THE EFFECT OF HUMAN RESOURCE COMPETENCE AND CHARISMATIC LEADERSHIP ON FINANCIAL ACCOUNTABILITY AT ISLAMIC BOARDING SCHOOLS IN BIREUEN REGENCY

This study examined and analyzed the effect of human resource competence and charismatic leadership on financial accountability in Islamic boarding schools in the Bireuen Regency. This study used primary and secondary data. The primary data was obtained by using a questionnaire to 98 respondents in Islamic boarding schools in Bireuen Regency, and the secondary data was obtained through online news, articles, and other news sources related to the required data. In addition, researchers also conducted observations. The data analysis method used was a structural method based on variance, namely Partial Least Square (PLS). The results showed that the competence of human resources and charismatic leadership positively influenced the financial accountability of Islamic boarding schools in the Bireuen Regency.


INTRODUCTION
Pesantren is a cultural center. Therefore, it is always worthy of discussion in various fields, for example; education, social change, multiculturalism, management, and so on (Zaini, 2016). The most important element of pesantren from an educational perspective is the ustad/teacher, as a human resource and also human capital (Bukit, 2017).
The practice of accountability is very much needed in every organization, including religious organizations such as Islamic boarding schools as a form of good corporate governance (Randa et al., 2011). When an organization's accountability is good, it will affect its transparency and will improve its organizational performance (Dubnick, 2005).
Even though Islamic boarding schools have shown their role so far, Islamic boarding schools still have to struggle to face various internal and external problems. These problems include the competence of human resources who manage the finances of Islamic boarding schools (Murdayanti & Purwita, 2019;Ahyar, 2020).
Human resources as a unit of human power in an organization and not just the sum of existing employees (Matindas, 2002). While the quality of human resources has a level of knowledge (knowladge), ability (skill) and willingness (ability) that can be shown by good human resources on the quality of the resulting performance (Ruky, 2003).
The majority of human resources owned by Islamic boarding schools work as teachers or lecturers. Islamic boarding schools still have few human resources that focus on accounting and finance (Murdayanti & Purwita, 2019). Islamic boarding schools have large enough human resources and if they are managed properly, they will be trusted by stakeholders so that they can develop properly and achieve independence.
Islamic boarding schools managed by kyai who have good managerial skills will be able to mobilize all the potential of students. Research from (Tucuan et al., 2014) states that there is a positive influence between leadership style and motivation. That is, the stronger the implementation and understanding of the leader, the better his work motivation and will increase accountability.

RESEARCH METHODS
The population in this study were 162 Islamic boarding schools in Bireuen Regency. The population characteristics (sampling frame) are Islamic boarding schools that are active and have more than 500 students. The increase is the large increase in the number of students owned by the pesantren, the problems regarding accountability are increasingly complex. Based on the sampling frame, the total population is 28 Islamic boarding schools. The sampling technique used is the census method. Respondents in this study were the leaders of Islamic boarding schools, teachers and staff of accounting and financial reports who were involved in the preparation of budgets and financial reports with a total of 84 respondents.
The data used in this study are primary data and secondary data. Primary data obtained by using a questionnaire that is distributing a list of questions to respondents. While secondary data is obtained through online news, articles and other news sources that are related to the required data. In addition, this researcher also conducted by observation (observation).
Data analysis was carried out using a variance-based structural method, namely Partial Least Square (PLS). PLS was chosen because the number of samples was less than 100 and did not require an assumption test. The model in PLS has a reflective latent variable. The reflective model is the indicator is seen as a variable that is influenced by the latent variable. This results in a change in one indicator will result in changes in other indicators in the same direction. Convergent validity test, which is a test to measure the level of accuracy of indicators or dimensions through measuring the magnitude of the correlation between constructs and latent variables. To measure convergent validity, the standardized loading factor and the Average Variance Extracted (AVE) value are used.
The standardized loading factor describes the magnitude of the correlation between each indicator and its construct. The loading factor value above 0.7 is stated as an ideal or valid measure as an indicator in measuring the construct. However, in the research development stage, a loading scale of 0.50 to 0.60 is still acceptable (Ghozali and Latan, 2015).
Average Variance Extracted (AVE) is the average percentage of variance scores extracted from a set of latent variables estimated through loading standardized indicators in the PLS algorithm iteration process. The rule of thumb required as a good model if the AVE of each construct is greater than 0.5, which means that 50% or more of the variance of the indicator can be explained.
Cronbach Alpha interprets the correlation between the created scale and all existing variable scales. Cronbach alpha coefficient (α) is a measuring tool that is in great demand in social science, namely an index that describes the reliability of a scale made with all existing variable scales. A value that can be said to be reliable when the alpha coefficient value is greater than or equal to 0.60.
Composite Reliability is a reliability test that measures the real reliability value of a variable. Ghozali and Latan (2015) state that a latent variable has high reliability if the composite reliability value is above 0.70. The reliability test parameters of the PLS measurement model can be seen in the table below.
Structural Model Evaluation Coefficient of Determination (R2) Used to find out how much exogenous variables explain endogenous variables with the following criteria: 0<R2≤0.33 = low 0.34<R2≤0.66 = moderate 0.67<R2≤100 = height Q-Square Predictive Relevant Used to find out how much the overall PLS model explains endogenous variables with the formula Q2= 1-((1-R12)(1-R22))1-Ri2)) Effect Size f-Square (f2) Used to determine the size of the influence of each exogenous variable on endogenous variables with the following criteria: 0.02<f2≤0.15 = small 0.15<f2≤0.35 = moderate F2≥ 0.35 = large 3. RESEARCH RESULTS AND DISCUSSION 3.1 Descriptive data analysis Based on table 2, the human resource competency variable has an average value of 4.5 and a standard deviation of 0.510. The average value of 4.5 which is lower than the median (5) indicates that according to respondents, the competence of human resources in Islamic boarding schools is still lacking. The charismatic leadership variable has an average value of 4.49 and a standard deviation of 0.510. The average value of 4.49 which is lower than the median (4.62) indicates that according to respondents, charismatic leadership in Islamic boarding schools is not optimal. The financial accountability variable has an average value of 4.53 and a standard deviation of 0.501. The average value of 4.53 which is lower than the median (5) indicates that according to the respondents, financial accountability in Islamic boarding schools is not optimal.

Assessing the Outer Model Convergent validity
The indicator that measures the construct is said to be high if the value of the outer loading is greater than 0.7, but according to Ghozali and Latan (2015) for the initial assessment of the development of the measurement scale for the loading value of 0.5 to 0.6 is considered sufficient. In this study, the loading factor limit used is greater than 0.7.  (2021) Covergent validity can also be assessed with the Average Variance Extract (AVE), which is the average variance value of the overall indicators extracted from the latent variables. An indicator is declared to have convergent validity if the AVE value is 0.50. This means that the average variance of all indicators explained by the latent variable is at least 50%  (2021) An indicator is declared to have discriminant validity if the loading factor of each indicator that measures the latent variable is greater than the cross loading value (the correlation of the indicator with other latent variables) Based on table 5, each indicator in each research variable has a cross loading value greater than the correlation value of the indicator with indicators on other variables, so that each indicator used in this study has a fairly good discriminatory validity.
In addition to looking at the value of cross loading, discriminatory validity can also be seen from the Fornell Lecker Criterion value (root AVE (Average Variance Extracted) vs correlation between latent variables) the value must be greater than 0.5.

Composite Reliability and Cronbach's Alpha
The construct reliability test was carried out using the composite reliability measure and Cronbah's Alpha. The p[p[testing criteria if the composite reliability value 0.70 or Cronbah'Alpha 0.60 is declared the construct to be reliable.  (2021) Based on table 7, it can be seen that the Composite Reliability value of the three variables is greater than 0.70. Thus the indicators that measure the AK, KK and KSDM variables are declared reliable. Meanwhile, the value of Cronbach's Alpha for the three variables is greater than 0.60. So that the indicators that measure the AK, KK and KSDM variables are declared reliable.

Inner Model Test Structural Model Test
Testing Evaluation of the structural model is carried out to determine the level of accuracy of the structural model in relation to the predictions to be made. This test is carried out by looking at the Coefficient of Determination (R2) , the value of Q-Square predivtive relevance (Q2), and Effect Size (f2) Used to find out how much the exogenous variable as a whole explains or represents the endogenous variable. Q-Square predictive relevance calculated manually. Q 2 = 1-(1-R 2 ) Q 2 = 1-(1-0,713) Q 2 = 1-0,287 Q 2 = 0,713 Based on the calculation above, it is known that the Q-Square predictive value (Q2) is worth 0.713 or 71.3%. This shows that the diversity of the AK variable can be explained by the KK and KSDM variables of 71.3% or in other words the contribution of the KK and KSDM variables to the AK variable is 71.3%. While the remaining 28.7% is the contribution of other variables that are not included in this model.

f-Square
Used to determine whether the exogenous variable has a large, large or small effect on the endogenous variable Based on table 9, it is known that the effect of KSDM produces f2 of 0.696. This shows that the KSDM variable has a large influence on the AK. Then the effect of the KK variable on Ak produces f2 of 0.129. This shows that KK has little effect on AK.

Hypothesis test
Structural model hypothesis testing is used to test the effect of exogenous variables on endogenous variables. Hypothesis testing (significance) was carried out using t-test statistics by applying the resampling (bootstrapping) method to obtain stable t-test statistics.

Hypothesis Testing 1
Based on table 5.10 it can be seen that the statistical T value of the influence of KK on Ak is 3.295 > 1.96 and p value is 0.001 < 0.005 with a coefficient of 0.271. This means that charismatic leadership has a positive and significant effect on financial accountability.

Hypothesis Testing 2
Based on table 5.10, it can be seen that the statistical T value of the effect of KSDM on AK is 6.510 > 1.96 and p value is 0.000 < 0.005 with a coefficient of 0.631. This can be interpreted that KSDM has a positive and significant effect on AK so that the Ha2 hypothesis is accepted.

CONCLUSION
Based on the results of data analysis that has been carried out, the conclusions of this study are as follows: Charismatic leadership has a significant positive effect on financial accountability. This means that the better the charismatic leadership, the better the financial accountability of Islamic boarding schools in Bireuen Regency. Human resource competence has a significant positive effect on financial accountability. This means that the higher the competency possessed by the members of the Islamic boarding school, the higher the financial accountability of the Islamic boarding school in Bireuen Regency.
To Islamic Boarding Schools in Bireuen Regency to apply more charismatic leadership so that they become role models for their members and are able to motivate their members to work optimally in achieving organizational goals. Improving the competence of its members so that they are able to increase the accountability of Islamic boarding schools to the maximum. Improvements can be made by providing education and training on financial reports that are provided continuously. In addition, providing assistance in the financial management of Islamic boarding schools including the preparation of financial reports for Islamic boarding schools in accordance with PSAK.
For further research, it is possible to expand the research location to other Islamic boarding schools in the province of Aceh, outside the province of Aceh and outside the island of Sumatra and can compare them and add other variables such as the benefits of information technology, collaboration between institutions.