7
Valid
Based on Table 3, it is known that the factor loading/outer loading value is greater
than 0.7. So the 28 indicators in this study are said to be valid and mean that these
indicators are significant in measuring a construct.
Furthermore, it is necessary to carry out convergent validity testing using average
variance extracted (AVE). Where this value is the value owned by each variable. If the
AVE value is below 0.5, it means that there is an invalid indicator, and analysis is needed
on the outer loading again. The results of obtaining AVE scores are listed in Table 4.
Table 4
Average Variance Extracted (AVE)
Variable
AVE
Information
BO
0.679
Valid
CM
0.592
Valid
HPWS
0.712
Valid
ITL
0.729
Valid
MEN
0.701
Valid
SW
0.635
Valid
Based on Table 11, it can be seen that all six variables have values above 0.5. This
shows that these indicators are all valid and do not need to be re-analyzed on the outer
loading.
Discriminant Validity
After conducting convergent validity, it is followed by discriminant validity where
an assessment is carried out from the Fornell Larcker criterion which is the correlation
value between the variable and the variable itself, and the variable with other variables.
In this assessment, the correlation of the variable with the variable itself, should not be
smaller than the correlation of the variable with other variables. The Fornell Larcker
criterion assessment in this study is listed in Table 5.
The Effect of High-Performance Work System on Intention to Leave and Safety Workarounds:
The Role of Burnout as Mediation and Mentoring & Coping Mechanism as Moderator
Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 9, September 2024 3499
Table 5
Fornell Larcker Criterion
Based on Table 5, the fornell larcker criterion assessment shows that the value of
the fornell larcker criterion in the top row is greater than the value in the bottom row with
diagonal analysis. This shows that the correlation of variables with the variable itself is
greater than the correlation of variables with other variables.
An assessment is carried out to see the correlation between the variables, then look
at cross-loading, which is the correlation between indicators and variables. In this
assessment, the indicator of the correlation variable must be greater than the correlation
between the indicator and other variables. The cross-loading assessment in this study is
shown in Table 6.
Table 6
Cross Loading
BO
CM
HPWS
ITL
MEN
SW
BO1
0,827
0,478
-0,286
0,513
0,167
0,494
BO2
0,809
0,533
-0,178
0,382
0,263
0,378
BO3
0,854
0,535
-0,250
0,560
0,093
0,487
BO4
0,870
0,480
-0,383
0,617
0,086
0,542
BO5
0,757
0,371
-0,196
0,499
0,216
0,413
CM1
0,254
0,710
0,309
0,223
0,567
0,120
CM2
0,524
0,835
-0,140
0,532
0,262
0,460
CM3
0,234
0,715
0,313
0,226
0,483
0,167
CM4
0,636
0,836
-0,117
0,542
0,242
0,461
CM5
0,288
0,742
0,216
0,294
0,511
0,279
HPWS1
-0,309
-0,006
0,873
-0,352
0,174
-0,249
HPWS2
-0,273
0,015
0,874
-0,261
0,242
-0,178
HPWS3
-0,272
0,081
0,813
-0,248
0,234
-0,325
HPWS4
-0,247
0,048
0,822
-0,206
0,311
-0,145
HPWS5
-0,255
-0,023
0,834
-0,209
0,178
-0,178
ITL1
0,383
0,403
-0,130
0,720
0,242
0,218
ITL2
0,508
0,448
-0,311
0,878
0,073
0,419
Variable
BO
CM
HPWS
ITL
MEN
SW
BO
0.824
CM
0.580
0.770
HPWS
-
0.323
0.027
0.844
ITL
0.633
0.543
-0.307
0.854
MEN
0.190
0.443
0.267
0.214
0.837
SW
0.568
0.461
-0.258
0.497
0.147
0.797
Syavergio Avia Difaputra, Hunik Sri Runing Sawitri
Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 9, September 2024 3500
ITL3
0,574
0,431
-0,316
0,866
0,148
0,411
ITL4
0,590
0,497
-0,245
0,877
0,249
0,488
ITL5
0,608
0,527
-0,284
0,915
0,205
0,526
MEN1
0,131
0,328
0,244
0,150
0,801
0,034
MEN2
0,090
0,416
0,383
0,104
0,809
0,136
MEN3
0,197
0,411
0,122
0,196
0,868
0,183
MEN4
0,139
0,368
0,326
0,175
0,827
0,111
MEN5
0,195
0,355
0,161
0,226
0,879
0,133
SW1
0,465
0,395
-0,126
0,341
0,153
0,834
SW2
0,437
0,343
-0,288
0,370
0,021
0,737
SW3
0,457
0,362
-0,210
0,478
0,171
0,817
Based on Table 6, the cross-loading value of each construct against its latent
variable is greater when compared to the construct value for other variables. Based on the
Cornell larger criterion and cross-loading, it can be concluded that the discriminant
validity in this study has been fulfilled and indicates that the meter used can measure the
variables.
Composite Reliability dan Cronbach’s Alpha
After the validity test is met, it is necessary to continue with the reliability test. The
reliability test is used to test whether the statement of each indicator is by the conditions
in the field. Where this is assessed by two assessments, namely composite reliability and
Cronbach alpha's. These two values must have a value above 0.7. Based on this
explanation, the reliability test assessment can be seen in Table 7.
Table 7 Composite Reliability and Cronbach's Alpha Values
Cronbach's
alpha
Composite
reliability
(rho_a)
Composite
reliability
(rho_c)
BO
0.882
0.892
0.914
CM
0.852
0.907
0.878
HPWS
0.899
0.903
0.925
ITL
0.906
0.922
0.930
MEN
0.896
0.926
0.921
SW
0.711
0.714
0.839
Based on Table 7, it can be seen that all variables have composite reliability values
above 0.7. This is also the same as Cronbach's alpha value where the values are all above
0.7. This shows that each indicator is reliable or effective and also by the conditions in
the field.
Structural Model Evaluation (Inner Model)
The Effect of High-Performance Work System on Intention to Leave and Safety Workarounds:
The Role of Burnout as Mediation and Mentoring & Coping Mechanism as Moderator
Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 9, September 2024 3501
In conducting data analysis, an evaluation of the structural model is needed. This
evaluation analyzes existing values, where testing is carried out to see the influence
between constructs using R square for endogenous variables and see the significance
value of the research model.
Coefficient of Determination (R Square)
R-square is a value that is only owned by endogenous variables, which is a value
that shows the large number of exogenous variables that affect endogenous variables.
Where the R-Square value is categorized as strong if it is more than 0.67, moderate if it
is more than 0.33 but lower than 0.67, and weak if it is more than 0.19 but lower than
0.33. The R-Square calculation in this study is shown in Table 8.
Table 8
Coefficient of Determination (R-square)
R-square
R-square adjusted
BO
0.329
0.312
ITL
0.444
0.429
SW
0.375
0.359
Based on Table 8, the output of the R-square value of the exogenous variable HPWS
can have a positive influence on the BO mediation variable of 0.329 or 32.9%.
Meanwhile, the exogenous variable HPWS through the BO-mediated variable was able
to have a positive influence on the endogenous variable ITL of 0.444 or 44.4% and the
endogenous variable SW of 0.375 or 37.5%.
Path Analysis (Hypothesis Testing)
Testing the structural influence model serves to explain the variables in a study.
The estimation for the structural model relationship, namely the coefficient path,
represents the hypothetical relationship between constructs. The path coefficient is a
value to show the direction of the variable relationship, which based on the existing
hypothesis has a positive or negative direction. After determining the coefficient path, it
is necessary to see the significance by looking for the T-statistic value through the
bootstrapping procedure. The worksheet framework needs to be calculated through
bootstrapping where the hypothesis has influence or significance if it has a T-statistic
above 1.96. The next criterion used in hypothesis testing is at a significance level of 5%
the P-value value is less equal to than 0.05. The results of the bootstrapping test in this
study can be seen in Figure 2 and Tables 9,10.
Syavergio Avia Difaputra, Hunik Sri Runing Sawitri
Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 9, September 2024 3502
Figure 2. PLS-SEM Structural Model
Table 9
Direct Influence Test
Original
sample
(O)
Sample
mean (M)
Standard
deviation
(STDEV)
T statistics
(|O/STDEV|)
P values
HPWS ->
BO
-0.328
-0.334
0.067
4.914
0.000
BO ->
ITL
0.527
0.534
0.061
8.609
0.000
BO -> SW
0.433
0.431
0.092
4.681
0.000
MEN x
HPWS ->
BO
-0.352
-0.339
0.074
4.744
0.000
MEN x
BO ->
ITL
0.194
0.193
0.074
2.622
0.009
CM x BO
-> SW
0.186
0.188
0.083
2.244
0.025
Table 10
Indirect Influence Test
Original
sample
(O)
Sample
mean
(M)
Standard
deviation
(STDEV)
T statistics
(|O/STDEV|)
P values
The Effect of High-Performance Work System on Intention to Leave and Safety Workarounds:
The Role of Burnout as Mediation and Mentoring & Coping Mechanism as Moderator
Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 9, September 2024 3503
HPWS -
> BO ->
ITL
-0.173
-0.179
0.042
4.069
0.000
HPWS -
> BO ->
SW
-0.142
-0.145
0.045
3.141
0.002
High-Performance Work System Negatively Affects Burnout
Hypothesis 1 in this study is that there is a negative influence of high-performance
work systems on burnout. Table 9 above shows that the T value is 4.914 above the critical
value of 1.96 and the p-value < 0.05. Therefore, the values of T and P are qualified so
that hypothesis 1 is supported.
Burnout Positively Affects the Intention to Leave
Hypothesis 2 in this study is that there is a positive influence of burnout on
intention to leave. Table 16 above shows that the T value is 8.609 above the critical value
of 1.96 and the p-value < 0.05. Therefore, the values of T and P are qualified so that
hypothesis 2 is supported.
Burnout Positively Affects Safety Workarounds
Hypothesis 3 in this study is that there is a positive effect of burnout on safety
workarounds. Table 16 above shows that the T value is 4.681 above the critical value of
1.96 and the p-value < 0.05. Therefore, the values of T and P are qualified so that
hypothesis 3 is supported.
Mentoring Moderates the Effect of High-Performance Work Systems on Burnout
Hypothesis 6 in this study is that mentoring strengthens the negative influence of
the high-performance work system on burnout. Table 16 above shows a T value of 4.744
above the critical value of 1.96 and a p-value < 0.05. Therefore, the values of T and P are
qualified so that hypothesis 6 is supported
Mentoring Moderates the Effect of Burnout on Intention to Leave
Hypothesis 7 in this study is that mentoring weakens the positive influence of
burnout on the intention to leave. Table 16 above shows a T value of 2.622 above the
critical value of 1.96 and a p-value < 0.05. Therefore, the values of T and P are qualified
so that hypothesis 7 is supported.
Coping Mechanism Moderates Burnout against Safety Workarounds
Hypothesis 8 in this study is that the coping mechanism strengthens the positive
influence of burnout on safety workarounds. Table 16 above shows a T value of 2.244
above the critical value of 1.96 and a p-value < 0.05. Therefore, the values T and P are
qualified until hypothesis 8 is supported.
Burnout Mediates the Effect of High-Performance Work System on Intention to
Leave
Hypothesis 4 in this study is that burnout mediates the influence of a high-
performance work system on the intention to leave. Table 17 above shows that the T value
Syavergio Avia Difaputra, Hunik Sri Runing Sawitri
Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 9, September 2024 3504
is 4.069 above the critical value of 1.96 and the p-value < 0.05. For this reason, the values
of T and P are qualified so that hypothesis 4 is supported.
Burnout Mediates the Effect of High-Performance Work Systems on Safety
Workarounds
Hypothesis 5 in this study is that burnout mediates the influence of high-
performance work systems on safety workarounds. Table 17 above shows that the T value
is 3.141 above the critical value of 1.96 and the p-value < 0.05. For this reason, the values
of T and P are qualified so that hypothesis 5 is supported.
Effect of High-Performance Work System on Burnout
The effect of HPWS on burnout based on data analysis obtained a statistical T value
of 4.914 and a p < value of 0.05 which means that there is a significant negative influence
on the influence of HPWS on burnout which means that the better an organization
implements a high-performance work system in the application to its employees, the less
burnout conditions from its employees, so that H1 is supported. This explains that the
implementation of a high-performance work system has occurred at the Diponegoro
National Hospital Semarang, among others, with the organization trying to educate
employees, making it possible for employees to take formal internal courses or training,
paying employee salaries above average, having a formal career plan from the
organization for its employees, and the freedom of the organization for employees to plan
their work.
The results of this study support previous research which says that HPWS facilitates
employees to obtain critical resources to meet the job demands of employees which
reduces the level of employee burnout. (Kilroy et al., 2016). Human resource (HR)
practices that improve skills and abilities related to employee duties and development,
also reduce employee emotional fatigue. Opportunities HR practices provide
opportunities for employees to participate in decision-making and incorporate
suggestions from workers for the well-being of the organization, which increases
employees' sense of attachment and responsibility to the organization. (Jyoti et al., 2015).
It is also in line with Macky & Boxall, that HPWS has revealed a positive impact on job
satisfaction, organizational commitment, and trust in management. (Macky and Boxall,
2008), On the same hand, it is also proven that training and performance management
hurt emotional fatigue, which is part of burnout (Harley, Sargent and Allen, 2010). So it
can be concluded that a high-performance work system hurts employee burnout in the
hospital realm.
Effect of Burnout on Intention to Leave
The effect of burnout on intention to leave based on data analysis obtained a T-
statistic of 8.609 and a p < value of 0.05, which is 0.000, which means that there is a
significant positive influence on the effect of burnout on intention to leave, which means
that the higher the burnout of an employee, the higher the intention to transfer employees
from the organization so that H2 is supported. Burnout in employees of Diponegoro
National Hospital Semarang is shown in Table 7, where on average has a low value. For
The Effect of High-Performance Work System on Intention to Leave and Safety Workarounds:
The Role of Burnout as Mediation and Mentoring & Coping Mechanism as Moderator
Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 9, September 2024 3505
Intention to Leave itself, it is shown in Table 5 with a fairly low average. It can be said
that a small number of employees experience burnout, which is characterized by
employees who feel emotionally drained, tired at work, and feel that they are working too
hard at Diponegoro National Hospital Semarang has the intention to leave the
organization or intention to leave as evidenced by the fact that some employees have the
thought of changing the organization since starting to work for the company. have the
intention to not be with the company for more than a few years, and are active or planning
to look for another job in a different organization.
Burnout refers to physical, emotional, and mental exhaustion, which can leave
individuals feeling tired, frustrated, and drained of energy. (Deery, Walsh and Guest,
2011), This is considered by the conditions of several employees at the Diponegoro
National Hospital Semarang. Burnout, which is considered a very serious occupational
health hazard, is often associated with employee turnover. (Schwarzkopf et al., 2017),
Where this is also to the condition of some employees who have the intention to leave the
organization or intention to leave. Thus this states that burnout has a significant and
positive effect on ITL, which is to the statement of (Jyoti et al., 2015), and it is proven
that high levels of burnout can cause employees to think about leaving the organization
the statement from (Boyas, Wind and Ruiz, 2013).
Effect of Burnout on Safety Workarounds
The effect of burnout on safety workarounds based on data analysis obtained a T-
statistic value of 4.681 and a p < value of 0.05, which is 0.000, which means that there is
a significant positive influence on the influence of burnout on safety workarounds, which
means that the higher the burnout of an employee, the higher the intention of deviations
in the work process to overcome obstacles in the organization so that H3 is supported.
Burnout in employees of Diponegoro National Hospital Semarang is shown in Table 7,
where on average has a low value. Meanwhile, safety workarounds are shown in Table
6, which on average have a low value. It can be said that a small number of employees
experience burnout, which is characterized by employees who feel emotionally drained,
tired at work, and feel that they work too hard at Diponegoro National Hospital Semarang
to regularize the work process to overcome an obstacle in the organization or safety
workarounds which are characterized by a small number of employees who ignore safety
regulations to complete their work and do not follow safety procedures by expecting of
faster working time.
The Effect of Burnout as a Mediation Variable The Effect of High-Performance
Work System on Intention to Leave
The effect of burnout as a mediating variable from the influence of a high-
performance work system on intention to leave based on data analysis obtained a T-
statistic value of 4.069 with a p > value of 0.05, which is 0.000. This means that the
influence of the high-performance work system has a negative and significant effect on
the intention to leave through burnout, so H4 is supported. This shows that the
implementation of a high-performance work system at Diponegoro National Hospital has
fostered a culture or environment that makes its employees minimize burnout. This is
Syavergio Avia Difaputra, Hunik Sri Runing Sawitri
Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 9, September 2024 3506
shown in Table 7 where the average burnout value has a low number. Things from
minimal burnout from employees make the intention to move from the employee also
small, this is shown in Table 5 which has a low average value.
The Effect of Burnout as a Mediation Variable The Effect of High-Performance
Work System on Safety Workarounds
The effect of burnout as a mediating variable from the influence of high-
performance work systems on safety workarounds based on data analysis obtained a T-
statistic value of 3.141 with a p > value of 0.05, which is 0.002. This means that the
influence of high-performance work systems has a negative and significant effect on
safety workarounds through burnout, so H5 is supported. This shows that the
implementation of a high-performance work system at Diponegoro National Hospital has
fostered a culture or environment that makes its employees minimize burnout. This is
shown in Table 7 where the average burnout value has a low number. The things of
minimal burnout from employees make safety workarounds from those employees will
also be small, this is shown in Table 6 which has a low average value.
The Effect of Mentoring as a Moderating Variable The Effect of High-Performance
Work System on Burnout
Based on the results of the analysis, it was explained that hypothesis 6 was accepted,
which means that mentoring was proven to moderate the influence of HPWS on burnout.
In this study, an average of 3.44 points were obtained for filling out the mentoring
questionnaire which was in the medium category, which means that the average employee
of Diponegoro National Hospital thought that when mentoring was implemented, the
negative influence of the high-performance work system on burnout was stronger. This
means that in a hospital environment that conducts a high-performance work system for
its employees, the implementation of mentoring is needed because this reduces burnout
among hospital employees.
A study conducted by Qian et al., 2024 said that leaders/mentors help employees
understand how HPWS works better. In this process, mentors provide varied
career/psychosocial support, which helps reduce stress, anxiety, emotional exhaustion,
and others among employees. (Qian et al., 2014). Similarly, HPWS also gives greater
autonomy and work control to its employees to reduce their burnout rates. So, the
interaction of HPWS attributes and mentoring has a synergistic effect in reducing
employee burnout. A well-defined career path established through HPWS along with
career support from mentors, helps employees to achieve their career goals and also
reduces their career-related anxiety. Furthermore, the psychological support of mentors
at HPWS helps employees build a better work-life balance and reduce work fatigue.
The Effect of Mentoring as a Moderating Variable The Effect of Burnout on
Intention to Leave
Based on the results of the analysis, it was explained that hypothesis 7 was accepted,
which means that mentoring was proven to moderate the effect of burnout on the intention
to leave. In this study, the average value of filling out the mentoring questionnaire was
The Effect of High-Performance Work System on Intention to Leave and Safety Workarounds:
The Role of Burnout as Mediation and Mentoring & Coping Mechanism as Moderator
Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 9, September 2024 3507
3.44 which was in the medium category, which means that the average employee of
Diponegoro National Hospital thought that when mentoring was applied, it would weaken
the positive influence of burnout on intention to leave. This means that in a hospital
environment, employees who experience burnout when mentoring is applied will reduce
the desire or intention to move from the organization.
The Effect of Coping Mechanism as a Moderating Variable The Effect of Burnout
on Safety Workarounds.
Based on the results of the analysis, it was explained that hypothesis 8 was accepted,
which means that the coping mechanism was proven to moderate the effect of burnout on
safety workarounds. In this study, the average value of filling out the coping mechanism
questionnaire was 3.16 which was in the medium category, which means that the average
employee of Diponegoro National Hospital thought that when the coping mechanism was
applied, it would weaken the effect of burnout on safety workarounds. This means that in
a hospital environment, employees who experience burnout when a coping mechanism is
applied will reduce the desire to deviate from work processes or workarounds.
Conclusion
The results showed that the implementation of HPWS had a significant influence
on burnout in non-doctor employees, with a negative relationship. This means that the
better the implementation of HPWS, the lower the burnout rate experienced by
employees. An enhanced work environment through HPWS helps employees reduce
emotional fatigue that can potentially arise. Furthermore, burnout has been shown to have
a positive influence on the intention to leave. This means that when employees experience
burnout, they tend to have a greater desire to leave the company. This condition shows
that burnout plays a driving factor that makes employees feel like looking for a better
work environment. In addition, burnout also has a positive effect on safety workarounds,
which is the act of ignoring safety regulations at work. Employees who experience
emotional fatigue tend to ignore safety procedures more often to get the job done faster.
This confirms that burnout conditions can increase the risk of non-compliance with
occupational safety standards.
This study also found that burnout mediated the influence of HPWS on intention to
leave. In other words, a good implementation of HPWS can reduce burnout, which in turn
lowers the employee's intention to leave the company. HPWS creates a more supportive
work environment so that employees feel more comfortable and motivated to continue
working in the hospital. Finally, burnout also mediates the influence of HPWS on safety
workarounds. Effective implementation of HPWS can reduce burnout rates, which
ultimately minimizes the tendency of employees to ignore safety regulations. In other
words, HPWS creates a more conducive work culture, so that employees are more
motivated to follow safety procedures properly. Overall, the results of this study
emphasize the importance of implementing HPWS in managing burnout and reducing its
negative impact on employees' intention to leave the company as well as compliance with
occupational safety procedures.
Syavergio Avia Difaputra, Hunik Sri Runing Sawitri
Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 9, September 2024 3508
Bibliography
Albougami, A. S., Almazan, J. U., Cruz, J. P., Alquwez, N., Alamri, M. S., Adolfo, C.
A., & Roque, M. Y. (2020). Factors affecting nurses’ intention to leave their current
jobs in Saudi Arabia. International Journal of Health Sciences, 14(3), 33.
Cruz, J. P., Cabrera, D. N. C., Hufana, O. D., Alquwez, N., & Almazan, J. (2018).
Optimism, proactive coping and quality of life among nurses: A cross‐sectional
study. Journal of Clinical Nursing, 27(9–10), 2098–2108.
Gaudenz, C., De Geest, S., Schwendimann, R., & Zúñiga, F. (2019). Factors associated
with care workers’ intention to leave employment in nursing homes: A secondary
data analysis of the Swiss Nursing Homes Human Resources Project. Journal of
Applied Gerontology, 38(11), 1537–1563.
Han, J., Sun, J.-M., & Wang, H.-L. (2020). Do high-performance work systems generate
negative effects? How and when? Human Resource Management Review, 30(2),
100699.
Jyoti, J., Rani, R., & Gandotra, R. (2015). The impact of bundled high-performance
human resource practices on intention to leave: Mediating role of emotional
exhaustion. International Journal of Educational Management, 29(4), 431–460.
Labrague, L. J., McEnroe‐Petitte, D. M., Leocadio, M. C., Van Bogaert, P., & Cummings,
G. G. (2018). Stress and ways of coping among nurse managers: An integrative
review. Journal of Clinical Nursing, 27(7–8), 1346–1359.
Rouleau, D., Fournier, P., Philibert, A., Mbengue, B., & Dumont, A. (2012). The effects
of midwives’ job satisfaction on burnout, intention to quit and turnover: a
longitudinal study in Senegal. Human Resources for Health, 10, 1–14.
Seaman, J. B., & Erlen, J. A. (2015). Workarounds in the workplace: a second look.
Orthopaedic Nursing, 34(4), 235–240.
Sekaran, U., & Bougie, R. (2016). Research methods for business: A skill building
approach. john wiley & sons.
Sohu, J. M., Hongyun, T., Akbar, U. S., & Hussain, F. (2023). Digital innovation, digital
transformation, and digital platform capability: Detrimental impact of big data
analytics capability on innovation performance. International Research Journal of
Management and Social Sciences, 4(3), 265–281.
Zaheer, S., Ginsburg, L., Wong, H. J., Thomson, K., Bain, L., & Wulffhart, Z. (2019).
The turnover intention of hospital staff in Ontario, Canada: exploring the role of
frontline supervisors, teamwork, and mindful organizing. Human Resources for
The Effect of High-Performance Work System on Intention to Leave and Safety Workarounds:
The Role of Burnout as Mediation and Mentoring & Coping Mechanism as Moderator
Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 9, September 2024 3509
Health, 17, 1–9.