The Effect Of Online Learning Service Quality On Student Satisfaction During COVID19 Pandemic In 2020

The coronavirus pandemic in 2019 (COVID 19) becomes the focus of all the countries in the world to be solved. The virus is so dangerous and its imposibble to stop its spread. To face this condition, Indonesian government advises all Indonesians to work from home in all sectors, including education. Universitas Kristen Krida Wacana obeys this suggestion by implementing the work from home for the lecturers to carry on online learning for the students. This study intends to investigate and analyze the online learning quality on the satisfaction of the management students in the Faculty of Economics and Business who was taking the subject of Human Resource Management and Organizational Behavior in 2020. Furthermore, they become the population of this study. This research grabs 64 students as the sample by using stratified random sampling. The data were analyzes bya a variance based Stuctural Equation Model (SEM) with the collected data associated with latent variables. After testing and analyzing the data, this study infers that the better quality of service online learning, the more satisfied the students. In other words, the implementation of this online learning is successful. Likewise, some practical and theoretical recommendations based on this evidence are available.

241 finally leads to death on infected persons (Utari, 2020). To stop the spread of this virus, the President of the Indonesia Republic, Joko Widodo, demands the Indonesians for supporting social distancing by doing everything from home, including worshiping, working, and studying (Cahya, 2020). As a consequence, the provincial government in Jakarta through Gubernatorial Regulation No.33 of 2020 on April 9, about the largescale social restriction does not permit all businesses to operate except for healthcare, food and beverage, energy, communications and information technology, financial services, and logistics, hotels, construction services, strategic industry, vital object services, daily need services.
Regardless of the release date of Gubernatorial Regulation No. 33, the accumulated number of infected people increased from 741 at the end of March to 4,138 at the end of April 2020. Similarly, the accumulated total of the deceased people grew from 84 at the end of March to 381 at the end of April (see Table 1).  (2020) In Indonesia, including Jakarta, education becomes one of the sectors affected by the pandemic of COVID 19. Therefore, the central government issues the policy for not executing the face-to-face-national exam for the students in elementary, junior, and senior high schools, as usual. Instead, an online exam result or the accumulated point in the report card becomes two alternative solutions to determine their graduation (Putri, 2020).
For the higher education world, the Minister of Education and Culture, Nadiem Anwar Makarim, recommends that all institutions of higher education temporarily stop the face-to-face meeting and requests them to utilize online learning media (Winahyu, 2020). As one of the higher education institutions in Jakarta following this suggestion, Universitas Kristen Krida Wacana switches the learning method for its lecturers and students, as suggested by the government, as declared in the Rector Decree No. 345/SK/UKKW/R/III/2020. Furthermore, this instruction is followed up by the dean of economics and business faculty by issuing decree No. 13/SK/UKKW/D/III/2020 to organize the implementation of this method.
Linked to online learning in this faculty, the students, as the consumers of the university, deserve to express their satisfaction after they use online learning. Therefore, this situation motivates this study by employing the management students taking the class of human resource management and behavioral organization in the even semester in the academic year 2019/2020 in April 2020. These students get chosen as the population because they are easily reachable, and the result is useful as the preliminary study before Universitas Kristen Krida Wacana applies the online survey about learning service quality to the whole students.

A. Customer satisfaction
Satisfaction reflects the response of customers when they compare the expectation with the experience after utilizing the product or service offered by a company (Tijptono & Chandra, 2016). According to Parasuraman, Zeithaml, & Berry (1985), satisfaction happens when expected service is the same as or lower than perceived service. Furthermore, the company has to maintain it to make the business succeed and survive in the tight market rivalry (Fen & Lian, 2005). Hence, the company needs to assume customers as the assets that can yield the profits (Anderson & Mital, 2000).

B. Service Quality, Online Service Quality, and Online Learning Service Quality
The service will brings quality to the users if it can satisfy their needs or expectations (Akbar & Parvez, 2009). Parasuraman et al. (1985) were the first scholar introducing the service quality with ten dimensions; they are tangibility, responsiveness, reliability, competence, access, courtesy, communication, credibility, security, and understanding. In their next manuscript, furthermore, Parasuraman et al. (1988) compress the number of 242 dimensions to be five: tangibility, empathy, responsiveness, reliability, and assurance. Moreover, the researchers use these five dimensions and their indicators when measuring the service quality based on the customer perception of the restaurant industry (Fen & Lian, 2005), the telecommunication industry (Akbar & Parvez, 2009), the salon and spa (Puung, Fudholi, & Dharmmesta, 2014).
By exploring data and conducting the focus group discussion, Santos (2003) successfully finds the dimensions of online service quality needed by firms having e-commerce business; they are usage, content, structure, linkage, search, and appearance. Unlike her research, Parasuraman, Zeithaml & Malhotra (2005) build four dimensions named efficiency, fulfillment, system availability, and privacy to measure online service quality and three dimensions called responsiveness, compensation, and recovery to measure online recovery service quality. Likewise, the instrument created by Parasuraman et al. (2005) gets used by other researchers, such as Suryani & Sumiati (2010), Wu (2011), Rafiq, Lu, & Fulford (2012), Komara (2013), Zemblyte (2015), Zehir & Narcıkara (2016), as well as Malhotra (2017).
In a higher education setting, one of the relevant quality aspects is about online learning. By this method, the education process can be everywhere, not depend on the meeting between lecturer and students in the classroom (Uppal, Ali, & Gulliver, 2018). For the research needs, the online learning service quality has to be able to measure. In their study, Pham et al. (2019) use three dimensions of online learning services. The first is the system. Secondly, instructors and course materials. Thirdly, the administrative and support service.
C. The relationship between online learning service quality with student satisfaction Lee (2010) finds that the service quality of online support positively influences student satisfaction in Seoul Digital University and Boston University. By investigating the students of the Palangka Raya University as his samples, Nahan (2013) proves that the quality of teaching and service positively determines student satisfaction. Also, after studying the active students from the economics faculty in the open university in Indonesia counting on long-distance learning, Maria, Hadiwidjaja, & Mulyana (2015) conclude that student satisfaction gets positively affected by service quality. In his research on the graduate students taking a Master in Management program at Mercu Buana University in Jakarta, Pawirosumarto (2016) concludes that the quality of the system, information, and service can raise the satisfaction of e-learning users. Additionally, by employing the students of Telkom University in Bandung as the samples of their study, Faoziah & Sembiring (2017) concludes that the speed of media response and security-and-privacy can reflect the quality of online learning and have a significant effect on student satisfaction. Furthermore, Osman, Saputra & Saha (2017) investigate the effect of quality of service and quality of program offered of satisfaction of the students form some universities in Bangladesh; they summarize that each determinant of student satisfaction has a positive sign. Likewise, the study of Pham, Limbu, Bui, Nguyen, & Pham (2019) in Vietnam infers that service quality in online learning has a positive effect on student satisfaction. Consistent with these mentioned previous research evidence, the theoretical framework gets built based on the student satisfaction model related to the online learning service quality of Pham et al. (2019), as seen in Figure 1, with the first hypothesis announced in this manner. H1: The better quality of service online learning, the more satisfied the students.

A. Variable Definition
The variable in this study is the Online Learning Service Quality (OLSQ) and Student Satisfaction (SS). a. OLSQ acts as the independent variable. The measurement of this variable refers to Pham et al. (2019).
According to them, it has three quality-based dimensions, i.e., system, instructors and course materials; administrative and support service. Because no transaction feature exists in the online learning system in Universitas Kristen Krida Wacana, three items related get removed. The relevant ones, furthermore, can be seen in Table 2. The instructions on the online learning website are easy to keep an eye on. SYS2 The information on the online learning website gets always renewed. SYS3 The online learning website provides a meaningful description of courses. SYS4 The information on the online learning website can be easy to obtain for the student. SYS5 The online learning website is stylish. SYS6 The campus fulfills what it promises based on the information on the online learning website. SYS7 Accessing information on the online learning website is quick. SYS8 The information provided by the online learning website is precise. Instructor and course materials

ICM1
The instructor is literally. ICM2 The course materials can be easily applied.

ICM3
The course materials motivate to think. ICM4 The duration and difficulties of exams provided are rational. ICM5 The instructor gives a speedy response to what students need. ICM6 The instructor steadily gives a lecture. ICM7 The instructor prepares and organizes the materials well. ICM8 The instructor of online learning creatively creates the situation to make the student get collaboratively involved. ICM9 The instructor gives attention to the student. ICM10 Course materials give benefits to the student. ICM11 Course materials follow the development of age. ICM12 The duration and difficulties of tasks provided are rational. Administrative and support

AS1
The campus has a supporting service unit to facilitate online learning for the student. AS2 The staff member notifies the student when the service needed will execute. AS3 The staff member is never late to serve the student. AS4 The staff member gives the students excellent service. AS5 The staff member knows what the students need. AS6 The staff member personally gives attention to the students. AS7 The helpdesk staff member provides comfortable operating hours.
b. Student Satisfaction (SS) has a role as the dependent variable. By denoting and modifying the content of the indicators used by Widowati & Tsabita (2017), three items exist to measure SS: (1) As a user of online learning, I am satisfied.
(2) As a user of online learning, I am glad.
(3) Online learning implementation in our campus is better than that in the others.

B. Population and sample
Performing as the population in this study is 76 management students in the Faculty of Economics and Business of Universitas Kristen Krida Wacana in Jakarta, taking the subject of human resource management and organizational behavior in 2020. Likewise, total samples (n) representing the population (N) are calculated by the Slovin formula by a 1% border of error (be) in the first equation, obtained from Umar (2008): By denoting this equation, the total samples (n) are 76 1+76(1%)(1%) = 63.87 ≈ 64 students. Because the students becoming the population take two subjects, we treat the subject as strata and utilize the stratified random sampling method to take the sample (see Table 3 for the distribution of the total samples from the population). Furthermore, 27 and 37 persons get randomly taken by the random number made by Microsoft Excel, as informed by Hartono (2012).

C. Method to collect the data
This study utilizes a survey method to collect data. According to Hartono (2012), this method uses the questionnaire to get the response of the participants based on the list of question items. Therefore, we use the five points of the Likert scale, from one to five, representing the states of totally disagree, disagree, neutral, agree, totally agree, respectively, as suggested by Sugiyono (2012). As a consequence, this response to the answer of participants needs validity and reliability test. To perform these tests, we use confirmatory factor analysis and Cronbach alpha analysis.

D. Method to analysis the data
In this study, a structural equation model-based on variance becomes the method for analyzing the data. It is due to the latent (Ghozali, 2014) and the number of respondents between 30 and 100 (Ghozali, 2008). Likewise, the model can get written in the following second equation.
The second and the third step is setting the significance level (α) of 5% and getting the probability value (pvalue) of t-statistic by Warp PLS software, respectively. As the fourth one, comparing α with p-value is essential to get a statistical conclusion related to the rejection of the null hypothesis by following this rule.
i. If the p-value surpasses or the same as 5%, the null hypothesis gets recognized. ii.
If the p-value goes under 5%, the null hypothesis gets forbidden.

A. The descriptive statistics for demography variables
In this study, all the students becoming the samples respond quickly to the distribution of the link of online questionnaires on 23-25 April 2020. Below are the statistics related to the demographic variable, i.e., gender 245 (see Table 4), and the academic information of students, i.e., batch (see Table 5) and range of grade point average (GPA) (see Table 6). Table 4 presents the number of students by denoting gender: 31 males and 33 females. Overall, the females dominantly participate in this survey (51.56%).  Table 5 indicates the number of students by denoting batch: 1 from 2014, 2 from 2017, 32 from 2018, and 29 from 2019. Overall, the students coming from batch 2018 dominantly participate in this survey (50%).  Table 6 exhibits the number of students by denoting three ranges of GPA: 2.51-3.00, 3.01 -3.50, and 3.51-4.00. The total students in the first range are 16, in the second range are 21, in the third range is 27. Overall, the students with a GPA between 3.51 and 4 dominantly participate in this survey (42.19%).

B. The test result of validity and reliability
The validity and reliability test is useful to know the accuracy and consistency of the answer of respondents. In the validity context, the answer will be accurate on each question item if the loading factor is above 0.5 (Sholihin & Ratmono, 2013). In reliability context, the answer to the group of items will be consistent, if the Cronbach Alpha is beyond 0.7 (Ghozali, 2016). Table 7 shows the validity of test result by confirmatory factor analysis and the reliability test result by Cronbach Alpha analysis on items in three dimensions of online learning service quality. a. In the validity test result, because all the loading factor values of each item in three dimensions, the system, the instructor-and-material, the administrative-and-support, are above 0.6 (see Panel A), the answer of the respondents is valid. b. In the reliability test result context, because all the Cronbach Alpha values of the valid items in each dimension are beyond 0.7 (see Panel B), the answer of the respondents to these items is reliable. Table 8 notifies the validity test result by confirmatory factor analysis and the reliability test result by Cronbach Alpha analysis on items in student satisfaction. a. In the validity test result, because all the loading factor values of each item for student satisfaction are above 0.7 (see Panel A), the answer of the respondents is valid. b. In the reliability test result context, because all the Cronbach Alpha values of the valid items for student satisfaction are beyond 0.7 (see Panel B), the answer of the respondents to these items is reliable.

C. The estimation result of the research model
From the previous section, the validity and reliability test already get fulfilled. After that, the next stage is resulting in the variance-based structural equation model estimation by Warp PLS 3, shown in Table 9. From 247 this table, the probability value of the t-statistic of OLSQ got is <0.001. Because this value is lower than a 5% significant level, the null hypothesis needs to refuse. Instead, the alternative hypothesis has to receive.

D. Discussion
This study successfully proves that online learning service quality positively contributes to the satisfaction of the students. Therefore, this situation also supports the finding of Pham et al. (2019) and the investigation employing students from the open university that counting on the long-distance online learning, as conducted by Maria et al. (2015). Also, the result of this study is consistent with the research taking the parts of online learning, as executed by Lee (2010), Pawirosumarto (2016), Faoziah & Sembiring (2017).
Practically, this research is useful to the Universitas Kristen Krida Wacana starts focusing on online learning. What the university gives attention to is on the low response of strongly agree and agree. This response exists if less than 60% of respondents with two statements are available. For the dimension of the system, the follow-up action given is on SYS1, SYS5, and SYS6 (see Table 10). Based on this evidence, the responsible unit has to able: a. to make the instruction on the online website easy to track for the students b. to make the website attract the student attention c. to make the promise on the online learning website immediately achieved For the dimension of the instructor and material, the focus of follow-up action has to be on ICM2, ICM3, ICM4, ICM5, and ICM12 (see Table 11). a. Linked to ICM2 and ICM3, the lecturer has to able to make the students understand and be interested in studying the course materials. b. Linked to ICM4 and ICM12, the lecturer should give a longer duration for the students to finish their exams and assignments. By executing it, the students have sufficient time to think to solve the hard problems. c. Linked to ICM5, the lecturer should quickly respond to fulfilling the student needs.
248 For the dimension of the administrative and support, the focus of follow-up action has to be on AS1, AS2, AS3, AS5, AS6, and AS7 (see Table 12). a. Related to AS1, the campus has to socialize the unit facilitating online learning to the students. b. Related to AS2 and AS3, the staff members have to make the schedule related to the service. c. Related to AS5 and AS6, the campus has to give the incentive to the staff members so that they prioritize the students. d. Related to AS7, the staff members must have the flexibility of time to serve the students. Universitas Kristen Krida Wacana views that the students are valuable costumers because they already trust in the offered learning process quality. Therefore, the university has to be able to provide lecturers with teaching performance. Previously, the teaching gets done by a face-to-face meeting. Once the COVID 19 happens, this activity gets executed online.
This research wants to investigate and analyze the service quality of online learning on the satisfaction of the management students taking the class of resource management and organizational behavior. Once the data get collected, analyzed, and discussed, this study finally concludes that this online learning service quality positively affects student satisfaction. It means the university has to prepare and equip (1) the online system with the attractive displays, the precise information, the sufficient bandwidth, (2) the lecturers as an instructor and a material provider with training, and (3) the administrators with the readiness to help the students, to satisfy the active students as the user of online learning.
Theoretically, this study has some limitations, such as the use of the limited number of students as a population and determinant of student satisfaction. Two situations can get improved by the next scholars by doing some following actions.
a. The first inadequacy can get overcome by employing all the students from one university or some universities. b. The second limitation can get fixed by adding factors like experiential marketing and student loyalty as the exogenous and endogenous variables, respectively. Therefore, the mediating effect of student satisfaction on the impact of experiential marketing and online learning service quality on student loyalty can get proven by a Sobel test.