ANALYSIS OF INTEREST IN SAVING IN ISLAMIC BANKING WITH A SOCIAL ENVIRONMENT AS A MEDIATOR IN THE RELATIONSHIP OF RELIGIOSITY AND STUDENT KNOWLEDGE

Saving is a good habit when it comes to financial arrangements, including for college students. The rapid development of sharia economy makes the emergence of various Islamic banking products for this saving activity. This is of course beneficial for students who should have started making financial arrangements. Although the interest in saving is still lacking, the use of Islamic bank products is quite profitable for their future. The purpose of the study was to analyze the factors that influence students' interest in saving in Islamic banking. Linear regression is an analysis technique with hypothesis testing to support the conclusions on RStudio. The results of analysis show that the direct impact between religiosity and knowledge of the social environment has a significant effect simultaneously, although religiosity cannot be a predictor of partial saving interest. While the results of the track analysis showed that social environment was able to be a mediator for religious relationships to saving interests, but not for knowledge relationships towards student saving interests in Islamic banking.


INTRODUCTION
Saving is one of the positive activities for all walks of life, as well as for students who still have to prepare for their future. This is driven by several factors, one of which is interest or belief. Confidence is a considerable motivation in encouraging individuals to consume a product (Andespa, 2017a). Interest is the liking or inclination of the heart to an attention or desire (Pusat Bahasa Kemdikbud, 2016). A mental device consisting of a mixture of feelings, expectations, the establishment of prejudices or other tendencies that lead the individual to a certain choice (Musruroh, 2015). Interest always develops along with the basic needs of the community as well as in terms of saving.
Along with the higher public interest in saving, the more varied banking products, both conventional and sharia. Banking is everything related to the bank, covering institutions, business activities and ways of carrying out its business activities. The main function of Indonesian banking is to collect and distribute public funds and aims to support the implementation of national development in order to increase equitable development and its results, economic growth and national stability, towards improving the standard of living of many people. (Otoritas Jasa Keuangan, 2016). Indonesia itself, which is a developing country, has a very rapid banking development. This can be seen both in bank office facilities and all banking products offered to the public, conventional and sharia. Islamic banks are financial institutions that carry out intermediary functions in collecting public funds and distributing financing to the public in accordance with sharia principles (Fajar Mujaddid, 2019).
Banking is everything related to a bank. The main function of Indonesian banks is as a collector and distributor of public funds and aims to support the implementation of national development (Fajar Mujaddid, 2019;Otoritas Jasa Keuangan, 2016). As a developing country, Indonesia has a very rapid banking development. This can be seen both in bank office facilities and all banking products offered to the public, conventional and sharia. Islamic banks are financial institutions that carry out intermediary functions in collecting public funds and distributing financing to the public in accordance with sharia principles (Fajar Mujaddid, 2019). This makes various Islamic banking products run based on a system that is far from usury'. This is one of the things that affects the increase in interest in the use of Islamic banking products in the community, as well as among students.
The concept of Islamic sharia-based finance has grown into a trend in the world economy, including in Indonesia. Based on Islamic banking data, the growth of conventional banks is smaller than that of Islamic banks where Islamic banks have experienced relative growth of around 40% per year in the last ten years while conventional banks are 20% Indonesian (Direktorat Perbankan Syariah, 2011). Along with the development of time, it is possible that the growth of Islamic banking will exceed the conventional development system. Based on the Indonesian Sharia Finance Development Roadmap (Otoritas Jasa Keuangan, 2018), The development of Islamic finance has produced various achievements, from the increasing number of products and services, to the development of infrastructure that supports Islamic finance. Even in the global market, Indonesia is among the top ten countries that have the largest Islamic financial index in the world. Nevertheless, the growth of Islamic finance has not been able to keep up with the growth of conventional finance. This can be seen from the market share of Islamic finance, which as a whole is still below 5%. However, when viewed from each type of Sharia product, until the end of December 2016, there were several Sharia products whose market share was above 5%, including Islamic banking assets amounting to 5.33% of all banking assets, state sukuk which reached 14.82% of the total government securities outstanding, Sharia financing institutions amounting to 7.24% of total financing, special Islamic financial service institutions amounting to 9.93%, and Islamic microfinance institutions at 22.26%. Meanwhile, Sharia products whose market share is still below 5%, include corporate sukuk in circulation amounting to 3.99% of the total value of sukuk and corporate bonds, net asset value of Islamic mutual funds of 4.40% of the total net asset value of mutual funds, and sharia insurance of 3.44%. In addition to the financial products above, shares of issuers and public companies that meet the criteria as Sharia stocks reach 55.13% of the stock market capitalization listed on the Indonesia Stock Exchange. The figures mentioned above show that Indonesia's Islamic finance still needs to be developed so that it can keep pace with the growth of conventional finance in order to grow the financial industry as a whole.
Based on the description above, this study aims to analyze the determinants of students' interest in saving in Islamic banking. Widya Gama Lumajang Institute of Technology and Business is one of the universities located in Lumajang Regency. Along with the development of the sharia concept in the field of economics, this Private University made several curriculum adjustments by adding several courses that are in accordance with the concept of sharia economics. Research assumes that the basis for the use of sharia-based products is religion, but for students, knowledge should also be the basis for decision making. Therefore, this study was conducted to determine the influence of religiosity and knowledge on the interest in saving in Islamic banking with the social environment as the mediator. These variables are considered the most basic factors in determining students' interest in saving in Islamic banking.

METHODS
This research uses quantitative concepts with a survey approach where the research does not make changes or there is no special treatment of research variables. The purpose of this study is to explain the causal relationship between one variable and another through hypothesis testing (Setyobakti, 2018). This method involves active students of the Accounting Study Program at the Widya Gama Lumajang Institute of Technology and Business for the 2019/2020 academic year as research objects. Sampling was carried out using the purposive sampling method with the criteria being (a) active students and (b) students already mastering sharia economics courses so that 39 respondents were obtained as samples in this study.
The data were collected using a research instrument in the form of a questionnaire with a likert type measurement scale. After the data were collected, the questionnaire was tested using testing research instruments using validity and reliability. Validity testing using data processing through RStudio with product moment correlation. Meanwhile, the reliability measurement method used is the Alpha Cronbach concept by showing the extent of respondents' consistency in answering the instruments assessed (Tavakol & Dennick, 2011). Furthermore, to determine the influence of each variable in this study,hypothesis and path analysis tests were used on RStudio.

RESULTS AND DISCUSSION
The collection of research data was carried out by providing questionnaires to research respondents. Furthermore, the collected data is analyzed with the help of the computer program RStudio version 1.4.1103.

Instrument Test
The research instrument test consists of validity and reliability tests. Validity tests are carried out to find out to what extent an item in the questionnaire can dig up the necessary data or information. The result of this test is said to be valid if the correlation value of each item is more than 0.5 . Meanwhile, reliability tests are carried out to measure the extent to which the proposed questionnaire can provide results that are no different from using the Cronbach Alpha concept.
The results of instrument testing are given in Table 1, which shows that all statement items on each variable have a correlation value of more than 0.5, so it can be concluded that each item in the questionnaire used to dig up data is valid so that it can dig up the necessary data or information. Meanwhile, the results of the questionnaire reliability test in this study showed that all variables were very reliable as shown in the results of Cronbach's Alpha > 0.7. So it can be concluded that all the measuring concepts of each variable from the questionnaire used in this study are reliable questionnaires.

Hypothesis Test
The linear regression equation in data processing in this study consists of: These three structural equations have met the classical assumption that the normality test using the Kolmogorov-Smirnov concept with the ks.test menu, pmulticollinearity test is carried out by ols_vif_tol on packages(' olsrr') and heteroskedasticity is carried out using the Breusch-Pagan test on packages(' lmtest'). While the coefficient of determination (R Square) in equation (1) is 0.5859 equation (2) of 0.7415, y and equation (3) is 0.6027, which means that for independent variables in each equation of the regression structure is quite strong in affecting its dependent variables.
The results of the hypothesis test show the signification of the research regression model. Equation (1) shows that the partial test p-value values are 0.276 and 7.79e-07 for each independent variable of the equation. From these results, it is concluded that variables X 1 do not have an effect on variable Y while on the contrary variables have a significant effect. Furthermore, for X 2 F-statistical testing, a p-value of 1.28e-07 was obtained, whose value was less than 0.05 so that a significant model was concluded simultaneously.
Equation (2) shows the relationship between the variable independen and the mediator directly. In hypothesis testing, p-values were obtained of 4.69e-10 and 0.105 for each independent variable of the equation (2). From these results, it is concluded that the variable X 1 has no effect on variable Z while on the contrary variable X 1 has a significant effect. Furthermore, for F-statistical testing, a p-value of 2,663e-11 was obtained, whose value was less than 0.05 so that a significant model was concluded simultaneously Equation (3) shows the relationship between variables in this study. In the hypothesis t test, p-values were obtained of 0.731, 3.97e-06 and 0.232 for each independent variable equation (2). From these results, it is concluded that the variable X 1 and have no effect on variable Y while on the contrary variable X 2 have a significant effect. Furthermore, for F-statistical testing, a p-value of 3,669e-07 was obtained, whose value was less than 0.05 so that a significant model was concluded simultaneously

Path Analysis
This analysis is used to test a model of the relationship between variables in the form of causation. Through this path analysis, it will be determined which path is the most appropriate and short for an independent variable to a dependent variable. This analysis uses hypothesis test analysis and the concept of Sobel Test on packages('bda') Equation (1) shows the direct influence between religiosity and knowledge on the interest in saving. Based on equation (1) it is known that religiosity has no significant effect on the interest in saving with a pvalue of 0.276 (>5%). Meanwhile, knowledge has a p-value of 7.79e-07 (<5%) so it is concluded that knowledge has a significant effect on the interest in saving partially. However, the results of the hypothesis test simulation simultaneously showed religiosity and knowledge had a significant effect on saving interest with a p-value of 1.28e-07 Equation (2) shows the influence of religiosity and knowledge on the social environment which is the mediator in this study. In the case of the application of the mediator, the independent variable should have a significant effect on the mediator. This must happen if you want the mediation process to take place in accordance with the initial assumptions of this study. The results of the simulation of equation (2) showed that religiosity had a significant effect on the research mediator, namely the social environment, with a pvalue of 4.69e-10 (<5%). Meanwhile, knowledge has no significant effect on the social environment with a large p-value of 0.105 (>5%). Hypothesis tests show simultaneously, religiosity and knowledge have a significant effect on the social environment as mediators in this study.
Equation (3) shows the influence of the mediator on the dependent variables of this study. The simulation results showed that the social environment did not have a partial significant effect on saving interest with a p-value of 0.232 (>5%). The same thing happens for religiosity where the p-value is 0.731 (>5%), while for knowledge it has a partial significant effect with a p-value of 3.97e-06 (<5%). The results of simultaneous hypothesis tests for religiosity, knowledge and social environment showed a significant influence with a p-value of 3,669e-07 (<5%) Based on the simulation results, it was found that there was a mediation process in the research model, where the social environment mediated the relationship between independent and dependent variables in this study. This is inferred based on the p-value of equation (1) which increases in equation (3) (Baron, R. M., & Kenny, 1986). Furthermore, the results of the Sobel Test show that variable X 1 with Z as a mediator variable that produces a p-value of 0.04307284 so that it can be concluded that variable Z can mediate the relationship between variables X 1 and Y. Whereas in variable X 2 where with a p-value of 0.1429638 it is concluded that Z cannot mediate variables X 2 and Y because the p-value is less than 0.05

Discussion
The simulation results in the previous subsection were used as a basis for the discussion of hypotheses in this study. The first hypothesis states that religiosity and knowledge have a significant effect on students' interest in saving in Islamic banking. This hypothesis refers to equation (1) which shows the direct influence between religiosity and knowledge on students' saving interests. Based on the results of the hypothesis test, it was concluded that religiosity did not have a partial effect on students' interest in saving. This is different from knowledge where with a p-value of 7.79e-07 (<5%) it is concluded that it has a significant positive effect on students' interest in saving. However, although the partial test has different conclusions for both independent variables, the results of the simultaneous test state that there is a significant influence on the interest in saving from religiosity and knowledge. This explains that for the religiosity of the object of study is not enough to make him increase interest in saving in Islamic Banking. However, this will have a significant effect if it is accompanied by knowledge in the process of learning the object of research. Together, religiosity and knowledge encourage the object of research to increase interest in saving in Islamic banking in a positive manner. This is in line with previous research on (Wahyuning, 2021) where religiosity is not capable of being a predictor for students' saving interests.
The second hypothesis states that the social environment can mediate the relationship between religiosity and knowledge to the interest in saving. This hypothesis is proven by the results of a path analysis that states that there is a mediation process in this independent and dependent variable. Based on the results of the sobel test, it was concluded that the social environment can be a mediator for religiosity relationships to students' interest in saving. However, different things are concluded for the relationship of knowledge to students' saving interests. The social environment is not able to mediate the relationship between knowledge and interest in student savings. This is also supported by equation (2) where knowledge cannot partially be a predictor for the social environment who is the mediator in this study. The results of this study are in line with the research (Rahmawati, 2019) and (Lestari, 2015) where the higher the student's religiosity value, it will encourage him to seek knowledge related to Islamic banking, both things that are prohibited and those allowed by religion. In this study, it was proven that knowledge variables are able to mediate religiosity towards the interest in saving in Islamic banking. The same is true of the social environment variables where (Andespa, 2017b) states that the cultural and family environment has a significant effect on the interest in saving. In general, social environments such as families, communities and schools will always contribute thinking to every individual decision. Likewise for students where they are social creatures who will always interact with their surroundings.