The Effect of Multiple Role Conflict and Work Stress on the Performance of Female Nurses at Permata Bunda General Hospital

ABSTRACT


INTRODUCTION
The hospital is one of the health service facilities that are often used by the community in seeking help with health problems.The success of a hospital in carrying out its functions is marked by an increase in the quality of hospital services.The quality of the hospital is greatly influenced by several factors.
The most dominant factor is human resources according to RI Law NO. 36 of 2009 concerning health, health workers are grouped according to their expertise such as medical personnel, pharmaceutical personnel, nursing personnel, community and environmental workers, nutrition workers, physical skills workers, medical technical personnel and others who support the provision of health services (Gillies , 2013).
Nursing as a profession and nurses as professionals are responsible for providing nursing services in accordance with the competencies and authorities owned independently or in collaboration with other health members.Nursing services provided in the form of a nurse's work appearance must be based on high abilities in accordance with nursing care standards so that quality can be guaranteed.
Burnout is a condition of work fatigue experienced by nurses, which is caused by personal, family and work environment factors.If burnout occurs, nursing care cannot be carried out properly,


The Effect of Multiple Role Conflict and Work Stress on the Performance … (Audilla Fahira) 163 because burnout has an impact on the financial, physical, emotional and social aspects of the profession, clients and organizations (Perrewe, 2002in Hera, 2016).
Hospital nurses are dominated by female workers.The contribution of women to economic development can be seen from the tendency of women's participation in the workforce.The participation of women today does not only demand equal rights but also states that their function has meaning for development in Indonesian society.Women's participation concerns traditional and transitional roles.Traditional or domestic roles include the role of women as wives, mothers and household managers.Meanwhile, the transitional role includes the notion of women as workers, members of society and human development (Alih Yutika, 2018).
Nurul Hudalifah (2014) conducted research related to the influence of multiple role conflicts and job stress on the performance of female nurses at Muhammadiyah Gresik Hospital with the results of the analysis showing that the dual role conflict partially had no effect on the performance of female nurses.Job stress partially affects the performance of female nurses.So it can be concluded that multiple roles and work stress simultaneously or together have an effect on the performance of female nurses.
Muhammad Iqbal ( 2016) conducted a study related to the influence of multiple conflicts and work stress on the performance of female employees at the Menggala regional public hospital.This is supported by preliminary investigations showing that the composition of employees is dominated by female employees who are married, which is 63%.With the conclusion from the results of statistical analysis for the regression coefficient value of 72.763 and a significance value of 0.000 where the value is smaller than the standard, namely 0.05.
Neneng Ngenda Arlinda ( 2019) conducted a study related to the effect of multiple conflict and work stress on the performance of female paramedics at Blud RS Konawe Selatan that the percentage for multiple role conflict variables is 76.1%, which means that respondents have high multiple role conflict, for stress variables.work is 66.9%, which means that the respondent has work stress which is the influence of multiple role conflict on performance with a significance value obtained is 0.006, that there is an effect of work stress on performance with a significance value obtained is 0.005, and the simultaneous influence of role conflict double and work stress on performance with a significance value obtained is 0.000.
Based on the above background, the writer wishes to conduct research on "The Effect of Work Stress Dual Role Conflict on the Performance of Female Nurses at Permata Bunda General Hospital Medan".The number of nurses in the hospital consists of 130 people, with the percentage of nurses being 80% women and 20% men.The results of this research are expected to be an alternative source in broadening the horizons for management science studies, especially hospital management in managing human resource management so that it can be used as a reference for future research development, and this research is expected to improve health services at Permata Bunda Hospital Medan.

RESEARCH METHOD 2.1. Place and time of research
This research was carried out in one of the public hospital companies owned by Alm.Father.H. M. Arbie, namely Permata Bunda General Hospital Medan, located at jl. Sisingamangaraja No.7, Mesjid Kec.Medan Kota, Medan City, North Sumatra.This research was conducted for 3 months from July 2020 to September 2020.

Types and Sources of Data
This research is a quantitative study using a survey method.Data consists of 2, namely, primary data taken directly from respondents which includes data from themselves about the respondent (age, gender, length of work, number of children, latest education and amount of income) through a questionnaire that is answered by the respondent, which in the questionnaire contains questions that explore aspects related to the influence of multiple role conflicts consisting of work-family and family-work conflicts, and stress on the performance of female nurses.
Respondents who answered the questionnaire list were paramedic nurses at Permata Bunda General Hospital Medan who had become permanent employees, and secondary data was data taken from data that had occurred in the human resources section of the Permata Bunda General Hospital in Medan, in the form of: number of paramedics.nurses, the entry and exit of nurses.

Population and Data Collection Techniques
The population referred to in this study are 202 nurses at Permata Bunda General Hospital Medan, while 132 female nurse paramedics at Permata Bunda Hospital Medan who are married are 132 people.In the study the sample is part of the population that has relatively the same characteristics and is considered to be representative of the population.In determining the required sample, the formula has been developed by Hair et.al (1995) in Augusty Ferdinand (2002).
The sampling technique used in this study is a census that is to make the population become respondents in the study, who become respondents.After the sample was taken by census, there were 115 paramedical nurses who were married and had children.This shows that the respondents used in this study meet the minimum sampling requirements according to SEM is between 100-200 samples.
Using this sample, it is hoped that the results of data analysis will provide a more valid picture of the respondent's condition.The reason for using nurses as samples is because nurses have unique characteristics with a profession that triggers a lot of stress, including: lots of pressure and demands from superiors to always serve patients kindly in any situation, nurses are also doctor partners who must be ready to carry out doctor's advice on the other hand, must be able to face complaints from the patient's family.If anything happens to the patient, the nurse also has to carry out many risky tasks, such as injecting, inserting catheters, sewing wounds, and caring for patients with infectious or dangerous diseases.

Method of collecting data
The data collection method used in this study is a questionnaire.The questions presented in the questionnaire are closed questions.Closed questions are made using an interval scale, to obtain data which, if processed, shows the effect or relationship between variables.The interval scale used in this study is the bipolar adjective, which is a refinement of the semantic scale with the hope that the resulting responses can be intervally scaled data (Ferdinand, 2006).
The scale used is in the interval range 1-10.The use of a scale of 1-10 (even scale) is to avoid the answers of respondents who tend to choose answers in the middle, so that it will produce responses that are gathered in the middle (gray area).The following is an overview of the scoring or value of the questionnaire in this study.

Validity Test and Variance Extract
Unidimensionality and reliability assessments are carried out for whether an Indicator has a good degree of suitability in a 1-dimensional model.Unidimensionality itself is an assumption used in calculating reliability.Reliability is a measure of the consistency of indicators in indicating a construct.Basically, the reliability test shows the extent to which a measuring instrument can provide relatively the same results when re-measured on the same subject.
There are two ways that can be used, namely by looking at construct reliability and variance extracted, both of which have a Cut Off Value, namely at least 0.70 and 0.50 respectively.However, the cut off values are not fixed numbers.

Data Analysis Techniques
Data analysis was performed using the Structural Equation Model (SEM) in modeling and hypothesis testing.SEM or structural equation modeling is a set of statistical techniques that allows testing a series of relatively complex relationships, simultaneously.(Ferdinand, 2006).What is meant by complicated is simultaneous models that are formed through more than one dependent variable at the same time acting as independent variables for other tiered relationships.In this study, two kinds of analysis techniques were used, namely: 1. Confirmatory analysis (confirmatory factory analysis) on SEM is used to confirm the factors that are most dominant in a group of variables.2. Weight regression in SEM which is used to examine how much influence between variables.
According to Ferdinand (2006) there are seven steps that must be taken when using the Structural Equation Model (SEM) modeling.A complete SEM modeling basically consists of two main parts, namely the Measurement Model and the Structural Model.Measurement Model or measurement model to confirm indicators of a latent variable as well as a structural model that describes the causality relationship between two or more variables.Structural Model is a model regarding the structure of relationships that form or explain causality between factors.

Analisis Deskriptif
Descriptive analysis in this study will describe the characteristic data of respondents based on age, education, latest, length of work, number of children, and monthly income.Descriptive analysis was also carried out to determine respondents' responses to the value of the research variables.Disclosure of descriptive analysis in the form of percentage data.
Characteristics of respondents based on the age of the respondents can be seen as many as 8.7% aged between 17-25 years, 67.0% aged between 26-35 years, 18.3% aged between 35-45 years, and the rest as much as 6.1% were at age range of 45 years and above.As can be seen from table 4.1 above, it can be seen that almost half of the female nurses who work at the Permata Bunda General Hospital in Medan are in the mature age range, namely between the ages of 26-35 years.Women in this age should experience little work stress and have have better emotional control abilities than women at a younger age.
Based on the latest education, it shows that there is a fairly even distribution at the four existing levels.The largest number of respondents were AKPER / equivalent graduates, as many as 101 people or 87.8%, followed by S1 graduates as many as 14 people or 12.2%.From the data presented in the table, it is concluded that more than half of the nurses who work at the Permata Bunda Medan General Hospital have just received education up to AKPER / equivalent, thus it can be said that in general nurses who work at Permata Bunda General Hospital Medan are included medium education category.
Based on the length of work, it is in the span of more than 6 years.This figure shows that most of the female nurses who work at the Permata Bunda General Hospital are classified as old.Generally, female nurses who work do not have sufficient work experience, so it is possible that their concentration is easily distracted while working.
Based on the number of children, it shows that respondents who have 2 children are 52 people or 45.2%, respondents who have 3 children are 38 people or 33.0%, respondents who have 1 child are 14 people or 12.2%, while the number of respondents who had 4 children was 9 or 7.8%, and the last one was the respondent who had 5 children, it was known that there were 2 children or 1.7%.This means that the female nurses at Permata Bunda Medan General Hospital mostly have 2 children.

Multivariate Validity and Reliability Test a. Reliability Test
The reliability test shows the extent to which a measuring instrument can provide relatively the same results when re-measured on the same object.The minimum reliability value of the accepted dimensions of latent variables is 0.70.Construct Reliability is obtained from the formula Hair, et.al, (1995).
Construct Reliability = ( ) () ∑∑ ∑ + jLoadingstd  Based on the observations in the table above, it appears that there is no reliability value that is smaller than 0.70.Likewise, in the variance extract test, there was also no value that was below 0.50.The results of this test show that all the indicators (observed) in the construct (work-family conflict, family-work conflict, work stress, and the performance of female nurses) used as the observed variable for the construct or the latent variable are able to explain the construct or the latent variable it forms. .

Description of Respondent Characteristics
This analysis was conducted to obtain a descriptive description of the respondents of this study, particularly regarding the research variables used.This analysis was carried out using an index analysis technique to describe the respondent's perception of the question items being asked.
Therefore, the number of respondents' answers does not start from number 0 but starts from numbers 23 to 230, the resulting index number will start from numbers 23 to 230 with a range of 207, without the number 0. By using the five box method (five box method), the range amounting to 207 divided by five, resulting in a range of 41.4 which will be used as the basis for the interpretation of the index value which in this example is as follows: 1. Index value 10.00 -28.00 = Very Low Interpretation 2. Index value 28.01 -46.00 = Low interpretation 3. Index value 46.01 -64.00 = moderate interpretation 4. Index value 64.01 -82.00 = High interpretation 5. Index value 82.00 -100 = Very High interpretation On this basis, the researcher determines the index of respondents' perceptions of the variables used in this study.

Work-Family Conflict
The family work variable was measured through 5 question items with descriptive statistics using the index number measurement technique as shown in the following table: Rata-rata total 66,14 The table above shows that from the index value range of 10-100, the average index of the family work conflict variable is High, which is 66.14%.This shows that the respondent's perception of family work conflicts is high on a per item basis indicating that work pressure is also high.
In the table, it is known that the conflict of commitment and responsibility towards family occupies the highest position in the variable of family work conflict, namely 67.39%.Then followed by the number of demands for tasks, and busy with work at 66.35%, next is work pressure where the index is 66.26%, and the last is the lack of family togetherness index reaching 64.35%.This shows that the five indicators can be used as a benchmark for family work conflict variables.

Family-Work Conflict
The family work variable was measured through 5 question items with descriptive statistical results using the index number measurement technique as shown in the following table: The table above shows that from the index value range of 10-100, the average index for the family work variable is moderate, namely 56.30%.This shows that the respondent's perception of family-work conflict is moderate.In the table, it is known that job intervention has the highest position in the family-work conflict variable, namely 59.83%.Then followed by marital pressure of 55.83%, then the lack of involvement as a parent where the index is 55.74%, the pressure index as parents reaches 55.48% and the last is the lack of involvement as a wife index of 54.61%.This shows that the five indicators can be used as benchmarks for the family-work conflict variable.

Work stress
The variable job stress is measured by means of 4 question items, the results of descriptive statistics using the index number measurement technique are as presented in the following table: The table above shows that from the index value range of 10-100, the average work stress variable index is high, which is 65.57%.This shows that the respondent's perception of job stress is high.In the table, it is known that tension and error occupy the highest position in the work stress variable, namely 67.57%.Then followed by a workload of 66.43%, then the demands / pressure from the top where the index is 65.48%, and the last is a decrease in the level of interpersonal relationship index by 62.78%.This shows that the five indicators can be used as benchmarks for job stress variables.

Nurse Performance
Nurse performance variables were measured through 9 items of descriptive statistical results using the index number measurement technique as shown in the following table: The table above shows that from the index value range of 10-100, the average index of the nurse performance variable is moderate, which is 53.84%.This shows that the respondent's perception of the nurse's performance is moderate.The table shows that effectiveness occupies the highest position in the work stress variable, namely 57.30%.Then followed by a quality of 56.52%, then commitment where the index was 55.91%, the attitude index reached 55.74%, the achievement index and productivity decreased to 54.09%, the quantity index reached 53.04%, the index the absentee level reached 51.65%, the punctuality index reached 51.30%, and the last one was late coming to work with an index of 48.96%.This shows that the five indicators can be used as a measure of the nurse's performance variables.

CONCLUSION
From the results of data processing, the value of Critical Ratio (CR) on the relationship between work-family conflict variables and work stress is 2.796 with a P (Probability) of 0.005, while the Critical Ratio (CR) value on the relationship between family-work conflict variables has a significant positive effect.to the occurrence of work stress for female nurses of 3.081 with P (Probability) of 0.002, the value of Critical Ratio (CR) on the relationship between job stress variables has a positive significant effect on the performance of female nurses of 2.639 with P (Probability) of 0.008.
Critical Ratio (CR) on the relationship between work-family conflict variables has a negative effect on the performance of female nurses of -2.514 with a P (Probability) of 0.012, the value of Critical Ratio (CR) on the relationship between variables, while the Critical Ratio (CR) value on the relationship Between the family-work conflict variable has a negative effect on the performance of female nurses of 3.234 with P (Probability) of 0.001.

Table 1 .
Variable Reliability Test Results the amount of variance of the extracted indicator by the developed latent variable.The acceptable extract variance value is a minimum of 0.50.The equation for getting the extracted variance value is: Variance Extracted = ∑∑ ∑ + jLoadingstd Loadingstd ε 2 2 The overall results of the reliability and variance extract tests are presented in Table 2 below:  Journal of Economics and Business (JECOMBI), Vol. 1, No. 2, January 2021: 162 -170

Table 2 .
Reliability Test and Extract Variance

Table 3 .
Work-Family Conflict Index  The Effect of Multiple Role Conflict and Work Stress on the Performance … (Audilla Fahira)

Table 4 .
Family-Work Conflict Index

Table 5 .
Job Stress Index

Table 6 .
Nurse Performance Index