1. Introduction

The millennial generation in Indonesia is becoming a majority. In the latest survey conducted by the Indonesia Bureau of Statistics (2018), 66.02% of Indonesia’s population that entered the workforce, 50.35% are millennials. This phenomenon has also taken place all over the world, the millennials are expected to outnumber their predecessors (Kurz et al., 2019). In Indonesia however, not only millennials outnumbered their previous generations, but the fact is that they were actually supported by the Indonesian Government 2045 golden generation plan, which mainly consisted of maintaining the millennial generation (Rokhman et al., 2014). This fact provided clear evidence on how important millennials are for Indonesia, and this is the reason why millennial are there in every sector of Indonesia’s economy.

Studies in Indonesia started to look at the millennials in every sector. Millennial as a consumer have attracted attention simply because how unique their behavior is (Handriana et al., 2020; Purwanto et al., 2020). Millennial are expected to fill the voluntary role in Indonesian society (Adha et al., 2019) and even millennials family behavior on Covid-19 pandemic has become an issue (Kasdi, 2020) with noteworthy attention on their resilience. Millennial on education actually, proposed a condition while according to Indonesia Bureau of Statistics (2018), 5.10% and 5.77% millennials both in countryside and city are working in this field, which is more than finance and insurance sector, health and social activities sector, and even governmental and public service sector. The millennial generation has played an important role in Indonesian education sector.

Indonesia’s education sector is not without its problems. Firstly, it is worth mentioning that several problems are still commonly found in this sector, mainly teachers shortage, uneven distribution of education equipment and personnel, and education quality in general (Nasution, 2006). Secondly, Indonesia has a variety of school systems (Suharno et al., 2020), each with their uniqueness, and one of them are Integrated Islamic School (Sekol ah Islam Terpadu-SIT in Bahasa). Integrated Islamic School is a school that required integration between affective, cognitive, and psychomotor in their education system, and as the name implies, Islam plays an important role in such schools (Hadi et al., 2014). According to the latest employee data, around 60% of Integrated Islamic School employees are coming from the millennial generation, and this employee counted on 8.378 in Central Java only (this data gathered from internal sources, no published article exists).

With education as a long-lasting work, especially with the special requirement regarding Islamic education, organizational commitment and loyalty are becoming an important part of millennial’s employees. A millennial employee on one hand can develop a sense of commitment and loyalty toward the organization, which in return will help the organization growth and survival (Scales & Quincy Brown, 2020). On the other hand, dealing with the millennial employee is necessary to put some perspective on their world view, attitude, and provide the right approach to nurture their organizational commitment and loyalty (Holtschlag et al., 2020). Organizational commitment and loyalty also played an important role in mitigating employee turnover, to keep employee slightly longer in the organization (Noor et al., 2020; Yao et al., 2019)Bandura’s Career Theory. This means that organizational commitment and loyalty, both are very important aspects in maintain millennial generation employees, especially in Integrated Islamic Schools.

This study aimed to gain an understanding of employees’ organizational commitment and loyalty among millennials employees in Integrated Islamic School. Several hypotheses are tested using several analytical methods, and the data gathered from an online questionnaire distributed among employees on three Islamic Education Institutions that use Integrated Islamic School format in Central Java, Indonesia. This study found several relationships amongst variables, as a result of confirmatory factor analysis. SPSS and Structural Equation Modelling (SEM) AMOS are used to help the analytical process.

2. Literature Review

2.1. Millennial Generation

A millennial generation is defined in many ways. The most common definition is based on the year of birth. Experts argued that millennial are those individuals who were born between years 1981–1992 according to Pew Research Center and United States Census Bureau (Kurz et al., 2019; Stewart et al., 2017). Others argued that millennials are those who were born between years 1980 to 1997 (Fry, 2018). Overall, experts agree that a millennial is an individual that was born in the ’80s and the ’90s, and Indonesia according to Indonesia Central Bureau of Statistics defined a millennial as an individual born between the years 1980–2000 (Indonesia Central Bureau of Statistics, 2018). This means in Indonesia, millennials are categorized as a person who were born between years 1980–2000.

Secondly, millennials usually relate to certain traits that are viewed as unique and different, compared to another generational cohort. Several experts argued that millennials to become ‘a millennial’ had something to do with technology, arguably, millennials are the first digitally suave generation (Pyöriä et al., 2017) and have a very strong connection with advance technology even personally in the workplace (Kim, 2018). Millennials also had a unique way of defining their point of view, mainly focused on things that aren’t usually necessary, like fashion and music taste (Stewart et al., 2017). Even in the job, millennials tend to search for something that satisfies their need, not only for money but also an opportunity for career development, and personal growth (Daud & Wan Hanafi, 2020; Holtschlag et al., 2020). These traits are correlated with how organizations viewed millennials, and this also happened in Indonesia.

Organizations in Indonesia viewed the millennial generation comparably with how international communities viewed millennials. It is common to relate millennials with high turnover and turnover intention (Mappamiring et al., 2020). Millennial lifestyle is associated with lifestyle, modernism, this affected their view on the traditional product (Fibri & Frøst, 2020) and sometimes only appears as a habit, not a well-implemented world view (Amalia et al., 2020). With this widespread view, almost certainly every aspect of the Indonesian economy has a certain tendency on the millennial generation. This is the reason why consumer behavior study in Indonesia is rich with millennial customer behavior, which can be found in the number of studies like (Amalia et al., 2020; Handriana et al., 2020; Purwanto et al., 2020), or Indonesian millennial financial studies like (Soekarno & Pranoto, 2020; Kunaifi & Akbar, 2019).

The Millennial generation are in itself a subject of an interesting study. An expert suggested that an organizational behavior study could be conducted based on the generational cohort, to provide in-depth insight into each generation (Liu et al., 2019). With the recent condition, in which organizations tend to have a much diverse generation, it is paramount to start to look at a things from a different generational perspective (Stewart et al., 2017). Different ways are needed in dealing with millennial organizational commitment and loyalty, and with the right treatment, a millennial could nurture their commitment and loyalty towards the organization (Scales & Quincy Brown, 2020). Hence it is important, to study the Indonesian millennial generation for their organizational commitment and loyalty.

2.2. Organizational Commitment and Loyalty

Organizational commitment and loyalty are important to any organization. Organizational commitment and loyalty are important keys in developing employee organization citizenship behavior (Robbins & Judge, 2013). Organizational commitment and loyalty are 2 (two) separate concepts but correlated according to research.

Organizational commitment is a concept that is related to the employee. Organizational commitment is related to several variables, that enabled the employee to do something, stay, and committed to the company. Studies found that employee trust and satisfaction had a positive correlation with organizational commitment (Rameshkumar, 2019; Yao et al., 2019; Yang & Chang, 2008). With these findings, employee trust and employee commitment are necessary in order to maintain employee organizational commitment.

Organizational commitment in itself is an interesting concept. Experts define organizational commitment in different ways, but they mostly talk about employee-related issues and their involvement in an organization (Mowday, Porter, & Steers, 1982). This definition derived from the dichotomy of attitudinal-behavioral, and organizational commitment is interchangeable with attitudinal commitment. Organizational commitment also grows, with experts conducting research on how to measure it as a variable, like Porter’s Organizational Commitment Questionnaire (Commeiras & Fournier, 2013). But some experts are not satisfied, Meyer and Allen (1991) argued that organizational commitment doesn’t stop on affective aspects, there is also an obligation and perceived aspect, hence the concept of organizational commitment is born, which contains affective commitment, continuance commitment and normative commitment (Meyer & Allen, 1991).

Organizational commitment and loyalty have a connection. Organizational commitment gained enough attention from scholars, mainly because organizational commitment affects employee behavior especially turnover (Mowday et al., 1982). Turnover is a concept that is strongly related to loyalty, even though loyalty is also found in marketing studies (Yao et al., 2019). With attitudinal-behavioral dichotomy, experts now argue that loyalty also consists of attitudinal and behavioral aspects, and both are strongly related (Bandyopadhyay & Martell, 2007). With organizational commitment as a catalyst on employee willingness to stay in the organization, the millennial generation is becoming a prime object for this type of study. The millennial generation are correlated to an attitude like job-hopping (Holtschlag et al., 2020), and for the organization, the millennial world view is not assimilated with the organization, which provides a challenge in integrating the millennials with the organization (Stewart et al., 2017). Nurturing millennial's organizational commitment could become a key in reassuring their loyalty (Scales & Brown, 2020; Graybill, 2014).

2.3. Integrated Islamic School

Integrated Islamic schools are one of the several school formats that are currently active and well known in Indonesia. Indonesia still viewed religion as an integral part of their life, as a result, it is commonly found that even the national education curriculum included religion-based education (Masuda & Yudhistira, 2020). This is the reason why Indonesia’s education still has a traditional format like Pondok pesantren or Islamic boarding school (Nasution, 2006), even though this educational institution has to comply with government rules and regulations.

Integrated Islamic school is one of several types of Islamic schools. It has a strong connection with religion-based teaching, but with a more modern approach reflected in this type of school effort in integrating the affective, psychomotor, and normative aspects of education processes (Hadi et al., 2014), this type of school is on the rise. According to the data gathered from Central Java Integrated Islamic School Network (Jaringan Sekolah Islam Terpadu (JSIT) Jawa Tengah, there are 8.378 teachers and employees currently working in a number of integrated Islamic schools in Central Java only (this data is not published publicly). This made the teachers and employees of integrated Islamic schools face several unique circumstances, firstly, they have to adapt to the challenges on Indonesia education. Secondly, their work culture are heavily influenced by religious aspects.

It is worth noting that according to a study, Indonesia’s religious education actually resulted in many moderate individuals (Masuda & Yudhistira, 2020). But on organizational terms, an employee from the millennial generation still has to deal with a common stereotype on the millennial generation, including organizational commitment and loyalty (Holtschlag et al., 2020; Mappamiring et al., 2020). This condition creates an opportunity on conducting organizational commitment and loyalty research on these types of millennials that worked in the heavily religious work environment.

2.4. Hypotheses

This study aimed to test several hypotheses. These hypotheses are developed to understand the relations between variables. These hypotheses and theoretical frameworks were mainly developed by (Yao et al., 2019), this study conducted research based on previous study’s limitations and future directions. Therefore this study also tested the robustness of the developed model, tested the hypotheses on different cultural backgrounds and work environments. Furthermore, this hypothesis development could be used to understand the millennial organizational commitment and loyalty.

2.4.1. Employee Trust and Satisfaction

Several studies have explained the relationship between employee trust and employee satisfaction. The results found mainly stated that employee trust had a positive relationship with employee satisfaction. Employee trust is defined as individual assumptions or expectations, that other’s actions in the future will be beneficial, favorable, or at least not detrimental to one’s interest (Perry & Mankin, 2004). Trust is related strongly to satisfaction (Ferres et al., 2004). Trust is also related to satisfaction among nurses and health workers (Yang & Chang, 2008). Regarding the manager attitude, the way manager or supervisors interacted with employees is strongly related to employee satisfaction. Trust in managerial leadership related to employee satisfaction (Jena et al., 2018; Brown et al., 2015; Kelloway et al., 2012). A leader's attitude toward mobile phones is also related to employee satisfaction, it is because the leader will focus more on a mobile phone rather than an employee (Roberts & David, 2020).

With previous studies mentioned above, we argued that the first hypothesis is:

H1: Employee trust has a positive effect on employee satisfaction.

2.4.2. Employee Trust and Organizational Commitment

Employee trust had a positive impact on organizational commitment according to several studies. Trust endorses employees to put more attention on company safety procedures (Liu et al., 2020). Trust also increased employee positive relationship toward organization and manager, and this further increased work-related positive behavior (Gill, 2008). It is also found that employee trust and satisfaction played an important part in maintaining employee commitment (Yang & Chang, 2008).

This impact-related on every aspect of organizational commitment. Organizational commitment also included obligation and perceived aspect, and this is the reason why there is 3 (three) conception of commitment which are affective, continuance, and normative commitment (Meyer & Allen, 1991). Trust strongly provides a basis for employee organizational commitment and organizational citizenship behavior (Robbins & Judge, 2013). Trust positively impacts the organizational commitment confirmed by other studies (Yao et al., 2019), including on the millennial generation (Scales & Quincy Brown, 2020), therefore we argued that employee trust has a positive impact on affective commitment, normative commitment, and continuance commitment, and these hypotheses are:

H2a: Employee trust has a positive impact on affective commitment.

H2b: Employee trust has a positive impact on normative commitment.

H2c: Employee trust has a positive impact on continuance commitment.

2.4.3. Employee Satisfaction and Organizational Commitment

Employee satisfaction makes many impacts on employee organizational behavior including organizational commitment. Employee satisfaction stated a condition that employees felt that the organization fulfilled their necessary needs (Podsakoff et al., 1996). Several studies pointed out that satisfaction affected organizational commitment, including satisfaction with the manager’s leadership (Kelloway et al., 2012). Job satisfaction is also positively related to organizational commitment among nurses (Yang & Chang, 2008). Other studies also explained how satisfaction played an important role in maintaining millennial commitment toward the organization (Holtschlag et al., 2020; Kim & Yang, 2020), and satisfaction had a positive impact on organizational commitment (Rameshkumar, 2019; Yao et al., 2019; Sila & Širok, 2018) On the other hand, if the employee is not satisfied with the organization and their leader, it affects the organizational commitment (Roberts & David, 2020). With the studies mentioned above, we argued that employee satisfaction has a positive impact on affective, normative, and continuance commitment. Therefore these hypotheses are concluded.

H3a: Employee satisfaction has a positive impact on affective commitment.

H3b: Employee satisfaction has a positive impact on normative commitment.

H3c: Employee satisfaction has a positive impact on continuance commitment.

2.4.4. Organizational Commitment and Employee Loyalty

The organizational commitment had a positive impact on employee loyalty according to several studies. Commitment is a form of employee willingness to partake in the organization and reach organizational goals (Mowday et al., 1982) this is why commitment is one of the turnover factors. Commitment influences many aspects, it is not just limited to affective aspects but it is also linked to obligation and perceived aspects (Meyer & Allen, 1991). Hence organizational commitment was found to be an antecedent for attitudinal loyalty (Izogo, 2015). An employee that is committed to the organization tends to be more loyal, even though they are from the millennia generation with a “job-hopping” reputation (Holtschlag et al., 2020). Some studies have also found that it general, an employee’s performance can also be greatly enhanced by organizational commitment, in all 3 (three) conception, affective, normative, and continuance commitment (Yao et al., 2019; Zopiatis, Constanti, & Theocharous, 2014). Therefore we conclude that organizational commitment has a positive impact on employee attitude and behavioral loyalty.

H4a: Affective commitment has a positive impact on attitudinal loyalty.

H4b: Affective commitment has a positive impact on behavioral loyalty.

H5a: Normative commitment has a positive impact on attitudinal loyalty.

H5b: Normative commitment has a positive impact on behavioral loyalty.

H6a: Continuance commitment has a positive impact on attitudinal loyalty.

H6b: Continuance commitment has a positive impact on behavioral loyalty.

2.4.5. Employee Attitudinal and Behavioral Loyalty

Employee attitudinal and behavioral loyalty are derived from the dichotomy between attitudinal and behavioral. Attitudinal and behavioral dichotomy starts from the importance of looking at certain issues or variables from the attitudinal perspective as a thinking stage, and behavioral on acting stage (Mowday et al., 1982). Experts argue that attitudinal and behavioral states are related to loyalty because an individual cannot take action without proper thinking (Bandyopadhyay & Martell, 2007). Thinking and action stages are also related, and the thinking part is an example of how a certain individual relates himself with the organization (McMullan & Gilmore, 2003). A positive impact between attitudinal and behavioral loyalty is found in several studies (Dehghan & Shahin, 2011), with the way organization or company designed a certain way of action, strongly enhanced consumer behavioral loyalty (Antwi et al., 2020; Bilgihan et al., 2016), or employee behavioral loyalty (Yao et al., 2019; Indrawan et al., 2018). With these findings, it is clear that loyalty is a concept that is available in both human resources and marketing studies, but it does not change the fact that attitudinal loyalty has a positive impact on behavioral loyalty, and therefore we conclude that attitudinal loyalty has a positive impact on behavioral loyalty.

H7: Attitudinal loyalty has a positive impact on behavioral loyalty.

Figure 1: Theoretical Framework

3. Research Methods and Materials

This research was conducted on 3 (three) Islamic educational institutions in Central Java, Indonesia. The reason behind this selection is the majority of the employee from this institution come from the millennial generation, and fulfill the definition from Indonesia Central Bureau of Statistics. Combined with the experience that these institutions have gained in their last 10 years of operation, several issues and problems regarding millennial organizational commitment and loyalty should be encountered. Lastly, the institutions could provide the necessary information regarding this study and could provide the possible way for the researchers to gain the important data and information (Sekaran and Bougie, 2016).

The research utilizes an online questionnaire, delivered throughout the respondents on each institution. This questionnaire is combined and selected carefully on each variable. Employee trust variable questions are selected from Perry and Mankin (Perry & Mankin, 2004), while employee satisfaction is taken and selected carefully from Podsakoff and others (Podsakoff et al., 1996). The organizational commitment with three conceptions, affective, normative, and continuance commitment questions are taken from Meyer and Allen (Meyer & Allen, 1991), and lastly, the questions for employee loyalty is selected from McMullan and Gilmore (McMullan & Gilmore, 2003).

The results are 261 responses from the respondents. It fits the minimum number of responses in SEM AMOS (Sarstedt et al., 2017), which is used as an analysis tool. The partial disaggregation approach is used to minimize random error (Hau & Marsh, 2004), while confirmatory factor analysis is implied, mainly to test the robustness of the model developed by the previous study, in Yao et.al (2019) (Yao et al., 2019) This test was the recommended for future research by a previous study.

4. Results and Discussion

From 261 responses, this study acquired a 100% rate of millennial generation filling the questionnaire from 3 (three) different Islamic education institutions. All of the respondents that finished the questionnaire are all had a year of birth between the year 1980 to 2000 which in line with the millennial generation definition stated by the Indonesia Central Bureau of statistics. From that number, 26.7% are male respondents, with 73.3% female respondents. The majority of the respondents (73.3%) had a bachelor’s degree whereas the remaining 26.7% had an academic degree, ranging from magister, high school graduate, and junior high school graduate. The respondent had a different structural background, but 73% of the respondent worked as a teacher in various positions and subjects, while the rest is working on administrative, and supporting system function.

Data gathered from the respondents are then treated with the outlier checked, resulted in 235 responses confirmed feasible to conduct the analytical processes. This number confirmed with the minimum data required for SEM analysis, which is 200 data (Sarstedt et al., 2017). The non-outlier data is then tested for validity and reliability test, for each variable and indicator. This determines whether the data is suitable for further analysis, especially using SEM AMOS analytical tool.

The validity test was conducted with the principal component analysis extraction method, using the Varimax extraction method and Kaiser normalization. The results from this test are the validity of the variable indicator. The results on the employee satisfaction variable are the elimination of ES14 and ES15 indicators, while affective commitment variable have two indicators (AC4 and AC8), are eliminated. For other variables, such as continuance commitment the test result eliminated the CC3 indicator, and lastly attitudinal loyalty variable, the results eliminated AL2, AL3, and AL5 indicators.

After the validity tests, the reliability test is conducted to check the reliability of each variable and indicator. The result explained in the Cronbach alpha value of each variable. The reliability test showed that the employee trust variable had a Cronbach alpha value of 0.963 while employee satisfaction had a 0.967 Cronbach alpha value. Other variables such as affective commitment had 0.840 Cronbach alpha, continuance commitment had 0.503, attitudinal loyalty had 0.770 and lastly, behavioral loyalty had Cronbach alpha value of 0.905, which concluded the reliability test. With this test, it is sure that variables here are prepared for the next test, including the model fit, and regression analysis.

Before the analysis explained it is worth noting that normative commitment only had a single indicator (Meyer & Allen, 1991) and the validity and reliability test are not applicable. Whereas the continuance commitment 0.503 Cronbach alpha does not fit the minimum requirement of variable reliability with respondents exceeding 200 (Sarstedt et al., 2017). But the first model fit and regression analysis does include the continuance commitment variable to check the model robustness.

The model fit analysis conducted with the help of SEM AMOS, resulted in several indicators that determined whether the model does fit with the samples. The first model fit and regression analysis does feature the continuance commitment variable, and the results are described in the Table below.

From Table 1, the CMIN/DF of this model is accepted, with a CMIN/DF value of 2.608 which is less than 3.00, in line with CMIN/DF model fit criteria. Whereas the GFI, AGFI, and RMR value each in 0.706, 0.668, and 0.090 are quite far from the minimum for the model fit, which is GFI ≧ 0.90, AGFI ≧ 0.90, and RMR 0.03, with a quite staggering number.

Table 1: Model Fit Analysis Result

Continuously, with CFI and TLI value, this model had a value of 0.856 on CFI and 0.845 on TLI. When the NCP value on 1170.324 is proved to be much higher than what is supposed to be. This means this model is not fit with the respondent, and the result becomes clearer while each variable and hypothesis is analyzed with regression analysis. The regression analysis resulted is provided in the Table below.

In Table 2, the regression analysis result, it is found that the Continuance commitment/CC variable does not fit the model. It has happened because firstly, the continuance commitment did not pass the reliability test, with Cronbach alpha value 0.503, which is less than the required Cronbach alpha value needed on more than 200 respondents, which is 0.700 (Sarstedt et al., 2017). Secondly, it is suspected that the respondent background worked in the education sector even though this suspicion needed further assessment. From the result of the model fit analysis and regression analysis, which pointed out how the continuance commitment variable does not fit the model, this study attempted a second analysis, without the continuance of the commitment variable.

Table 2: Regression Analysis Result

The model fit analysis was conducted without continuance commitment variable results described below. The analysis resulted in CMIN/DF, GFI, AGFI, RMR, TLI, CFI, NCP, and RMSEA value.

Model fit analysis in Table 3 explained the fitness of the model, and the result described below. With CMIN/DF result on 2.646, this result complies with the fitness criteria, which is ≦2.00 or ≦3.00, because the result 2.646 is less than 3.00 value. The GFI result, 0.711, AGFI on 0.674, and RMR on 0.058 are still on the poor acceptability range (Sarstedt et al., 2017), however, without continuance commitment, the model fit result is getting closer to the criteria stated by SEM expert. The RMR value on 0.058 is smaller than the previous result with the continuance commitment variable on 0.090, which made the newest analysis result closer to the criteria which are = 0.03, so did the GFI and AGFI, both are closer to the standard ≧0.90.

Table 3: Model Fit Analysis Result without Continuance Commitment Variable

The CFI and TLI values, each on 0.865 and 0.855 are considered acceptable because the number is still greater than 0.80 (Sarstedt et al., 2017). However, this number is still below the expected standard, ≧0.95 for CFI and ≧0.90 for TLI. But after all, it is sufficiently acceptable to fit as a model to continue the analytical processes (Sarstedt et al., 2017). The RMSEA result on 0.080 comply with the expected criteria, which stated on ≦0.80, hence the RMSEA value is equal with the expected criteria. Whereas the NCP value of 1079.733 is accepted with the NCP criteria.

The model fit analysis resulted in GFI, AGFI, and RMR value that fallen into poor acceptability criteria. However, the CFI, TLI, CMIN/DF, RMR, and NCP all entered the acceptable criteria, which is the minimum needed to continue the regression analysis, which aimed to understand the impact on each variable, and tested the hypotheses constructed in this study. The regression analysis result is available in the Table below.

From Table 4, the result for each hypothesis testing is visible and explains the significance impact on each hypothesis. Employee trust has a positive impact on employee satisfaction according to the first hypothesis (H1). From Table 4, with the β value of 1.122, employee trust does have a positive impact on employee satisfaction and H1 is accepted. The second hypothesis (H2) argued that employee trust has a positive impact on effective commitment (H2a) and normative commitment (H2b) which was then analyzed using the regression analysis. The results of the value of employee trust impact affective and normative commitment both in the number of 0.655 and 0.825 respectively. Hence the hypotheses H2a, the employee trust has a positive impact on affective commitment, and H2b, the employee trust has a positive impact on normative commitment, both hypotheses are accepted.

Table 4: Regression Analysis Result without Continuance Commitment Variable

It is worth noting that hypotheses H2c, which stated that employee trust has a positive impact on continuance commitment are not applicable for the testing. This is due to the reason for continuance commitment variable elimination from the model in the second analysis, and from this result, the H2c hypothesis is rejected.

Further hypotheses, in this case, the 3rd hypotheses (H3) stated that employee satisfaction has a positive impact on affective commitment (H3a) and has a positive impact on normative commitment (H3b), on Table 4.4 the result could be found also. The β value of 0.294 for the affective commitment and 0.123 for the normative commitment both mean that employee satisfaction does have a positive impact on employee affective commitment and normative commitment. From this result, both hypotheses H3a and H3b are accepted. On the other hand, hypotheses H3c which stated that employee satisfaction has a positive impact on continuance commitment is rejected because of the elimination of the continuance commitment variable from the model.

The fourth hypothesis (H4) that stated that affective commitment has a positive impact on employee attitudinal loyalty (H4a) and behavioral loyalty (H4b) which can be seen in Table 4.4. With the β values of 0.199 and 0.264, each suggested that affective commitment had a positive impact on attitudinal loyalty and behavioral loyalty, so the H4a and H4b hypotheses are accepted. The fifth hypothesis (H5), stated that normative commitment has a positive impact on attitudinal (H5a) and behavioral loyalty (H5b) also confirmed. From Table 4.4, it is found that normative commitment does have a positive impact on attitudinal loyalty, with β 0.405, however, normative commitment impacts the behavioral loyalty which was found to be not-significant with β 0.014 even though this is still on positive impact. Considering the previous study resulted in β value only 0.001 on the impact of normative commitment toward behavioral loyalty (Yao et al., 2019), this means that in this study β value of 0.014 can be accepted as an insignificant positive impact. Hence both hypotheses H5a and H5b are accepted.

The sixth hypothesis (H6) which includes the continuance commitment variable is non-applicable for the testing, which means that H6a, which stated that continuance commitment has a positive impact on attitudinal loyalty, and H6b stated continuance commitment has a positive impact on behavioral loyalty is rejected.

Lastly, the result from the seventh hypotheses (H7) stated that attitudinal loyalty has a positive impact on behavioral loyalty, Table 4.4 confirmed the result. With the β value of 0.838, this means that attitudinal loyalty does have a positive impact on behavioral loyalty, and this result confirmed and accepted the H7 hypotheses. The Figure below showed the theoretical framework after the analysis and testing.

Figure 2: The Theoretical Framework after Testing

This study was conducted on 3 (three) different Islamic education institutions, namely Yayasan Wakaf Bina Amal Semarang in Semarang City, Central Java. The others are Yayasan Nur Hidayah Surakarta in Surakarta City, Central Java, and Yayasan Tarbiyatul Mukmin Pabelan, in Magelang, Central Java. Total respondents gathered from this study is 261 with 235 responses passed through the outlier test. All of the respondents have come from the millennial generation, they were born between the years 1980–2000, and these millennials occupied various organizational positions in each institution.

From the validity and reliability testing, the Cronbach alpha for each variable is 0.963 for the employee trust variable, 0.967 for employee satisfaction, and 0.840 for affective commitment. The rest of the variables, such as continuance commitment had a Cronbach alpha value of 0.503, whereas each attitudinal and behavioral loyalty had Cronbach alpha value of 0.770 and 0.905 respectively. With a single indicator, the normative commitment validity and reliability test are non-applicable, and with Cronbach alpha of 0.503, the continuance commitment variable did not pass the reliability test. However, the first model fit and regression analysis still included the continuance commitment variable.

The first model fit analysis, with Table 1 served the result showed that this model has an accepTable fit, however, when looking at the regression analysis result shown in Table 2, it is clear that the continuance commitment variable does not fit in the model. It has mainly happened because firstly, this variable did not pass the reliability test. The Cronbach alpha value of 0.503 fits with the minimum Cronbach alpha which is 0.5 minimum. However, with 200 more respondents, each variable should at least had a 0.7 Cronbach alpha value (Sekaran and Bougie, 2016; Sarstedt et al., 2017). Hence, with this result, the continuance commitment variable did not pass the reliability test and this affected the regression analysis result.

Secondly, it has something to do with the respondents. According to Meyer and Allen (1991), continuance commitment is a commitment that occurred with skill-based acquired by the employee, because they work in the organization (Meyer & Allen, 1991). This is not the case for respondents who work in the educational field, their knowledge or experience base enables them to work in the other fields especially with 73.3% of the respondents already acquired a bachelor degree. This is linear with the Indonesian education problem, it is commonly found that a teacher or employee had different or even double job outside their profession. The result is an education process that is not optimal for enhancing students’ capabilities (Nasution, 2006). This problem was exacerbated by the lack of understanding between the stakeholders in Indonesia education system, shown by how teachers and staff did not have access to at least decent pay (Suharno et al., 2020).

And lastly, this study’s main object is the millennial generation. This generation is very much able to acquire other knowledge by themselves with their familiarity with the information technology (Pyöriä et al., 2017). Hence the better way to dealing and nurturing with the millennial organizational commitment and loyalty is not through materialistic aspects but a better and clearer career development path (Holtschlag et al., 2020), or a group that valued more personal contribution (Scales & Quincy Brown, 2020) becoming a way on dealing with millennial. Millennial did not view the skills they gained through working in a certain organization is important because they are more than able to expand their skill and knowledge outside the organization, and this made the continuance commitment does not comply with them.

Using another organizational commitment questionnaire could become the solution to portray the millennial generation’s overall view on organizational commitment. The three conceptions is an idea conveyed by Meyer and Allen (1991). However, there is a previous study according to organizational commitment which resulted in the development of the Organizational Commitment Questionnaire (OCQ) by Porter (Porter et al., 1974). A future study could utilize this questionnaire, although this questionnaire did not perfectly depict the idea given by Meyer and Allen (1991). Hence this study continues using Meyer and Allen’s concept because one of the main goals for this study is to test the robustness of the developed model.

The second model fit and regression analysis was conducted without the continuance variable. This resulted in an acceptable model explained in Table 3, with the regression analysis on Table 4 that explained the impact for each variable, and the result of hypothesis testing. Hypothesis 1 (H1) that employee trust has a positive impact on employee satisfaction is accepted. This result is similar to the previous studies (S.-X. Liu et al., 2020; Roberts & David, 2020), which concluded that the millennial generation valued trust, and with trust in their manager, it positively impacted job satisfaction.

The second hypothesis (H2) had 3 components, first, that employee trust has a positive impact on affective commitment. Second and third, employee trust has a positive impact on normative commitment, and continuance commitment. The first and second component (H2a and H2b) is accepted, thus employee trust had a positive impact on affective and normative commitment. This result confirmed the previous studies (S.-X. Liu et al., 2020; Yao et al., 2019), and with the millennial generation case, it also strengthened the fact that trust, played important role in millennials organizational commitment (Graybill, 2014). The third component, H2c contains continuance commitment is non-applicable because the variable is removed from the second analysis.

The third hypothesis also had 3 (three) components that consisted of H3a, employee satisfaction has a positive impact on affective commitment. The second, or H3b stated that employee satisfaction has a positive impact on normative commitment, and the last component, H3c stated that employee satisfaction has a positive impact on continuance commitment. The result confirmed that both H3a and H3b are accepted, hence this study confirmed previous research results (Kim & Yang, 2020; Sila & Širok, 2018; Zopiatis et al., 2014). This means that the millennial generation that is satisfied with the job, tends to have a greater organizational commitment toward the organization. However the third component, H3c, the analysis result is non-applicable because the variable is removed from the model.

The fourth and fifth hypotheses (H4 and H5) both had two components. The fourth hypothesis first component (H4a) stated that affective commitment has a positive impact on attitudinal loyalty, and the second component (H4b) stated affective commitment has a positive impact on behavioral loyalty. The fifth hypothesis also had two components, each stated that normative commitment has a positive impact on attitudinal loyalty (H5a) and normative commitment has a positive impact on behavioral loyalty (H5b), and the result of the hypotheses testing accepted all hypotheses. This is can be seen in Table 4.4 and resulted accordingly with previous research (Yao et al., 2019; Izogo, 2015; Zopiatis et al., 2014).

However, in the previous research, hypothesis H5b on normative commitment impact on behavioral loyalty, this study confirmed that it had an insignificant positive impact. This result is quite contrary to the finding of a previous study, Yao (2019) that concluded that normative commitment had a very insignificant impact on behavioral loyalty. These findings concluded that further research is needed to understand the role of normative commitment on behavioral loyalty because the seventh hypothesis (H7) proved that both loyalties are important in managing employees and in this study, especially millennial employees.

The seventh hypothesis (H7) stated that attitudinal loyalty has a positive impact on behavioral loyalty, and the hypotheses testing accepted this statement. This result also confirmed previous studies (Antwi et al., 2020; Yao et al., 2019; Bandyopadhyay & Martell, 2007), this means that attitudinal loyalty does provide better behavioral loyalty among millennial employees. This also means, variables like the affective and normative commitment that does have a significant positive impact on attitudinal loyalty, played an indirect important role in employee behavioral loyalty because from this result, attitudinal loyalty impacted behavioral loyalty significantly.

Lastly, the sixth hypothesis that consisted of two components, H6a and H6b is non-applicable for testing. It has happened because H6a and H6b had continuance commitment variable, which has been removed from the model in the second analysis, hence both H6a and H6b that stated continuance commitment has a positive impact on attitudinal loyalty and behavioral loyalty was rejected in this study. This finding is contradicted with a previous study, that found continued commitment has a positive impact on both attitudinal and behavioral loyalty (Yao et al., 2019). Further research is needed in these terms, but this study suggested bigger respondents if future researchers wanted to conduct this type of study on the educational environment, or tested the robustness of this model on different job environments.

This study concluded that millennial employees can construct their organizational commitment and loyalty toward the organization. All variables tested, employee trust, employee satisfaction, affective commitment, normative commitment, attitudinal loyalty, and behavioral loyalty all showed a significant impact on each hypothesis tested. This means that organizations and businesses needed to enhance their effort to nurture employee trust and employee satisfaction, especially among millennial staff and employees. The organizations cannot rely on the perspective that their work experience could provide necessary skills which in turn could nurture millennial generation continuance commitment toward organization because the millennials already possessed all the tools needed to expand their set of skills themselves.

Secondly, although working in a religious work environment such as an Islamic educational institutions, millennials still had an organizational commitment and loyalty toward the organization. This provides an insight that Indonesia Islamic education provides a basis for moderate Muslims, that still valued things like organizational good governance (Masuda & Yudhistira, 2020). A religious institution cannot rely only upon the religion itself. Creating trust, employee satisfaction is much more important to grow the millennial generation's commitment and loyalty toward the organization. This also means that all three of the Islamic educational institutions could provide trust and satisfaction among their employee.

Lastly, the millennial generation still holds a pristine view toward organizational commitment and loyalty, especially if that started from trust and satisfaction. This means the millennial generation in Indonesia valued commitment and loyalty. This could occur as long as their trust in the organization, and their satisfaction as an employee, are well maintained.

5. Conclusion

This study concluded that organizational commitment and loyalty is an integral part of the millennial generation that worked in Islamic educational institutions. This finding strengthened the fact that the millennial generation that entered the workplace can bring up their commitment and loyalty toward the organization if the organization provided trust and satisfaction among its employee. This finding also explained that the developed model can depict the millennial generation’s organizational commitment and loyalty in educational organizations. Future research could further emphasize in this model with the same organizational type, however, the respondents number should be increased to gain a better result. Secondly, this model’s robustness should be tested in another organizational and cultural background. Lastly, another organizational commitment questionnaire could be utilized to enrich organizational commitment studies.

[script] var kistiSpellChecker_replaceCKEditorOptionId = setInterval(function() { if ( ! window.CKEDITOR ) return; clearInterval(kistiSpellChecker_replaceCKEditorOptionId); var isMarkTagAllowed = function(config) { if (config.allowedContent === true) return true; if (typeof(config.allowedContent) === 'string' && config.allowedContent.includes('mark') ) { return true; } if (typeof(config.allowedContent) === 'object' && config.allowedContent.keys().includes('mark') ) { return true; } if (typeof(config.extraAllowedContent) === 'string' && config.extraAllowedContent.includes('mark') ) { return true; } if (typeof(config.extraAllowedContent) === 'object' && config.extraAllowedContent.keys().includes('mark') ) { return true; } return false; }; for (var key in CKEDITOR.instances) { var editor = CKEDITOR.instances[key]; var element = editor.element; var config = editor.config; if ( isMarkTagAllowed(config) ) { continue; } config.extraAllowedContent = 'mark; ' + (config.extraAllowedContent || ''); var originEditorKey = null; for (var key in window) { if ( window[key] === editor ) { originEditorKey = key; break; } } var extraProperties = { listSeq: editor.listSeq // KISTI 기계학습데이터 구축 작업화면 대응 }; console.log('listSeq before destroy', editor.listSeq); editor.destroy(); window[originEditorKey] = editor = CKEDITOR.replace(element, config); for (var key in extraProperties) { editor[key] = extraProperties[key]; } console.log('extraProperties after recreate', extraProperties); } }, 100);[/script] [script] var kistiSpellChecker_replaceCKEditorOptionId = setInterval(function() { if ( ! window.CKEDITOR ) return; clearInterval(kistiSpellChecker_replaceCKEditorOptionId); var isMarkTagAllowed = function(config) { if (config.allowedContent === true) return true; if (typeof(config.allowedContent) === 'string' && config.allowedContent.includes('mark') ) { return true; } if (typeof(config.allowedContent) === 'object' && config.allowedContent.keys().includes('mark') ) { return true; } if (typeof(config.extraAllowedContent) === 'string' && config.extraAllowedContent.includes('mark') ) { return true; } if (typeof(config.extraAllowedContent) === 'object' && config.extraAllowedContent.keys().includes('mark') ) { return true; } return false; }; for (var key in CKEDITOR.instances) { var editor = CKEDITOR.instances[key]; var element = editor.element; var config = editor.config; if ( isMarkTagAllowed(config) ) { continue; } config.extraAllowedContent = 'mark; ' + (config.extraAllowedContent || ''); var originEditorKey = null; for (var key in window) { if ( window[key] === editor ) { originEditorKey = key; break; } } var extraProperties = { listSeq: editor.listSeq // KISTI 기계학습데이터 구축 작업화면 대응 }; console.log('listSeq before destroy', editor.listSeq); editor.destroy(); window[originEditorKey] = editor = CKEDITOR.replace(element, config); for (var key in extraProperties) { editor[key] = extraProperties[key]; } console.log('extraProperties after recreate', extraProperties); } }, 100);[/script]