Establishment of Word of Mouth in Mediation of Customer Satisfaction

Ignatius Septo Pramesworo^*1 and Tiolina Evi²

¹Management Department, Faculty of Economic and Business, Perbanas Institute Jakarta, Indonesia

²Accounting Department, Faculty of Economic and Business, Perbanas Institute Jakarta, Indonesia

INTRODUCTION

The quality of human resources must have competitive advantages in undergoing the era of globalization and modernization which is always growing. One of the things that becomes a means of developing the quality of human resources is education. Education which is one of the important factors in improving the quality of human resources is needed by the community from day to day. This is one of the main causes of the proliferation of institutions that offer higher education with the same study program or offer new study programs to the public. The emergence of these higher education institutions has triggered increased competition between universities, both public and private, to get users of educational services.

In Indonesia, private universities (PTS) are one of the formal educational institutions that have duties and responsibilities from the Government of the Republic of Indonesia in preparing graduates from high school or equivalent schools. Prospective students certainly try to find private universities that match their interests and talents. In addition, there are many factors that influence prospective students to be interested in entering a certain PTS. One of them is the quality of the PTS itself. This quality factor is influential because there is an assumption that some private universities tend to be weak in providing quality services to their customers, especially students. In addition, there are private universities that are unable to comply with the quality, such as the requirements, which have been set by the regulations of the Government of the Republic of Indonesia. Therefore, it can be said that private universities must try to compete in finding prospective students and satisfying students and complying with the rules set by the government.

In this regard, private universities must be able to make prospective students and their students as customers. This is in accordance with the statement by (Akareem, H. S., & Hossain, S. S. (2016) that universities are currently considered service companies like other profitable and/or unprofitable organizations. Therefore, universities should be able to group and target their market, namely students. (Bunce, L. et al., 2017) added that globally, students studying in colleges have been defined as customers by governments since the introduction of student tuition fees.

Data for Higher Education Service Institutions (LLDikti) in Indonesia has now replaced the role of Coordinator for Private Universities (Kopertis). LLDikti has been also in charge of State Universities (PTN) since July 2018. LLDikti stated that the number of PTS universities in 2018 was 325 universities in the Regional Kopertis III. Universities in Kopertis Region III can be seen in table 1 with the following details:

Table 1: Data on the Number of Private Universities LLDikti Region III

Source: LLDikti Region III, Jakarta

Based on the PTS data at LLDikti Region III above, PTS inevitably have to find and use the right way to maintain service quality and maintain their brand image which in the end can satisfy and even increase the number of students. Various ways have been implemented by PTS to get prospective students and retain students by maintaining and improving the quality of higher education services. PTS hopes that by providing and improving the quality of services provided to its customers, customers (students) get satisfaction in the teaching and learning process during college. Satisfied PTS students will be happy to share information and even invite others to register for the PTS. This shows that positive Word of Mouth (WoM) is a powerful way to promote and eventually get students.

PTS develops like a service industry and changes all the time along with the globalization process, therefore PTS needs to be marketed. In accordance with the marketing interests of the education sector industry, one of the PTS marketing targets is prospective students and PTS students themselves. According to (Kundu, S., & Sundara Rajan, R. 2016), the most effective promotion is through positive word of mouth (WoM). Furthermore, (Sijabat, R. 2017) states that a satisfied customer will become a more effective and convincing spokesperson for a product or service. Customer satisfaction, in this case students, will not happen or impossible without the excellent service provided by PTS. (Shao-Chang, L. 2013) asserts that a satisfied customer will be the spokesperson for a product more effectively and convincingly than any type of advertisement. (Naz, F. 2014) adds that positive word of mouth (WoM) can clearly be the key that directs potential customers to the place he suggests.

PTS in maintaining and increasing student satisfaction is influenced by various things. Therefore, IKPIA Perbanas or Perbanas Institute as a private university must be smart in getting prospective students. Table 1 below shows the number of new student admissions for IKPIA Perbanas Jakarta, Faculty of Economics and Business, Regular Class, Accounting and Management Undergraduate Study Program (SI) from the 2009/2010 Academic Year to the 2018/2019 Academic Year.

There are many things that have been done by Perbanas Institute in improving itself both from within and from outside. For example, to overcome competition with other private universities in finding increasingly stringent prospective students, Perbanas Institute has opened new study programs, awarded scholarships, and provided facilities and infrastructure. Perbanas Institute has implemented a number of strategies in facing the competition for new student admissions, including: renovating the canteen, renovating the lobby, providing a student center, and opening the Perbanas Institute Bekasi branch. In addition, Perbanas Institute has made improvements to the facilities supporting the teaching and learning process, one of which is renovating the facilities of the Institute's Academic Secretariat. Perbanas Institute has also provided a pricing strategy. The pricing strategy carried out by Perbanas Institute is in the form of tuition fees by offering payment schemes. Perbanas Institute has also implemented world-class services according to Quality Management ISO 9001: 2008 starting in August 2012, by routinely conducting an ISO 9001:2008 Survillance Assessment every 6 (six) months by an external auditor from the United Registrar of System (URS). Another thing that is not less important is that the Perbanas Institute has received an A Accreditation from the Government of Indonesia, in this case the National Accreditation Board for Higher Education (BAN-PT).

Perbanas Institute wants and always tries to provide good quality education services. Quality services for students are always improved by providing facilities to support the teaching and learning process. There are many factors for Perbanas Institute to be able to maintain and improve the quality of its services so that students can be satisfied and then want to become one of the consumers who use services. Based on the quotations above, in meeting customer satisfaction, namely students, a university must be able to apply the concept of marketing in running its college, namely by providing quality services and maintaining the brand image of the university. The satisfaction of these service users will have an impact on increasing customer satisfaction, which in turn will have an impact on the willingness of students to do positive WoM. Brand image and quality of service to students are important because students and the public judge the progress and the pros and cons of a college can be judged from these two things.

LITERATURE REVIEW:

Service Quality:

(Zeithamel, et al., 2006; Green, P. 2014) state that service quality is the extent to which a service meets or exceeds customer expectations. Furthermore (Parasuraman et al., 1985; Islam, R. et al., 2015) explain that in Service quality there is a difference or gap between customer expectations and service perceptions. So service quality is based on the premise that the level of service quality perceived by customers can be measured between the customer's expectations and what they actually receive from the provider.

These services depend on the type of product and differ across organizations. Services may be defined in different ways depending on the area in which the term is used. (Keller, K. L. 2009; Lalla, N. et al., 2015) define service as an intangible act or performance offered by one party to another that does not result in any ownership. Apart from that, service can also be defined as an intangible offer by one party to another in exchange of money for pleasure.

(Mmutle & Shonhe 2017) state quality is one of the things that consumers look for in an offer. One of the offerings is service. Previously, (Keller, K. L. 2009) also stated that quality can be defined as the totality of features and characteristics of a product or service that depend on its ability to meet stated or implied needs. It is clear that quality is also related to the value of the offering, which can evoke satisfaction or dissatisfaction on the part of users.

The measurement of service quality has been illustrated as on the nature of the difference between the service expected and the service perceived by the customer. On the one hand, if expectations are greater than perceptions, the perceived service quality is unsatisfactory and customer dissatisfaction is said to occur. On the other hand, if the expectation is less than the perception, the perceived service quality is said to be satisfactory and will tend towards the ideal quality with an increasing positive difference between the expected and perceived service quality.

In order for a company's offerings to reach customers, there is a need for service. These services depend on the type of product and differ across organizations. Service can be defined in many ways depending on the area in which the term is used. (Keller, K. L. 2009) define a service as "any intangible act or performance that one party offers to another that does not result in the ownership of anything. Overall, a service can also be defined as an intangible offering by one party to another in exchange of money for pleasure.

(Bunce, L. et al., 2017) stated that quality is one of the things that consumers look for in an offer, of which service is one. Furthermore, (Sultan & Yin Wong 2012) state that quality can also be defined as the totality of features and characteristics of a product or service that depend on its ability to meet stated or implied needs. It is clear that quality is also related to the value of the offering, which can evoke satisfaction or dissatisfaction on the part of users.

(Zeithaml et al., 1990; Mmutle & Last Shonhe 2017) defines service quality as the extent to which customers' perceptions of services meet and/or exceed their expectations. Thus service quality can be the way in which customers are served in an organization which can be good or bad. (Parasuraman 1988; Jain & Aggarwal 2015) defines service quality as "the difference between customer expectations and service perceptions". Furthermore, it is said that measuring service quality can be done by looking at the difference between perceived and expected service. This is a valid way and can allow management to identify gaps in what they offer as a service.

The SERVQUAL scale which is also known as the gap model by (Parasuraman, et al., 1988; Shahin 2006) has proven to be one of the best ways to measure the quality of services provided to customers. This service evaluation method has proven to be consistent and reliable. (Mmutle & Shonhe 2017) argue that when the perceived or experienced service is less than the expected service; this implies an unsatisfactory quality of service; and when the perceived service is more than the expected service, the clear conclusion is that the service quality is more than satisfactory. The idea of SERVQUAL seems best suited to evaluating service quality from a customer perspective. This is also due to the "felt" and "expected" elements of the service delivered. Customers or consumers can be seen as people who will or are consuming services.

Brand Image:

(Deheshti, M. et al., 2016) stated that brand image is a set of tangible and intangible properties, such as ideas, beliefs, values, interests, and features that make it unique. Furthermore, (Keller, K. L. 2009) states that visually and collectively, a brand image must represent all internal and external characteristics that are able to influence how a brand is perceived by the target market or customers. In other words, brand image can be interpreted as a number of beliefs about the brand.

This is in accordance with the idea that companies usually build organizational strategies starting from establishing a brand or brand as self-identity in offering products or services to the public. A brand or brand according to (Phillip, K., & Keller, K. L. 2012; Naatu, F. 2016) is a name, term, sign, symbol or design, or a combination of all of them that is used to identify the products or services of one seller or group of sellers and as a differentiator from competitors' products or services. The brand will later be managed by the company in order to create a positive image that can add value to the benefits of a product or service.

(Keller, K. L. 2009; Ansary, A., & Nik Hashim, N. M. H. 2018) says that brand image is a series of associations or perceptions that exist in the minds of consumers towards a brand. This series of associations or perceptions is usually organized and forms a meaning for a person. Someone who has experience or gets enough information on a product or service will have a strong relationship with the product or service. Images or associations represented in perceptions can reflect objective reality or not. The image formed from the association (perception) is what underlies the person to become a customer and then tells the good and bad of a product or service to others. In addition, he or the customer can also recommend a product or service to buy or not.

(Peter and Olson 2002; Wijaya, B. S. 2013) state that brand image includes knowledge and belief in brand attributes which are expressed in cognitive aspects. The use of brands and appropriate use situations, as well as evaluations affect feelings and emotions. Things related to brands and feelings are perceived as affective aspects.

(Peter and Olson 2002; Wijaya, B. S. 2013; & Labrecque, L. I. et al., 2013). suggests a brand image which is defined as consumer perceptions and preferences for brands as reflected by various brand associations that exist in consumers' memories. Although brand associations can occur in various forms, they can be divided into performance associations and image associations related to brand attributes and advantages.

According to (Ansary, A., & Nik Hashim, N. M. H. 2018) brand image is a representation of the overall perception of the brand and is formed from information and past experiences of the brand. Brand image is associated with attitudes in the form of trust and preference for a brand. Consumers who have a positive image of a brand will be more likely to provide positive information about a product or service and provide good recommendations for it. Even some customers make purchases or repurchases.

If a brand image has been formed in a person, in this case the customer, then the perception of the brand image of the product or service will be difficult to change. Usually the image of the product or service that is formed has advantages compared to similar products or services. (Maurya, U. K., & Mishra, P. 2012) reiterate the definition of brand image as a form of perceived identity of a product or service offered to customers that can distinguish the company's products or services from competitors' products or services formed by a name, word, sign, symbol or design, or a combination of these.

(Wijaya, B. S. 2013) summarizes the dimensions of brand image based on sources developed and elaborated by experts which can be depicted in Figure 1. This image summarizes the results of studies of various literatures and relevant research, it can be concluded that the dimensions that influence and shape the image of a brand.

Image is availabe at PDF file

Source: (Wijaya, B. S. 2013)

Figure 1: Brand Image

Customer Satisfaction:

(Cronin & Taylor 1992; Lemon, K. N., & Verhoef, P. C. 2016) stated that customer satisfaction can be conceptualized as a special meaning of the transaction. It is based on customer experience at specific service encounters. Furthermore, (Wu, H. C. et al., 2015) suggests that customer satisfaction is cumulative based on an evaluation of the overall service experience. This means that customer satisfaction is based on the experience with the service provider as well as the outcome of the service.

(Phillip, K., & Keller, K. L. 2012; Duy, P. N. N., & Hoang, T. M. 2017) gives the idea that customer satisfaction is the level of a person's feelings which is the result of a comparison between the performance appraisal/product end result in relation to customer expectations. (Oliver 1997; Pham, T. S. H., & Ahammad, M. F. 2017) suggests that customer satisfaction is a form of appreciation or assessment of a process of utilizing a product or service that has been received based on customer subjectivity. Therefore, companies in carrying out their business must be customer-oriented. This means that the company must provide all services and provide quality products or services based on the needs and expectations of users so that users will get satisfaction.

According to (Phillip, K., & Keller, K. L. 2012; Awoke, H. M., & Wollo, S. 2015) satisfaction is a feeling of pleasure or disappointment felt by a person after he compares the perception of performance or or the results of a product with his expectations. So, in satisfaction there are two major elements that are compared, namely the performance or results of a product and customer expectations. On the other hand, customer expectations are also influenced by several factors, such as experience, other people's opinions, as well as the company's promises. Further, (Dominici, G., & Guzzo, R. 2010) add that companies that are able to read and pay attention to customer needs will make companies with their products or services closer to their customers. Companies should also maintain relationships with their customers and thus companies will look for ways to be able to compete with their competitors.

(Ngo, V. M. 2015) states that the most widely used method by companies to measure customer satisfaction is surveys.

Word of Mouth:

(Huete-Alcocer, N. 2017) states that word of mouth (WoM) communication is communication from person to person or between the sender of the message and the recipient of the message. Usually the recipient of the message receives the message in a non-commercial way about a product, service, or brand. WoM becomes a reference that shapes customer expectations.

According to (Voyer, P. A., & Ranaweera, C. 2015) WoM is a customer's compliments, recommendations and comments about their experiences with services and products that influence customer decisions or their buying behavior. This is a positive thing to promote a product or service at no cost (Evi, T. & Septo, I. 2021). In the world of marketing, positive word of mouth communication (WoM) is one of the effective ways to introduce a product or service to users. Positive WoM usually doesn't cost a lot of money to introduce the product or service. (Berger, J. 2014) stated that WoM in the marketing world cannot be separated from the culture of the world community who likes to share or tell stories. There are stories that have been passed down from generation to generation, including legends that were not originally written, but were told by word of mouth. For example, stories about the adventures of the Kancil from Indonesia, the Legend of King Arthur, and the Legend of Lady Godiva explain that the effectiveness of telling something is very, very important for human civilization (Evi, T. & Septo, I. 2021).

According to (Shi, W. et al., 2016) WoM is defined as informal communication among the public that contains a person's experience or satisfaction/dissatisfaction with the product or service he uses. This information is conveyed verbally (word of mouth) from one person to another so that it becomes an infinitely long chain.

WoM is an important part of marketing studies considering that communication in WoM can influence consumer purchasing decisions. Consumers trust WoM more in assessing a product, and are able to influence someone's recommendation for a purchase. Stories and experiences of someone using a product sound more interesting and this can affect their hearing to try the product. In addition, (Auf, M. A. A. et al., 2016) stated that the power of WoM also increases considering that humans are social creatures who like to interact and share with others, including the problem of purchasing preferences. WoM is also able to spread very quickly if the individual who spreads it also has a large network. Positive WoM is an effective, inexpensive and credible marketing communication tool.

Image is availabe at PDF file

Figure 2: Research Conceptual Framework

Research Hypothesis:

H1: Service quality has a significant positive effect on customer satisfaction.

H2: Service Quality has a significant positive effect on WoM

H3: Brand Image has a significant positive effect on Customer Satisfaction

H4: Brand Image has a significant positive effect on Positive WoM.

H5: Customer Satisfaction has a significant positive effect on Positive WoM

H6: Customer Satisfaction is able to mediate the effect of Service Quality on Positive WoM

H7: Customer Satisfaction is able to mediate the influence of Brand Image on Positive WoM

RESEARCH METHODOLOGY:

Research Design:

In accordance with (Bist, B. B. 2014) the approach used in this study is a quantitative approach or quantitative – hypothesis testing research. The form of research used is causality with a quantitative approach. The dependent variable is the dependent variable which is influenced by other independent variables. The dependent variable in this study is Positive WoM.

Population and Sample:

Number of students of SI Management and SI Accounting (N) = 1,067 (Class of 2014) + 601 (Class of 2015) + 423 (Class of 2016) + 299 (Class of 2017) + 420 (Class of 2018) = 2810 studentsn = N/ 1 + N(e)2

n = 2810 / 1 + 2810 (0,1)2

n = 2810 / 1 + 2810 (0,01)

n = 2810 / 29.1

n = 96.89 rounded up to 97 votes

So, the minimum number of samples studied was 97 respondents from students of the Undergraduate Management study program and the Undergraduate Accounting study program, Perbanas Institute Jakarta.

Respondents were taken randomly, which included students of the Undergraduate Management study program and the Undergraduate Accounting study program, Perbanas Institute Jakarta. Of the 300 respondents who received the questionnaire, 115 respondents returned the questionnaire. This is less than the minimum limit of the SEM model. (Rachbini, W. & Evi, T. 2021) states if the respondent's data is less than the minimum limit, then the alternative model is to use the PLS-SEM (Partial Least Square-Structural Equation Model) model which can use a minimum respondent data of 30-50 samples.

RESEARCH RESULTS AND DISCUSSION:

Characteristics of Respondents:

The data processed for this study were only for 50 respondents. Therefore, the characteristics of the 50 respondents from the data collection that will be used are as follows:

Image is availabe at PDF file

Figure 3: Number of Respondents by Age

Source: Data processed

Based on Figure 3, 2% or 1 respondent is 18 years old; 26% or 13 respondents aged 19 years; 26% or 13 respondents are 20 years old; 20% or 10 respondents are 21 years old; 8% or 4 respondents are 22 years old; and 18% or 9 respondents are 23 years old.

Image is availabe at PDF file

Figure 4: Respondents by Class (Academic Year)

Source: Data processed

Based on Figure 4, 10% or 5 respondents from the 2019/2020 class; 24% or 12 respondents entered the 2018/2019 class; 20% or 10 respondents from Class 2017/2018; 32% or 16 respondents entered the 2016/2017 class; 2% or 1 respondent from Class 2015/2016; and 12% or 6 respondents from the 2014/2015 Class.

Image is availabe at PDF file

Figure 5: Respondents based on Academic Achievement Index (GPA)

Source: Data processed

Based on Figure 5, none of the respondents had a GPA below or equal to 2.50; 8% or 4 respondents have GPA ranging from 2.51 to 2.75; 16% or 8 respondents have GPA ranging from 2.76 to 3.00; 30% or 15 respondents have GPA ranging from 3.01 to 3.25; 24% or 12 respondents have GPA ranging from 3.26 to 3.50; 12% or 6 respondents have a GPA of 3.51 to 3.75; and 10% or 5 respondents have a GPA more than or equal to 3.76.

Outer Model Evaluation:

The value of R-Square (R²) is used to see each dependent variable as the predictive power of the structural model. Changes in the R-Square value can be used to explain the effect of the independent variable on the dependent variable, namely whether the independent variable has a substantive effect on the dependent variable? The R-Square value can be classified into three, namely the R-Square value is said to be strong if = 0.75, the R-Square value is moderate if = 0.50, and the R-Square value is weak = 0.25. The results of the R-Square from this study can be seen in Table 10

Table 1 - R-Square

Source: Data processed

From Table 6 it can be said that the R-Square and R-Square Adjusted values ​​of Customer Satisfaction 0.508 and 0.487 and Positive WoM 0.657 and 0.635 are in the range of 0.50. This means that both variables are included in the moderate category. From the output above, it can be said that the effect of Service Quality (KL=X1) and Brand Image (CM=X2) on Customer Satisfaction (KP=Y) is 0.508 or 50.8% and the remaining 49.2% is influenced by other variables outside this research model. Furthermore, the same thing happened to the effect of Service Quality (KL=X1) and Brand Image (CM=X2) on Positive WoM (WOM=Z) with a value of 0.657 or 65.7% and the remaining 34.3% was influenced by other variables in outside of this research model.

Then the Fit Test of the entire model or the combined model (Goodness of Fit Model) was measured using the R-Square dependent latent variable with the same interpretation as the regression. Q-Square predictive relevance for construct models. Based on Table 4.10 the value of Q² can be calculated as follows:

Q² = 1 – (1 – 0,508) (1 – 0,657)

Q² = 1 – (0,49 x 0,34)

Q² = 1 – 0,16

Q² = 0,84

Based on the results of the model test above, it can be seen that the model can explain the phenomenon studied by 84%, the rest of the other variables outside the study explain this model. The value of Q² = 84% explains that the variables used in this study are strong because they are able to influence 84%.

Hypothesis Test Results:

After evaluating the Outer Model and Inner Model, the data studied were processed using PLS analysis. From the results of Path Coefficients, it can be seen that all hypotheses can be accepted. This is shown in Table 7

Table 2: Path Coefficients

Source: Data processed

Interpretation of Research Results:

The discussion or hypothesis testing is based on the values ​​contained in the structural analysis of the model, the path coefficient significance level obtained from the t-value and the standardized path coefficient value. The threshold value or the threshold for testing the hypothesis, namely the results of the coefficient analysis of the structural model can be seen in table 7. If P-Value is < 0.05 then the relationship is significant, otherwise if P-Value is > 0.05 then the relationship is not significant. From Table 7, it can be seen that the P-Value for H1 and H5 0.05, while H2, H3, H4, H6, and H7 have P-Value values ​​> 0.05.

Table 7 shows that the service quality variable has a significant effect on customer satisfaction with a P-value of 0.000 0.10. This means that for every increase in Service Quality, Customer/Student Satisfaction will increase. Thus the results of the first hypothesis (H1) are accepted.

The service quality variable has no significant effect on Positive WoM with a P-Value value of 0.711 > 0.10. This means that every improvement in Service Quality at Perbanas Institute Jakarta is not followed by Positive WoM students. Thus the results of the second hypothesis (H2) are rejected.

Furthermore, the Brand Image variable has a positive but not significant effect on Customer Satisfaction with a P-Value of 0.645 > 0.10. This means that every increase in Brand Image, Customer Satisfaction will increase but this is not significant or not give a big impact on Customer Satisfaction of Perbanas Institute students. Thus the results of the third hypothesis (H3) are rejected.

Image is availabe at PDF file

Figure 6: The Result of The Analysis

Based on Figure 6 Positive WoM is a latent variable because it is multidimensional and cannot be measured directly. There are three independent latent variables that positively affect WoM, namely Service Quality (KL/X1), Brand Image (CM/X2), and Customer Satisfaction (KP/Y). The path analysis model in this study is the mediating path model. Positive WoM is influenced directly through Service Quality and Brand Image, and through the mediating variable of Customer Satisfaction. The four latent variables use a reflective measurement approach which is indicated by a one-way arrow from the latent variable to the indicator variable. Each latent variable is measured by several indicator variables with different amounts. The latent variable of Service Quality is measured by 15 indicators, the latent variable of Brand Image is measured by 5 indicators, the latent variable of Customer Satisfaction is measured by 4 indicators, and the positive WoM latent variable is measured by 3 indicators. This is shown in Figure 7.

Image is availabe at PDF file

Figure 7: Path Diagram Model of Relationship between Latent Variables in SmartPLS in SmartPLS

Source: Data processed

The Effect of Service Quality on Customer Satisfaction:

Based on table 7 the results of the analysis of the coefficients of the structural path model are used to determine which relationships have a significant effect. The effect of Service Quality variable (KL = X1) on Customer Satisfaction (KL = Y) with a P-Value <0.05, which is 0.000, means Service Quality has a significant influence on Student Satisfaction. Therefore, the path coefficient value of 0.635 means that the Service Quality variable and the Customer Satisfaction variable have a positive relationship, so it can be concluded that there is a significant positive effect between the Service Quality variable on Customer Satisfaction. When service quality increases, customer satisfaction in this case students will also increase. From the results of this hypothesis, it means that Ho is rejected and Ha is accepted.

Effect of Service Quality on WoM:

Based on table 7 the results of the analysis of the coefficients of the structural path model are used to determine which relationships have a significant effect. The influence of Service Quality on WoM with a P-Value value <0.05, which is equal to 0.711, means that Service Quality has a significant influence on WoM. The coefficient value obtained is 0.062, which means that the quality of service on WoM has a positive relationship. Thus, there is a significant positive effect between Service Quality variables on WoM when Service Quality increases, WoM will increase as well. From the results of this hypothesis, Ho is rejected and Ha is accepted.

The Effect of Brand Image on Customer Satisfaction:

Based on table 7 the results of the analysis of the coefficients of the structural path model are used to determine which relationships have a significant effect. The effect of the Brand Image variable (CM = X2) on Customer Satisfaction (KL = Y) with a P-Value <0.05, which is 0.645, means that the Brand Image variable has a significant influence on the Customer Satisfaction variable. Therefore, the path coefficient value of 0.043 means that the Brand Image variable and the Customer Satisfaction variable have a positive relationship so that it can be concluded that there is a significant positive effect between the Brand Image variable on Customer Satisfaction. When Brand Image increases, Customer Satisfaction, in this case students, will also increase. From the results of this hypothesis, it means that Ho is rejected and Ha is accepted.

The Effect of Brand Image on Positive WoM:

Based on table 7 the results of the analysis of the coefficients of the structural path model are used to determine which relationships have a significant effect. The effect of the Brand Image variable (CM = X2) on Positive WoM (Z) with a P-Value value <0.05, which is 0.722, means that Brand Image has a significant effect on Positive WoM students. Therefore, the path coefficient value of 0.043 means that the Service Brand Image variable and the Positive WoM variable have a positive relationship, so it can be concluded that there is a significant positive effect between the Brand Image variable and the Positive WoM variable. When the Brand Image increases, the Positive WoM of students will also increase. From the results of this hypothesis, it means that Ho is rejected and Ha is accepted.

The Effect of Customer Satisfaction on Positive WoM:

Based on table 7 the results of the analysis of the coefficients of the structural path model are used to determine which relationships have a significant effect. The effect of the variable Customer Satisfaction (KP = Y) on Positive WoM (Z) with a P-Value <0.05, which is 0.000, means Customer Satisfaction has a significant effect on Positive WoM students. Therefore, the path coefficient value of 0.739 means that the Customer Satisfaction variable with the Positive WoM variable has a positive relationship so that it can be concluded that there is a significant positive effect between the Customer Satisfaction variable on the Positive WoM. When Customer/Student Satisfaction increases, the Positive WoM of students will also increase. From the results of this hypothesis, it means that Ho is rejected and Ha is accepted.

The Effect of Service Quality on Positive WoM with Customer Satisfaction Variables as Mediation Variables (Intervening):

Based on table 7 the results of the analysis of the coefficients of the structural path model are used to determine which relationships have a significant effect. The indirect effect of the Service Quality variable (KL = X1) on Positive WoM (WoM = Z) with a P-Value value <0.05, which is 0.709, means that the Service Quality variable indirectly has a significant influence on the Positive WoM variable.

Based on the research results, it is known that the indirect effect of Service Quality (KL = X1) on Positive WoM (WoM = Z) through Customer Satisfaction (KP = Y) is 0.87. This can be seen in Table 8.

Table 8: VAF Calculation for the Effect of Service Quality on Positive WoM with Customer Satisfaction Variables as Mediation Variables (Intervening)

Source: Data processed

According to (Hair et al., 2013; Sholihin & Ratmono 2013) if the VAF value is above 80%, in this case 87% then this indicates that the role of Customer Satisfaction (KP = Y) is a full mediation. This means that the Customer Satisfaction variable is able to mediate the effect of Service Quality on Positive WoM because Service Quality has a significant effect on Positive WoM through Customer Satisfaction.

The Effect of Brand Image on Positive WoM with Customer Satisfaction Variable as a Mediation Variable (Intervening):

Based on table 8, the results of the coefficient analysis of the structural path model are used to determine which relationships have a significant effect. For the indirect effect of the Brand Image variable (CM = X2) on the Positive WoM (WoM = Z) with a P-Value <0.05, which is 0.698, which means that the Brand Image variable indirectly has a significant influence on the Positive WoM variable.

Based on the research results, it is known that the indirect effect of Brand Image (CM = X2) on Positive WoM (WoM = Z) through Customer Satisfaction (KP = Y) is 0.62. This can be seen in Table 9.