Determinant of Student Loyalty with Satisfaction as Intervening Variables

Reader Impact Factor Score
[Total: 2 Average: 4.5]

Published on International Journal of Economics & Business
Publication Date: May, 2020

Ayunda Kumala Nusantarawati, Diah Yulisetiarini & Deasy Wulandari
The Student of Management Department, Economics and Business Faculty, University of Jember
The Lecturer of Management Department, Economics and Business Faculty, University of Jember
Jember, East Java, Indonesia

Journal Full Text PDF: Determinant of Student Loyalty with Satisfaction as Intervening Variables (Studied of Student Loyalty, Jember University).

Abstract
The research design in this study is an explanatory research that is to explain the status of the variables studied and to determine the relationship between one variable with another variable. The sample size required was 379 students who were randomly selected in each faculty according to their proportions. The foundation of theory and previous research that has been done is the basis of the conceptual framework in this study. Price variables and image variables have an influence on customer satisfaction (students) and customer loyalty (students), besides customer satisfaction (students) have an influence on customer loyalty (students). The test results show that there is a price effect on satisfaction with an Estimate value of 0.167, the influence of the two variables is ascertained from the CR value of 2,329 with a significance level of less than 0.05 which is equal to 0.020. The positive sign of the test results shows if the price increases, the satisfaction of buying education services will also increase with a significant increase. The test results show that there is an influence of Image on aging with an Estimate value of 0.645. The influence of the two variables is ascertained from the CR value of 6.287 with a significance level of less than 0.05, which is equal to 0.000. education services will also increase with a significant increase. The test results show that there is an influence of Loyalty on Prices with an Estimate value of 0.131, the influence of the two variables is ascertained from the CR value of 1,996 with a significance level of less than 0.05 which is equal to 0.046. A positive sign from the test results shows if the price increases, the loyalty of buying education services will also increase with a significant increase. The test results show that there is an influence on the Image of Loyalty with an Estimate value of 0.287 the influence of the two variables is ascertained from the CR value of 2.138 with a significance level greater than 0.05 that is equal to 0.032. buying education services will also increase with a significant increase. The test results show that there is an effect of satisfaction on Loyalty with an Estimate value of 0.348, the influence of the two variables is ascertained from the CR value of 2.120 with a significance level of less than 0.05, amounting to 0.034. The positive sign of the test results shows that if satisfaction increases, the loyalty of buying education services will also increase with a significant increase. With these results, all research hypotheses are accepted.

Keywords: Price, Image, Loyalty & Satisfaction.

1. Introduction
Jember University is one of the State Universities in Indonesia and the only State University in Jember. The University of Jember was established in 1964 based on the Decree of the Minister of PTIP Number: 151 of 1964 on 09 November 1964. At the time of its establishment, the University of Jember had five faculties namely (1) the Faculty of Law located in Jember with its branches in Banyuwangi; (2) Faculty of Social and Politics located in Jember; (3) The Faculty of Agriculture is located in Jember; (4) The Faculty of Economics is located in Banyuwangi; (5) The Faculty of Literature is located in Banyuwangi. In 2018 Jember University already has 15 faculties consisting of 5 social faculties and 10 exact faculties which in total have 100 study programs.
Jember University as a State University in 2018 proposed the determination of BLU status, namely State Universities with a pattern of financial management of public service bodies that refer to the Law of PT, PP no. 74 of 2012 jo PP No. 23 of 2005 with the technical guidelines of the Minister of Finance Decree on the proposal Minister of Research and Technology and Higher Education. The purpose of the submission is due to the rapid development of the University of Jember which is marked by the increasing number of study programs, the increasing number of students, lecturers and education personnel, as well as the amount of the budget that is managed. The phenomenon of the rapid development of the University of Jember is expected to provide opportunities for all levels of society to pursue tertiary education at the University at prices in the form of UKT / Single Tuition in accordance with the ability of parents. In addition to the price, the rapid development of the University of Jember is expected to also influence the image of the University of Jember so that with a good image the satisfaction and loyalty of students towards the University of Jember can increase.
Jember University version of the Ministry of Research, Technology and Higher Education in the last 4 years, namely the year from 2015 to 2018 had experienced a very significant increase in 2017 ranked 20th, while in 2018 it was ranked 22th out of 100 universities in Indonesia. The main components used to assess the performance of Indonesian tertiary institutions include 5 (five) main components. (1) HR quality, which includes the percentage of the number of S3 educated lecturers, the percentage of the number of head professors and professors, and the ratio of students to lecturers; (2) Institutional Quality, which includes the accreditation of institutions and study programs, the number of internationally accredited study programs, the number of foreign students, and the number of university collaborations; (3) Quality of Student Activities, which include student performance; (4) Quality of Research and Community Service, which includes research performance, community service performance, and the number of scopus indexed scientific articles per number of lecturers. (5) Quality of innovation, which includes innovation performance.
Jember University is one of the State Universities (PTN) which is most in demand by the community to study tertiary education. According to the Minister of Research, Technology and Higher Education in 2019 Jember University is one of the ten (10) most popular State Universities in Indonesia. Based on the top 10 ranking of the high drop in demand, the impact on student interest from 2014/2015 to 2018/2019 was very volatile and experienced a significant increase in 2018/2019 with an interest of 105,206 students, received 8,567 students, and those who registered were 7242 students. This condition shows that Jember University is a tertiary institution that has quite a large amount of interest among prospective students.

2. Literature Review
2.1 Prices
Kotler and Keller (2009: 72) say many economies assume that consumers are recipients of prices and accept prices at face value or as given. Marketers realize that consumers often process price information actively, translate prices based on their knowledge from previous experience, formal communication (advertisements, telephone sales, and brochures), informal communication (friends, colleagues, or family members), point of purchase or online resources or other factors. Lupiyoadi (2016: 210) said that the determination of prices in education services is very important because educational institutions depend on tuition fees and other service fees for operational activities, around 80-90% of operational budget, while others come from donated income, as well as other financial assistance.
2.2 Image
The term image has been gaining in popularity since the 1950s, expressed in various contexts such as image of the organization, image of the company, national image, image of the brand or brand image, public image, self-image and so on (Buchari Alma, 2018: 379 ). Image of a college, formed based on the elements gathered from the components. The components that have been studied are “Academic reputation, campus appearance, cist, personal attention, location, distance from home, graduate and professional school preparation, career placement, social activities, program of study, and size”. The components that ultimately form the image, namely academic reputation or academic quality, campus appearance, cost, location, distance from residential homes, possible future careers, social activities in Higher Education (Alma. 2018: 382).

2.3 Customer Satisfaction
Satisfaction or dissatisfaction is a feeling of pleasure or disappointment from someone who comes from the comparison between his impression of the real / actual product performance with the expected product performance. Vinna Sri Yuniarti (2015: 234) argues that if a product or service that is expected by consumers is not in accordance with the reality, it is certain that consumers will not feel satisfied. Conversely, if the product is suitable or better than consumers expect, consumers will feel satisfied. Sangadji and Sopiah (2013: 180) state that from a managerial perspective, maintaining and / or increasing customer satisfaction is very critical, therefore, managers must look at programs that must increase customer satisfaction as an investment. This level of satisfaction / dissatisfaction will also be influenced by consumers’ evaluations of exchange equity, as well as by their attribution of product performance.

2.4 Customer Loyalty
Understanding consumer loyalty according to Sangadji and Sopiah (2013: 104) consists of several definitions, including “loyalty is defined as non-random purchases expressed over time by some decision making units.” Based on these definitions, it can be explained that loyalty refers more to the manifestation of the behavior of decision-making units to make continuous purchases of goods or services from a selected company. Loyalty can also be defined as a customer’s commitment to stay in depth to re-subscribe or re-purchase selected products or services consistently in the future, even though the influence of the situation and marketing efforts have the potential to cause behavior change. Customer loyalty in the context of service marketing as a response is closely related to pledges or promises to uphold commitments that underlie relationship continuity, and are usually reflected in ongoing purchases from the same service provider on the basis of dedication and pragmatic constraints. Another definition of customer loyalty is customer commitment to a store brand, or supplier, based on a very positive attitude and is reflected in consistent repeat purchases. The conclusion based on these definitions shows that loyalty is more directed at a behavior, which is indicated by routine purchases and is based on the decision making unit.

3. Methodology / Methodology
3.1 Research Design
The research design in this study is explanatory research that is to explain the positions of the variables studied and to determine the relationship between one variable with another variable (Sugiono, 2003: 100). The purpose of this study using explanatory research is to determine the effect of price and image factors on the satisfaction and loyalty of Jember University students.

3.2 Research Population and Samples
Population is the whole subject or totality of research subjects that can be people, objects, or something that can be obtained and or can provide information (data) research. In other words the population is the whole object of research (Johar Arifin, 2017: 7). The population in this study were students of the University of Jember in 2018 with 7242 students. The research sample is part of the whole object studied and is considered to represent the entire population (Johar Arifin, 2017: 7). The sampling technique in this study uses the Slovin formula as follows:

Where :
n = sample size
N = population size
e = percent leeway inaccuracy due to sampling errors that can still be tolerated or desirable for example by 5% or 1% (for this study taken 5%).
The sampling method used in this study is proportional stratified random sampling, namely the number of samples obtained based on the size of the sample divided by the proportion of existing stratification (Johar Arifin. 2017: 8).

3.2 Types and Data Sources
There are 2 types and sources of data used in this study, namely:
1. Primary data, is data obtained directly from sources or respondents by distributing questionnaires to students.
2. Secondary data, is data obtained indirectly in the form of documents in the collection and processing that are used as support and complement research. This secondary data is in the form of documentary evidence obtained from Jember University.

3.2 Data Collection Methods
Various data collection methods that can be done in a study, in this study the data collection method used is by distributing questionnaires / questionnaires. Questionnaire is a method of data collection which is done by giving a set of questions or written statements to respondents to be answered. Questionnaires are a more efficient method of data collection if the researcher already knows for certain the variables to be measured and knows what is expected of the respondent. The distribution of questionnaires was distributed to students of the University of Jember in 2018 according to the number of samples in each faculty.

3.3 Identification of Variables
Variable identification are the variables that will be analyzed in this study, grouped into three variables, namely:
a. Exogenous variables or independent variables or causal variables in their position are not affected by other variables in the model. In this research, the exogenous variables are price (X1) and image of Jember University (X2).
b. Intervening endogenous variables, in this study that satisfaction (Y1) is an experience felt by students while studying at Jember University.
c. Endogenous variables, in the study are loyalty (Y2) in the form of student loyalty to the institution, namely Jember University.

3.4 Data Analysis Methods
1. Variable Measurement Scale
The scale of measurement of variables in this study is to use a Likert scale. William G. Zikmund and Barry J. Babin (2013) say that the Likert scale is a measure of attitude made so that respondents can judge how strongly they agree or disagree with statements made carefully, ranging from very positive attitudes to very negative to some objects.
2. Test Data Validity
According to Mohamad Dimyati (2009: 115-116) a validity test was conducted to see which items were appropriate (representative) to be used to represent the independent variables used. This test was carried out using confirmatory factor analysis on each of the latent variables, namely price, image, satisfaction, and loyalty using the AMOS version 5.0 program. The indicators of a variable are said to be valid if they have a significant loading factor at α = 5%. The research instrument is called valid unidimensional if it has a goodness of fit index (GFI) ≥ 0.09.
3. Reliability Test
The reliability test in this study aims to show how much the measurement can give relatively no different results if the measurements are taken again on the same subject (Mohamad Dimyati. 2009; 116). The instrument is said to be reliable if the measuring instrument gets consistent results, so that this instrument can be used because it can work well at different times and different conditions. This reliability test uses the following formula:

Construct reliability ≥ 0.70 indicates that the reliability of the construct is good, while reliability with 0.60 – 0.70 can still be accepted with the validity requirements in a good model (Ghozali 2014).

3.5 Structural Equation Model (SEM) Analysis
SEM (Structural Equation Modeling) is one type of multivariate analysis that is formed through more than one dependent variable explained by one or several independent variables, where a dependent variable at the same time acts as an independent variable for other known tiered relationships as an intervening variable and a moderating variable.
According to Mohamad Dimyati (2009: 58) SEM model has an advantage compared to other multivariate statistical methods because in the latent variable the measurement error (error = e) is included in the model. Its ability to present a comprehensive model together with its ability to confirm the dimensions of a concept or factor and its ability to measure the influence of relationships that theoretically exist. SEM technique allows testing several dependent variables at once with several independent variables.

3.6 Assumption Test of Structural Equation Model (SEM)
Testing SEM assumptions has the aim to ensure that the regression equation obtained can have accuracy in estimation, unbiased and consistent. After doing the validity and reliability tests on each latent variable, the assumption test is performed to see whether the required prerequisites have been fulfilled in SEM modeling. The prerequisites are:
1. Test for normality
Normality test needs to be done both for normality of univariate and multivariate data where several variables are used at once in the final analysis. According Ghozali (2005: 128) normality assumption testing can be done by using z statistical values for skewness and kurtosis and empirically can be seen Critical Ratio (CR). If a significance level of 5% is used, then the CR value is between -1.96 to 1.96 (-1.96 ≤ CR ≤ 1.96) then it can be said that the data is normally distributed, both univariate and multivariate.
2. Multicollinity Test
Multicollinearity test can be seen through the determinant of the covariance matrix. According to Ghozali (2005: 131) a very small or near zero determinant value indicates an indication of multicollinearity or singularity problems so that this data cannot be used for research.
3. Test Outliers
Outliers are observations that emerge with extreme values both univariate and multivariate, which arise because of the combination of unique characteristics that are possessed and look very much different from other observations. The criteria used are based on the value of Chi Squares on the degree of freedom of the number of indicator variables at the significance level of p≤0.05 (Ghozali. 2005: 130).

4. Results and Discussion
The research design in this study is explanatory research that is to explain the position of the variable under study and to determine the relationship between one variable with another variable. The sample size required was 379 students who were randomly selected in each faculty according to their proportions. The foundation of theory and previous research that has been done is the basis of the conceptual framework in this study. Price variables and image variables have an influence on customer satisfaction (students) and customer loyalty (students), besides customer satisfaction (students) have an influence on customer loyalty (students).

4.1 Analysis of Test Validity and Reliability
The construct of the Confirmatory Regression technique is called the latent variable (Variables that cannot be measured directly) and the indicator is the observed variable (The observed variable, as the operationalization of the measurement of latent variables). After the constructs and indicators are stated in the equation, then using the Confirmatory Regression procedure the validity and reliability tests are performed.
A test or instrument can be said to have high validity if the device performs a function of measurement or provides a measurement result in accordance with the measurement objectives. A valid measurement tool, not only able to disclose data accurately and provide a careful picture of the data. Careful means that the measurement is able to provide a picture of the smallest differences between subjects with one another. The validity of a data if the factor loading of the indicator variable has a value above 0.50, it can be said that the question items as constituents of unobserved variables in confirmatory regression are valid (Ghozali, 2005: 26)
In addition to being valid, the instrument must also be reliable. The instrument is said to be reliable if the measuring instrument obtains consistent results. Thus this instrument can be used safely because it can work well at different times and different conditions. So, religiosity is not different, it is measured again on the same subject. Reliability in this study was calculated using composite (construct reliability) with a cut of value of 0.60. (Ghozali, 2008: 233).
The formula used to calculate construct reliability is as follows:

4.1.1 Test the Validity and Reliability of Price Constructions
Price is a latent variable measured by three indicators, namely study programs (X1.1), parental abilities (X1.2), and State Universities (X1.3). Based on the test results with the confirmatory factor analysis presented in table 1 Test the validity test found that the value of t (determined by the value of CR) for loading each variable value is greater than the critical value at a significance level of 0.05, likewise the probability value is smaller than α (0 , 05) so that it can be concluded that all variables are significantly related to construct Price (all indicators are valid). Contruct-Reliability The price of 0.714 is above the recommended value, construct reliability ≥ 0.70 indicates that the reliability of the construct is good, while the reliability with 0.60 – 0.70 is still acceptable with the condition of validity in the good model (Ghozali 2014), with thus all price indicators are reliable.

Table 1 Test Results Validity and Reliability of Price Constructions

Source: Attached data processed

4.1.2 Test Validity and Reliability of Image Constructions
Image is a latent variable measured by three indicators: 1) Visual (seen): nice and magnificent campus buildings, comfortable classrooms, adequate sports and arts facilities (X2.1); 2). Verbal (spoken / oral): promotion, publication, personal information submitted by Jember University to the public (X2.2); 3). Beehave (community attitude), is the department / study program that is most preferred / favored by the community especially students (X2.3).
Based on the test results with the confirmatory factor analysis presented in table 2, the validity test shows that the t value (determined by the CR value) for loading each variable is greater than the critical value at the 0.05 significance level, as well as the probability value is smaller than α (0 , 05) so that it can be concluded that all variables are significantly related to the Image construct (all indicators are valid). Contruct-Reliability price of 0.837 is above the recommended value, construct reliability ≥ 0.70 indicates that the reliability of the construct is good, while reliability with 0.60 – 0.70 is still acceptable with the validity conditions in a good model (Ghozali 2014), with thus all Image indicators are reliable.

Table 2: Test Results of Image Validity and Reliability of Constructions

Source: Attached data processed

4.1.3 Test the Validity and Reliability of the Satisfaction Constructions
Satisfaction is a latent variable measured by three indicators namely 1). Competence of study programs, is the accreditation status of study programs owned by the study program (Y1.1); 2). The quality of academic programs, which are the qualities of lecturers, curriculum / syllabus and SISTER applied by Jember University (Y1.2); 3). Student interaction with administrative staff, is the relationship and communication between Jember University students and the admin staff serving them (Y1.3).
Based on the test results with the confirmatory factor analysis presented in Table 3 Test the validity test found that the value of t (determined by the value of CR) for loading each variable value is greater than the critical value at a significance level of 0.05, likewise the probability value is smaller than α (0 , 05) so that it can be concluded that all variables are significantly related to the construct of satisfaction (all indicators are valid). The construct-reliability decision of 0.646 is above the recommended value, construct reliability ≥ 0.70 indicates that the reliability of the construct is good, while reliability with 0.60 – 0.70 can still be accepted with the condition of validity in a good model (Ghozali 2014), with all satisfaction indicators are reliable.

Table 3: Results of Validity and Reliability Test for Satisfaction Constructions

Source: Attached data processed

4.1.4 Test the Validity and Reliability of the construct of Loyalty
Loyalty is a latent variable measured by three indicators, namely 1). Saying positive things, is a good word about the University of Jember (Y2.1); 2). Recommendation, which is providing information and giving recommendations to others to study at Jember University (Y2.2); 3). Willingness to rejoin institutions, is a willingness to pursue higher education at Jember University (Y2.3).
Based on the test results with the confirmatory factor analysis presented in table 4 Test the validity test found that the value of t (determined by the value of CR) for loading each variable value is greater than the critical value at a significance level of 0.05, as well as the probability value is smaller than α (0 , 05) so that it can be concluded that all variables are significantly related to the construct of loyalty (all indicators are valid). The construct-reliability decision of 0.746 is above the recommended value, construct reliability ≥ 0.70 indicates that the reliability of the construct is good, while reliability with 0.60 – 0.70 is still acceptable with the validity requirements in the good model (Ghozali 2014), with thus all loyalty indicators are reliable.

Table 4: Test Results of Loyalty and Validity and Reliability Test

Source: Attached data processed

4.2 Data Analysis
4.2.1 Regression Analysis with Confirmatory Approach
At this stage will be discussed regarding the suitability of the model test and a significant test of causality. The test results with the amos version 5.0 program provide a regression model with a confirmatory approach as shown in Figure 1 below which shows the effect of Price, Image, and satisfaction variables on student loyalty continuing study at Jember University.

4.2.1 Confirmatory Regression Assumption Test
After testing the validity and reliability of each latent variable, the assumption test is performed to see whether the prerequisites needed in confirmatory regression modeling can be met. The next requirement that must be fulfilled is the assumption of normal multivariate, absence of multicollinearity or singularity and outliers.
1. Normality Test
Normality test needs to be done both for normality of univariate data and multivariate normality in which several variables are used at once in the final analysis. To test the presence or absence of normality assumptions, it can be done with the z statistic for skewness and kurtosis empirically can be assessed on the Critical ratio (CR) used a significance level of 5%, then different CR values between -1.96 to 1, 96 (-1.96≤CR≤1.96) is said to be normally distributed data, both univariate and secivariously multivariate (Ghozali, 2005: 128). The normality test results obtained a CR value of 1,910 which means that CR is between -1.96 to 1.96. So that it can be stated that multivariate data is normal. In addition, there is also a normal univariate data shown by all critical ratios of all indicators located between -1.96 ≤ CR ≤ 1.96.
2. Multicollinearity Test
Multicollinearity can be seen through the determinant of the covariance matrix. The determinant value is very small or close to zero, indicating an indication of multicollinearity or singularity problems, so that the data cannot be used for research (Tabachnick and Fidell, 1998, in Ghozali, 2005: 131). The multicollinearity test results give a determinant of sample covariance matrix value of 1,002. The value is far above zero so that it can be concluded that there is no multicollinearity and singularity problem in the analyzed data, so the data can be continued in this study.
3. Test Outliers
Outliers are observations that emerge with extreme values both univariate and multivariate, which arise because the combination of unique characteristics that are owned and look very much different from other observations, if outliers occur, special treatment can be made on the outliers provided that the outliers are known. Detection of multivariate, outliers is done by taking into account the value of Mahalnobis Distance. The criteria used are based on the chi-square value of the degree of freedom of the number of indicator variables at the significance level of p≤0.05 (Ghozali, 2005: 130). Outliers test results appear on expensive d-squared aerobics. To calculate the mahalanobis distance value based on the value of chi squares at free degree 12 (number of indicator variables) at the level of p <0.05 (X2 0.05) is 21,026 Outliers test results are presented in (attachment) observations fathest from the centroid ) Group number 1. Outliers test results as presented in (appendix) shows that none of the cases has a mahalanobis distance value, so it can be concluded that there are no multivariate, outliers in the analyzed data.

4.2.2 Model Suitability Test
Testing the model in Confirmatory Regression aims to see the suitability of the model. The results of testing the suitability of the model in this study are presented based on Table 5 on the eight criteria used to assess whether a model is feasible or not, and all criteria are met. So it can be said that the model is acceptable, which means there is a compatibility of the model with the data. The model test results are shown in Table 5

Table 5 Model Conformance Index

Source: Appendix

Based on Table 5, it is known that from the eight criteria used to assess the suitability of a model, there is only one criterion that is declared marginal, but overall it can be said that the suitability of the model is good and acceptable, which means there is a fit between the model and the data.

4.2.3 Hypothesis Testing
Testing the research hypothesis will be based on the results of estimation of structural models.

Table 5 Hypothesis Testing

Source: Appendix

The test results show that there is a price effect on satisfaction with an Estimate value of 0.167, the influence of the two variables is ascertained from the CR value of 2,329 with a significance level of less than 0.05 which is equal to 0.020. The positive sign of the test results shows if the price increases, the satisfaction of buying education services will also increase with a significant increase.
The test results show that there is an influence of Image on aging with an Estimate value of 0.645. The influence of the two variables is ascertained from the CR value of 6.287 with a significance level of less than 0.05, which is equal to 0.000. education services will also increase with a significant increase.
The test results show that there is an influence of Loyalty on Prices with an Estimate value of 0.131, the influence of the two variables is ascertained from the CR value of 1,996 with a significance level of less than 0.05 which is equal to 0.046. A positive sign from the test results shows if the price increases, the loyalty of buying education services will also increase with a significant increase.
The test results show that there is an influence on the Image of Loyalty with an Estimate value of 0.287 the influence of the two variables is ascertained from the CR value of 2.138 with a significance level greater than 0.05 that is equal to 0.032. buying education services will also increase with a significant increase.
The test results show that there is an effect of satisfaction on Loyalty with an Estimate value of 0.348, the influence of the two variables is ascertained from the CR value of 2.120 with a significance level of less than 0.05, amounting to 0.034. The positive sign of the test results shows that if satisfaction increases, the loyalty of buying education services will also increase with a significant increase. With these results, all research hypotheses are accepted.

4.2.4 Discussion
1. Effect of Price on Purchasing Satisfaction
The results of this study tested the price hypothesis influencing the satisfaction of study students at Jember University. The test results show that the Estimate value of 0.167. The influence of these variables is ascertained from the CR value of 2,329 with a significance level of less than 0.05 which is equal to 0.020. A positive sign on the Estimate value means that the Price variable influences student satisfaction, this shows that the Price is an integrated learning plan as a guideline for conducting academic and / or professional education. The existing study programs are organized on the basis of a curriculum aimed at allowing students to master knowledge, attitudes, and behaviors / skills that are in accordance with the stated goals. The accreditation status of a tertiary institution is a reflection of the performance of the tertiary institution concerned and illustrates the quality, efficiency, and relevance of a study program held, with the existence of an accreditation status, students will be more confident and confident to continue their studies at the tertiary institution.
The results of the study are consistent with Laurencia Retha Christi’s research findings (2003) that Price has a significant and dominant influence in choosing cellular operator services in the city of Jember. These results also support the theory of Henry Assael (1995), Kotler’s theory (2000), and Louden’s theory (1993), and Kasali’s theory (1999), which states that Price has a significant effect on consumers’ decisions to buy a product or service.

According to Indah (2012) The results of testing the hypothesis of the Effect of Price (X1) on Satisfaction (Z) show that the linear regression equation formed is Z = 3.325 + 0.715X1. The correlation between price and satisfaction is 0.676 which is located in a strong interval (number> 0.50 – 0.75) so it can be concluded that the strength of the linear association (relationship) between the price variable and the satisfaction variable is strong. The effect given by the price variable on the satisfaction variable is 45.7% while the remaining 54.3% can be influenced by other factors besides the price. T count value of 9.085 with a significance of 0.05 (5%) and see the t table for the 2-party test adjusted for the value of df (degree of freedom) = n-2 = 100-2 = 98 that is equal to 1.98. So it can be concluded that the value of t arithmetic> t table, which is 9.085> 1.98. This means that Ho is rejected and Ha is accepted so that it can be said that there is a significant effect between price (X1) on satisfaction (Z).

2. Effect of Image on Purchasing Satisfaction
The results of this study tested the price hypothesis influencing the decision of study students at Jember University. The test results show that the Estimate value of 0.645. The influence of these variables is ascertained from the CR value of 6,287 with a significance level of less than 0.05 which is equal to 0,000. A positive sign on the Estimate value means that the image variable influences student satisfaction, this shows that the image in the world of education is reflected as behavior during studying at the University of Jember. Behavior is a form of effort in educating prospective students who have a calm and authoritative personality that will be accepted by students while studying at Jember University.
The results are in accordance with Kho research findings (2018) Brand image has a positive effect on customer satisfaction of TX Travel through perceived value as a moderating variable. The value of the loading factor of the direct influence of brand image to customer satisfaction shows 0.073, which means it is greater than the indirect effect through the perception of the company’s value of (0.241 X 0.255 = 0.061). While the indirect effect of brand image on customer satisfaction through the perception of the value of tour leaders is greater than the direct effect of brand image on customer satisfaction (0.337 X 0.740 = 0.249).

3. Effect of Price on Purchasing Loyalty
The results of this study test the price hypothesis to affect the student loyalty at the University of Jember. The test results show that the Estimate value of 0.131. The influence of these variables is ascertained from the CR value of 1,996 with a significance level of less than 0.05 which is equal to 0.046. A positive sign on the Estimate value means that the variable Price influences student loyalty, this shows that Price is an economic value that will be spent for study at Jember University. Prices make a determinant for the continuity of student studies, so prices become part of determining the purchasing power of one consumer, namely students.
The findings are in accordance with the findings of the study according to Indah (2012) The results of testing the Effect of Price (X1) on Loyalty (Y) produce a simple linear regression equation formed Y = 4.690 + 0.926X1. The correlation between price and loyalty is 0.654 which is located in a strong interval (number> 0.50 – 0.75) so it can be concluded that the strength of the linear association (relationship) between the price variable and the loyalty variable is strong. The effect given by the price variable on the loyalty variable is 42.8% while the remaining 57.2% can be influenced by other factors besides the price. T value of 8.555 with significance of 0.05 (5%) and see the t table for the 2-party test adjusted for the value of df (degree of freedom) = n-2 = 100-2 = 98 that is equal to 1.98. So it can be concluded that the value of t count> t table, which is 8.555> 1.98. This means that Ho is rejected and Ha is accepted so that it can be said that there is a significant effect between price (X1) on Loyalty (Y).

4. Effect of Image on Purchasing Loyalty
The results of this study tested the Image hypothesis influencing the study of students at the University of Jember. The test results show that the Estimate value of 0.287. The influence of these variables is ascertained from the CR value of 2,138 with a significance level greater than 0.05 which is equal to 0.032. A positive sign on the standardized value means that the Iamge variable influences student loyalty, this shows that Image is the delivery of information to the public whose role is to attract students to choose Jember University. Jember University formed the Image in various ways including: visiting high school and vocational schools, making brochures and banners, exhibiting, placing advertisements in newspapers. This shows that the higher the intensity of promotion in delivering information about the University of Jember, it will affect the increase in student loyalty through education at the University of Jember.
The results of the study support some of the results of previous studies conducted by Dyah Ayu Anisa Pradipta (2012). From the results of research conducted to support the positive influence of user image and product image of consumer loyalty. But there is one conflicting dimension which is the dimension of company image. In previous studies stated that the company’s image has no effect on consumer loyalty, which is a comparison from previous research with current research. In the research that has been conducted at this time stated that the dimensions of company image have a positive and significant effect on consumer loyalty.

5. Effect of Loyalty on Purchasing Satisfaction
The results of this study tested the Loyalty hypothesis effect on student satisfaction at the University of Jember. The test results show that the Estimate value of 0.348. The influence of these variables is ascertained from the CR value of 2.120 with a significance level greater than 0.05 which is equal to 0.034. A positive sign on the standardized value means that the variable Loyalty affects student satisfaction, this shows that Loyalty is the delivery of information to the public whose role is to attract students to choose the University of Jember. Customer loyalty is the core goal of Jember University which is pursued by marketers. This is because with customer loyalty as expected, it is certain that Jember University will reach students according to the target capacity of each faculty. Another impact is that Jember University is able to maintain its business (existence). Loyalty is now the belle of universities in maintaining the education business, supported by several factors, one of the main factors is the factor of customer satisfaction, to get customer satisfaction, the need for supporting factors, including quality of service to customers, customer value (customer value) and product advantage. These three factors are closely interrelated and are able to create business profits which also have an impact on the existence of higher education businesses.
The research results support some of the results of previous studies conducted by Rina Rahmawati (2014) Customer loyalty is customer loyalty to a product by trusting the product and using it repeatedly. Customer loyalty (customer loyalty) is the strength of the relationship between a person’s relative attitude (individual’s relative attitude) and repeat business (repeat patronage). Loyalty can be fulfilled because of customer satisfaction. Customer satisfaction is created due to several factors, including 1) the quality of service to customers, 2) customer value, 3) product advantage. Customer loyalty becomes the belle of entrepreneurs especially entrepreneurs engaged in food. Customer loyalty, will have an impact on the benefits obtained, so that it will also affect the company’s survival (the company’s existence). With customer loyalty, companies in the food sector will be able to survive. This is consistency with the aim of a business standing, which is getting profit (return) and gaining existence or the business can survive in any condition.

5. Conclusions
The price has a significant effect on satisfaction where the more competitive UKT prices at Jember University will further increase the satisfaction of students carrying out their studies at Jember University. This is based on the results of confirmatory regression analysis testing which shows the first hypothesis is accepted which states the price has a significant effect on satisfaction.
The price has a significant effect on loyalty where the more competitive UKT prices at Jember University will increase student loyalty towards Jember University. This is based on the results of confirmatory regression analysis testing which shows the second hypothesis is accepted which states the price has a significant effect on loyalty.
Image has a significant effect on satisfaction, the better the image of Jember University will increase the satisfaction of students of Jember University. This is based on the results of testing the confirmatory regression analysis which shows a third hypothesis is accepted which states the image has a significant effect on satisfaction.
Image has a significant effect on loyalty, the better the image of Jember University will increase the loyalty of students to the University of Jember. This is based on the results of testing confirmatory regression analysis which shows the fourth hypothesis is accepted which states the image has a significant effect on loyalty.
Satisfaction has a significant effect on loyalty where the higher the satisfaction of students, the higher the loyalty of students to the University of Jember. This is based on the results of confirmatory regression analysis testing which shows the fifth hypothesis is accepted which states satisfaction has a significant effect on loyalty.

6. References
Amoon Jawaid Austin, Sara Pervaiz. 2017. The Relationship Between ‘Student Loyalty’ and ‘Student Satisfaction’ (A case of College / Intermediate Students at Forman Christian College). European Scientific Journal. January / SPECIAL / ISSN edition: 1857 – 7881 (Print) e – ISSN 1857- 7431 pp 100-117.
AWH. Sallaudin, SM Farid, M. Ishamuddin, S. Athirah, A. Rohaizan, BR Rosla. 2018. Work Framework Proposed Between The Quality Of Services, Image Of Company, Satisfaction And Loyalty In Higher Tvet University. Journal of Fundamental and Applied Sciences. ISSN 1112-9867 pp 2473-2486.
Buchari Alma. 2018. Marketing Management and Marketing Services, Bandung: Alfabeta.
Mamta Sangadji and Sopiah Etta. 2013. Consumer Behavior. Yogyakarta: Andi
Ferdinand and Agusty. 2002. Structural Equation Modeling in Management Research. Issue 2. Semarang: BP Diponegoro University.
Johar Arifn. 2017. SPSS24 for Research and Thesis, Jakarta: PT Elex Media Komputindo.
Kotler, P. 2005. Marketing Management in Indonesia. Book 2. Adapted by A.B. Susanto. 2001. Jakarta: Four Salemba.
Kotler and Keller. 2007. Marketing Management. Volume 1 Twelfth Edition. Jakarta: Index.
Leonnard, Heny K.S Daryanto, Dadang Sukandar, Eva. Z. Yusuf. 2014. The Loyalty Model of Private University Student. International Research Journal of Business Studies Vol 7 No. 1.ISSN 2089-6271 pp 55-68.
Lujain Mohammed Hasan, Ridzuan bin Masri. 2013. Factors Influence the Satisfaction of International Students at Private Universities in Malaysia. International Journal of Science and Research (IJSR) ISSN (Online) Volume 4 Issue 8, pp. 136-142.
Lupiyoadi Creepers. 2013. Competency Based Marketing Management Services. Jakarta: Salemba Empat.
Masoodul Hassan, Saad Hassan, Muhammad Saqib Nawaz and Ibrahim Aksel. 2013. Measuring Customer Satisfaction And Loyalty Through Service Fairness, Service Quality And Price Fairness Perception: An Empirical Study Of Pakistan Mobile Telecommunication Sector. Science International. 25 (4), 971-980, ISSN 1013-5316 pp 971-980.
Muhammad Adam. 2018. Marketing Management Services. Bandung: Alfabeta.
Muhammad Ali Baber, Muhammad Sualeh Khattak. 2017. The Impact of Relational Dynamics on Students ‘Loyalty and the Mediating role of Students’ Satisfaction in the Higher Education Sector. Journal of Accounting & Marketing Volume 6. Issue 1. 1000215.ISSN: 2168-9601 pp 1-5.
M. Dimyati. 2009. SEM Analysis in Test the Influence of Several Variables on the Loyalty of Research Based Research on Small Business Credit Debtors. Jakarta: Mitra Wacana Media.
Nabil El-Hilali, Sara Al-Jaber, Lina Hussein. 2015. Students’ Satisfaction and Achievement and Absorption Capacity in Higher Education. Procedia – Social and Behavioral Sciences. doi: 10.1016 / j.sbspro.2015.02.384 pp. 420 – 427.
Parasuraman. A. 2002. Service Quality. Coral Gables: University of Miami, Monterey Library Assessement and Banchmarking Institute.
Sam Thomas. 2011. What Drives Student Loyalty in Universities: An Empirical Model from India, International Business Research Vol. 4, No. 2 pp. 183-192.
Sanjai K. Parahooa, Heather L. Harveyb, Rana M. Tamimc, Hamdan Bin Mohammed. 2013. Factors Influencing Student Satis Faction In Universities In The Gulf Region: Does Gender Of Students Matter ?. Journal of Marketing for Higher Education http://dx.doi.org/10.1080/08841241.2013.860940 pp 1-20.
Sugiyono 2013. Business Research Methods. Bandung: Alfabeta
Tina Shahsavar, Frantisek Sudzina. 2017. Student satisfaction and loyalty in Denmark: Application of EPSI methodology. PLOS ONE 12 (12): e0189576. https://doi.org/10.1371/journal.pone.0189576 December 14, 2017 pp1-18.
Tjiptono Fandy. 2005. Marketing Services. 1st edition. Malang: Bayumedia.
Vinna Sri Yuniarti. 2015. Consumer Behavior. Bandung: Faithful.
William G. Zikmund and Barry J. Babin. 2013. Exploring Marketing Research. Jakarta: Salemba Empat.
Yulisetiarini Diah. 2014. The Relationship Between Service Quality, Customer Satisfaction and Loyalty in Restaurant Business in East Java. International Journal of Business and Management Invention Issue. Vol.3 Issue 5 (version 3) pp 1-10.
Yulisetiarini Diah. 2016. The Effect of Relationship Marketing Toward Customer Satisfaction and Cutomer Loyalty on Franchised Retails in East Java. Menditteranean Journal of Social Sciences MCSER Publishing Rome Italy. Vol 7 No. 1 January pp 333-339.
Yulisetiarini Diah, Arnis Budi Susanto. 2018. Relationship Marketing: Satisfaction and Loyalty on Expedition Customers in Indonesia. International Journal of Business and Management Science (IJBMS). ISSN 1985-692X. Vol. 8 No. 3 pp. 567-575.
Yulisetiarini Diah, Yongky Ade Prahasta. 2019.The Effect Of Price, Service Quality, Customer Value, And Brand Image On Customers Satisfaction Of Telkomsel Cellular Operators In East Java Indonesia. International Journal Of Scientific & Technology Research Volume 8, Issue 03.ISSN 2277-8616 pp 5-9.
Yunia Wardi, Abror, Okki Trinanda. 2018. The Marketing of Higher Education: Managing Student Loyalty Based on Tuition Fee Policy and Service Qualit. Journal of Business and Management, Volume 19, No. 2, September 2018, p. 101-108. ISSN: 1412 – 3681 (printed), ISSN: 2442 – 4617 (online), Website: http://journal.feb.unpad.ac.id/index.php/jbm.