Determining the Effect of Banking Service Quality on Customer Loyalty Using Customer Satisfaction as a Mediating Variable: An Applied Study on the Jordanian Commercial Banking Sector

Banks operate in an environment characterized by competition between banking service providers and the continuous development of financial technology. On this basis, the quality of banking services determines the success of banks. This study aims to test customer satisfaction as a mediating variable in the relationship between banking service quality and customer loyalty using the convenience sample method. Using the SmartPLS software, the data analysis and hypothesis test results show that the quality of banking service in all its dimensions positively affects customer loyalty. Moreover, customer satisfaction mediates such a relationship. Accordingly, the Jordanian commercial banking sector should focus on the tangibility dimension and influence customer satisfaction to promote customer loyalty.


Introduction
The Jordanian financial sector is characterized by competitiveness. The financial sector Consists of various institutions. Where it includes the central bank, thirteen commercial banks, seven foreign commercial banks, three Jordanian Islamic banks, and a foreign Islamic bank, in addition to specialized credit institutions, exchange companies, and representative offices. The banking sector indicators recorded growth at the end of 2019 in various fields. Assets increased by 5.4%, and credit facilities and deposits grew by 3.7% and 4.3%, respectively (Central Bank of Jordan & Association of banks in Jordan, 2019).This statistics indicates that the banking sector enjoys a large market share in terms of assets, facilities and deposits, which requires continuous improvement and development in banking services to gain a competitive advantage reflected in customer satisfaction and loyalty.
Clients have a significant role in the success of the banking business. Hence, the banking departments should provide banking services characterized by innovation and diversification. These services should be in line with the desires of customers or even exceed their expectations. Ultimately, they should increase customer satisfaction and guarantee their loyalty. This study highlights a pressing topic in the banking industry.
2. Does the quality of banking service affect customer satisfaction?
3. Does customer satisfaction affect customer loyalty? 4. Does customer satisfaction mediate the relationship between banking service quality and customer loyalty? Kranias and Bourlessa (2013) found that adequate proximity to a bank significantly affects customer loyalty. Consequently, Greek banks have redesigned their services to increase customer loyalty. Kheng et al. (2010) evaluated the banking sector in Malaysia. They asserted that banks should improve service quality to promote customer loyalty. Leninkumar (2016) revealed that various dimensions of service quality (i.e. Tangibility, reliability and empathy) significantly affect customer loyalty in the Sri Lankan banks. Likewise, Onditi et al. (2012) found that the quality of service greatly affects customer loyalty, given that the quality of service determines customer loyalty. Liu and Wang (2017) concluded that organizations that work to improve service quality through the influence customer loyalty could enhance their performance effectively. On the basis of the above discussion, this study proposes the following hypothesis:

Literature Review & Development Hypotheses
Hypothesis 1: The quality of banking service positively affects customer loyalty. Raza et al. (2020) and Hammoud et al. (2018) showed that the quality of banking service in its five dimensions has a large and positive effect on customer satisfaction. Al-Azzam (2015) mentioned that the dimensions of banking services positively affect customer satisfaction in Arab Banks. Nautiyal (2014) found that the dimensions of banking service quality affect customer satisfaction in Delhi. Mohammad and Alhamadani (2011) determined whether service quality is an important precedent for customer satisfaction in the Jordanian commercial banking sector. They concluded that the dimensions of service quality have a great effect on customer satisfaction. Recently, Vetrivel et al. (2020) extended the literature to the dimensions of Internet banking services. They revealed that these dimensions positively affect customer satisfaction. On the basis of the above discussion, this study proposes the following hypothesis: Hypothesis 2: The quality of banking service positively affects customer satisfaction.
Prior studies (e.g. Mohsan et al., 2011;Arokiasamy, 2013) concluded that customer satisfaction has positive and negative relations to customer loyalty and intention to switch. Chiguvi and Guruwo (2017) found a positive relationship between customer satisfaction and customer loyalty. Leninkumar (2017) also found significant positive relationships between customer confidence and loyalty, customer satisfaction and loyalty and customer satisfaction and trust. Moreover, customer satisfaction significantly affects customer loyalty. On the basis of the above discussion, this study proposes the following hypothesis: Hypothesis 3: Customer satisfaction positively affects customer loyalty. Fida et al. (2020) showed that the quality of banking service and customer satisfaction and loyalty are highly correlated. The respondents, on average, agreed with the dimensions of the banking service in Oman. Siddiqi (2011) concluded that all features of the quality of banking service have a positive correlation with customer satisfaction. In turn, customer satisfaction positively influences customer loyalty in the banking sector in Bangladesh. Khatoon et al. (2020) found that the quality of banking service dimensions has a significant and positive impact on customers' purchasing intentions. Customer satisfaction (as the mediating variable) has a partial effect on the relationship between the independent and dependent variables. Khan and Fasih (2014) explored the relationship between banking service quality and customer satisfaction and loyalty. Using all of the former's dimensions, they found significant and positive correlations. On the basis of the above discussion, this study proposes the following hypothesis: Hypothesis 4: Customer satisfaction mediates the relationship between the quality of banking service and customer loyalty.

Study Contribution
This study is a continuation of previous studies related to the issue of the quality of banking services. It focuses on continuously enhancing the quality of banking services. It analyses and evaluates customer behaviour to help banks meet their various needs to achieve their loyalty to the service. This study is characterized by the following features: Environment and sector: This research applies to branches of Jordanian commercial banks in Tafila Governorate. Its economic activity is diverse in funding sources, foreign investments, competitive small and medium enterprises and many natural resources. This situation also indicates the diversity of clients in this governorate.
This study is one of the first few studies from the perspective of the researcher which determines whether customer satisfaction mediates the relationship between banking service quality and customer loyalty. , especially in the middle-east environment.

Research Model
On the basis of related studies, the researcher developed the model shown in Figure (1).

Population and Study Sample
The study population consisted of all customers in the Jordanian commercial banking sector, including 13 banks.
The study community is difficult to restrict, given the expansion of the Jordanian commercial banking sector and an increase in customers. Thus, this study only considered clients in Tafila Governorate. The study sample only included Jordan Ahli Bank, Arab Bank, Housing Bank and Cairo Amman Bank.

Data Collection
The researcher developed a study tool (questionnaire) to obtain data from the respondents. The study tool comprises four sections. The first section explains the demographic data, and the second section discusses the independent variable (i.e. Quality of banking service). The discussion follows prior studies (e.g. Raza et al., 2020;Fida et al., 2020;Siddiqi, 2011;Lomendra et al., 2019;Panichpathom, 2016). The third section explains the mediating variable (e.g. Customer satisfaction) following existing research (e.g. Hammoud et al., 2018;Siddiqi, 2011;Awuor, 2009). The fourth section covers the dependent variable and follows the literature (Mohsan et al., 2011;Leninkumar, 2017). The study tool uses a Likert-type scale ranging between strongly agree and strongly disagree.
On this basis, the researcher selected a convenience sample and distributed 300 questionnaires. Five questionnaires were excluded from the 290 retrieved questionnaires. A total of 285 questionnaires was considered suitable for analysis, accounting for 95% of the total distributed questionnaires. Table 1 presents the demographic characteristics of the study sample.

Validity and Reliability
After developing the questionnaire, the researcher presented it to a group of seven arbitrators with experience and competence in financial and banking sciences and business administration. The necessary adjustments were made to obtain the final version of the questionnaire. On this basis, the study tool (the questionnaire) could achieve the research aim.
The researcher used Cronbach's alpha coefficient to identify the extent of internal consistency between the paragraphs of the questionnaire. The scale test result is statistically acceptable if the Cronbach's alpha value is greater than the threshold of 0.70 (Hair et al., 2010). The closer the Cronbach's alpha to 1, the higher the degree of internal consistency (Sekaran & Bougie, 2009).
The Cronbach's alpha values of the study variables ranged from 0.808 to 0.915), as shown in Table 2. All of them were greater than 0.70 and considered acceptable for scientific research, indicating internal consistency and stability. Thus, the questionnaire was valid for analysis.
To verify the internal consistency, the researcher tested the composite reliability. With a value ranging between 0 and 1, its value is considered acceptable if it is greater than 0.70. In this study, all the independent variables were acceptable, as shown in Table 2. The average variance extracted (AVE) value is considered acceptable when it exceeds 0.50. In this study, the AVE was within acceptable limits. Table 2 indicates that all factor loadings reach more than 0.70 and are within the acceptable limits of this test, except for Paragraph No. 4 within the (tangibles) axis and Paragraph No. 29 within the (customer loyalty) axis. Hence, these paragraphs were dropped in the second-level analysis. Moreover, multicollinearity was not a problem amongst the independent study variables. All tolerance values were greater than 0.2, and all VIF values were less than 5 and within the permissible limits.   Table 3 shows the arithmetic mean, standard deviation and the relative importance of all study variables. The arithmetic mean and the relative importance of all the variables reach the average level. Thus, the quality of banking service in all its dimensions, customer satisfaction and customer loyalty are moderate, considering the study sample under examination and analysis.  Table 4 shows the test results of Hypothesis 1. The effect of the quality of banking service has a positive explanatory power of 43.1% of customer loyalty. This effect is statistically significant at the 0.05 level and supported by the t-test result of 19.403 which is greater than its tabulated value. Therefore, Hypothesis 1 is accepted. The quality of banking service in all its dimensions has a positive effect on customer loyalty.

Analysis of Hypothesis 1 Test Results
Moreover, all dimensions of banking service quality are statistically significant at a level of less than 0.05. The dimension of empathy (β = 0.207) has a greater effect on customer loyalty than reliability (β = 0.194) and responsiveness (β = 0.155).  Table 5 shows the test results of Hypothesis 2. The value of R² (0.605) indicates that the quality of banking service has a positive explanatory power amounting to 60.5% of the change in customer satisfaction, and this effect is statistically significant at 0.05 level. This finding is enhanced by the t-test value of 31.776, which is greater than its tabulated value. Thus, Hypothesis 2 is accepted. The quality of banking service in all its dimensions has a positive effect on customer satisfaction.

Analysis of Hypothesis 2 Test Results
The results in Table 5 also indicate that all the dimensions of banking service quality are statistically significant at the 0.05 level. This finding is consistent with those of Hypothesis 1 in the first and second dimensions. The dimension of empathy (β = 0.245) has the greatest effect, followed by reliability (β = 0.231) and safety (β = 0.181).

Analysis of Hypothesis 3 Test Results
The results in Table 6 refer to the test results of Hypothesis 3. Customer satisfaction has a positive interpretation of customer loyalty (R² = 0.605). This result is confirmed by the t-test value of 72.004, which is greater than its tabulated value. Thus, Hypothesis 3 is accepted. Customer satisfaction has a positive effect on customer loyalty.

Analysis of Hypothesis 4 Test Results
The results in Table 7 refer to the test results of Hypothesis 4. It states that customer satisfaction mediates the relationship between banking service quality and customer loyalty. To test the intermediate variable hypothesis, the researcher relied on the method proposed by a prior study (Preacher & Hayes, 2008), which is based on the following two tests:

1-Bootstrap the indirect effect (Total effect)
The effect of the independent variable on the dependent variable is tested (the total effect). The first condition indicates a statistically significant relationship between the independent and dependent variables, even with the mediating variable. In Table 7, the effect of the quality of banking service has a positive explanatory power amounting to 79.3% of the change in customer loyalty in the presence of customer satisfaction. Customer satisfaction as an intermediary variable has increased the above direct effect compared with Hypothesis 1 (43.1%) in Table 4. The t-test value is 16.363, which is greater than its tabulated value. The second condition verifies the effect of the mediating variable (Preacher & Hayes, 2008). The upper and lower limits were bootstrapped at a confidence level of 95%. In Table 8, the lower and upper limits of the confidence interval are 0.600 and 0.791.
Zero does not interrupt these two values. Thus, customer satisfaction mediates the relationship between banking service quality and customer loyalty, supporting Hypothesis 4.  Figure 3. the model of standardized loading and path coefficient for the trimmed model, after deleting questions (4&29)

Conclusion
The arithmetic averages of all the variables reach the average level, indicating that customer satisfaction and loyalty are generally average for the quality of banking services provided to the sample under examination and analysis.
The researcher also validated the SERVQUAL model in the five dimensions investigated in this study using SmartPLS. The dimensions with the greatest effects on customer loyalty are empathy, reliability and responsiveness. The evaluation of customer satisfaction shows empathy as the best dimension, followed by reliability and safety. The weakest dimension is the tangibility dimension.
The results of the SmartPLS test support Hypothesis 1. The quality of banking service in all its dimensions has a positive effect on customer loyalty. This finding agrees with prior studies (Kheng et al., 2010;Leninkumar, 2010;Liu & Wang, 2017). Thus, the quality of banking services is a critical determinant of customer loyalty. It should stay consistent with customer behaviour and perception management, which is positively reflected in the banks' market share.
The analysis results support Hypothesis 2. The quality of banking service in all its dimensions has a positive effect on customer satisfaction. This finding is consistent with prior studies (Raza et al., 2020;Hammoud et al., 2018;Al-Azzam, 2015;Nautiyal, 2014;Mohammad & Alhamadani, 2011). Thus, banks should improve all dimensions of their service quality. In turn, they can enhance the satisfaction of existing customers and attract more customers in the future.
The test results support Hypothesis 3. Customer satisfaction has a positive effect on customer loyalty. This finding is consistent with prior studies (Leninkuman, 2017;Arokiasamy, 2013;Mohsan et al., 2011;Chiguvi & Guruwo, 2017). Thus, if customers are satisfied with the provided banking services, they will stay loyal. Satisfaction precedes loyalty; that is, the greater the customer satisfaction, the greater their loyalty to the bank.
Finally, customer satisfaction mediates the relationship between the quality of banking service and customer loyalty. This finding is partly consistent with prior studies (Fida et al., 2020;Siddiqi, 2011;Khan & Fasih, 2014). Thus, mutual relationships exist between the quality of banking service and customer loyalty through customer satisfaction as a mediating variable. These relationships enhance and develop the options offered in banking services, which lead to advantages in the competitive banking market.

Recommendations
On the basis of the above results, the Jordanian commercial banking sector should continuously develop and improve the quality of its banking services, with a focus on the tangible dimension.
Banks should influence customer satisfaction to achieve their loyalty whilst paying attention to their needs.
Banks should invest in factors that may positively affect customer satisfaction and loyalty, such as reducing the processing time, adding more parking spaces and handling complaints quickly. Bank employees should show sincere concern to clients' needs to provide effective services.
Finally, banks should consider increasing their branches in the governorate of Tafila to provide more services in line with the wishes and aspirations of customers.

Limitations and Future Research Directions
This study was limited to Tafila governorate. Hence, the results should not be generalized easily to other contexts. Future research should focus on larger cities with a higher number of clients.
Future works should also cover other variables, such as confidence, bank location and confidentiality, and customers of different demographic and cultural variables.