Customer Loyalty towards Islamic Banks: The Mediating Role of Trust and Attitude

: Islamic banks have faced many challenges either from rivals or customers who look for better products and services. This study examined both the direct and mediating effects of selected variables inﬂuencing Islamic banks’ customer loyalty. The direct links were satisfaction, attitude towards Islamic banks, customer trust, and religious obligation to customer loyalty. On the other hand, the indirect links were through the mediators’ trust and attitude. A sample of 416 working Muslims in the UAE was collected and analysed. We used PLS-SEM techniques and the SmartPLS 3.0 software package to investigate customers’ loyalty towards Islamic banks. It was found that all the direct links were positive and signiﬁcant, and were mediating relationships. This result implied that the respondent’s attitude toward Islamic banks and customer trust plays a signiﬁcant role in understanding the relationship between customer satisfaction, religious obligation, and customer loyalty. In other words, both customer satisfaction and religious obligation indirectly created customer loyalty through the mediators. This study is one of the few studies that incorporate religious obligation as one of the determinants of customer loyalty.


Introduction
Financial intermediaries create a link between lenders and borrowers to mobilise funds [1]and savings into investments to maintain growth of the production process [2,3].Financial intermediaries involve both lenders and borrowers in trust and attitude to ensure the equal rights of each party [4,5].An essential element of trust is satisfaction in a financial transaction, which leads to customer loyalty with Islamic banks [6,7].Moreover, the most crucial element of attitude is the religious belief in a financial transaction, which can reflect loyalty toward Islamic banks [8][9][10].Customers experiencing low levels of satisfaction may lead to diminishing public trust in Islamic banks.
Financial development consists of the development of both financial institutions and markets.The process of financial development is a problematic phenomenon for financial services organisations, and it is influenced by different factors, such as satisfaction, attitude, trust, religious obligation, and loyalty [8,11].Low levels of customer satisfaction as well as the lack of religious belief and trust may cause customers to be misled when they receive financial services from banks, which, in turn, ensure stiff competition to retain customers [9].Overall, the total value of Islamic bank assets worldwide has been estimated at US$1.5 trillion in 2018, and it was estimated that it would reach 3.2 trillion by 2020 [12].Islamic banking in the UAE has been growing fast in the past two decades.For instance, the Islamic bank penetration rate jumped from 47% in 2015 to 58% in 2020, while conventional banks decreased from 70% in 2015 to 64% in 2020.In addition, fully-fledged Islamic banks' total assets grew by 5% in 2020, representing 19% of the total assets of banks in the UAE.Islamic banks are banks that must follow Islamic law in their transactions.Islamic banks are prohibited from dealing with interest, gambling activities, and selling what they do not own, among others [1].In addition, Islamic banks' transactions are either based on trading or partnership, and parties have to share the risk and the benefits.Therefore, Muslims using Islamic banks means simultaneously combining the business aspect and the religious obligation aspect.However, given that the dual banking system exists in the UAE, Islamic banks face marketing challenges that are not faced by their rivals [13].Therefore, customer loyalty in Islamic banks does not have the traditional marketing aspect of the business, but has religious and moral dimensions.Customer loyalty is one of the common and essential factors banks need to emphasise.This is because the banking sector is part of the service industry and differentiation of service and product is crucial to maintain customer loyalty.Customer loyalty determines the intensity of the long-term relationship between customers and banks.The more loyal the customer, the more the customer is spending on the firm's products and services, and the higher the customer advocacy of the firm [14].This, in turn, leads to higher revenue obtained by the bank and, eventually, higher profitability.Therefore, better loyalty leads to a better bank perspective given high profit.
To ensure their success, Islamic banks must remove any customer doubts about their financial transactions.Financial transactions with Islamic banks are compliant with Shariah law [15,16].Islamic banks face stiff competition from conventional banks concerning the financial services and products on offer [17,18].Customer retention and loyalty are the fundamental strategies for a financial service provider's business sustainability in the longterm evolving economy [3].Financial service providers, particularly Islamic banks, should ensure that their customers are satisfied, trusting, and comfortable with their attitude towards financial transactions with products and services.Customer loyalty toward banks is reflected by customer retention, which, in turn, generates revenue, increased deposits, and evolving marketing.
The significant role of customer loyalty in different organisations is widely highlighted in consumer-oriented product offerings and services [2,17,19].However, the researchers above ignored the importance of empirical investigations concerning trust, religious beliefs, and Islamic banks' attitudes toward financial transactions.Besides, the Covid-19 pandemic currently presents some downside risks to the development of Islamic banks [20].For example, Islamic banks' products are based on buy and sell contracts as well as partnerships.In the current Covid-19 pandemic, finding customers who will take the risk is difficult and buying and partnering might not be favourable to the customer [21].It is crucial to investigate how customer satisfaction and religious beliefs reflect trust and attitude towards loyalty for Islamic banks.
Previous studies that looked at religious motivation from an Islamic banking perspective found that it is not one of the determining criteria [22][23][24].On the other hand, some studies found religious obligation related to behavioural and attitudinal factors [25,26].Given that there are mixed results in the literature about the impact of religious obligation on loyalty and the limited studies looking at this link, it is crucial to investigate this relationship.While religious obligation is regarded as an important determinant of customer attitude and behaviour in general [26,27], surprisingly few studies have examined this factor as a loyalty determinant in Islamic banking [9,28].Additionally, current studies in Islamic banking loyalty primarily focus on service quality and customer satisfaction as loyalty drivers [13,28,29].
Driven by these knowledge gaps, this study measured customer loyalty by adopting customers' satisfaction, religious obligation, trust, and attitude towards financial transactions with Islamic banks.Therefore, this study aimed to investigate the direct links between customer satisfaction, attitude, trust, and religious obligation, influencing customer loyalty toward Islamic banks.In addition, this study examined whether trust and attitude mediated the relationship between satisfaction and loyalty, as well as religious obligation and loyalty toward Islamic banks.

Literature Review and Hypotheses Development
From the Islamic banks' perspective, ref. [30] investigated the determinants of customer loyalty in Islamic banks in Indonesia.They indicated that Islamic banks should create and promote various products that follow Sharia Law and, at the same time, match the ones provided by competitors.In addition, others found that Islamic banks should improve their services to maintain current customers and attract new ones.This in turn will enhance long-term relations between customers and banks, and eventually lead to loyalty [31].Moreover, it was found that loyalty in Islamic banks can be enhanced by improving service commitment, managing conflict, and better communication with customers [5,32].Therefore, Islamic bank loyalty differs from the conventional bank in the sense that Islamic banks do not only provide a service, but there is always a spiritual dimension emerging from the Islamic laws [13,33].Ref. [34] studies the link between service justice, service quality, social influence, image with customer satisfaction and loyalty, and the moderating impact of ownership in the banking sector in Bangladesh.They found that service justice, service quality, and social influence are directly linked to customer loyalty.Similarly, ref. [35] studies the service quality impact on behavioural intention of banking customers in Bangladesh.They found a positive link between service quality and behavioural intention in banking customers.

Customer Loyalty
Banks provide homogenous services, regardless of whether they are Islamic or conventional.The differences only relate to the added value of the services offered and who those services target.As a result, the banking sector faces stiff competition worldwide to ensure its competitiveness to retain customers [36].In this respect, Islamic banks are no different, and they join conventional banks in the race to retain customers.Bank customer loyalty is defined as the ability to maintain customers' patronage over rival banks and recommend the services provided to other customers [37].It is an ongoing process that is not concluded with a single purchase, but continues with repeated purchases of specific services.Hence, bank customer loyalty is the long-term willingness and commitment to repeat the purchase of products or services [38].The importance of loyalty stems from repeated purchases, and recommendations of services to other customers generate higher profit levels, reduced customer turnover, and reduced marketing costs [39,40].In addition, customer loyalty can benefit customers by minimising risk and cost [41].However, customer loyalty is difficult to achieve and is challenging to maintain [42].Bank customer loyalty, by itself, embeds a certain level of trust in the bank's values and the satisfaction of customer needs.Refs.[43,44] explained that the formation of customer satisfaction could influence loyalty toward banks, and the formation of customer satisfaction could influence bank trust.Refs.[45,46] showed that customer satisfaction had been acknowledged as a crucial factor influencing customer loyalty, while refs.[47,48], emphasised the link between trust and customer loyalty.

Customer Trust
Refs. [49,50] defined trust as a willingness to have self-confidence in an exchange partner.Others, such as [51] defined trust as confidence in the reliability and quality of a firm's goods and services.Similarly, ref.
[52] indicated that trust was more concerned with accepting the level of risk or adverse outcomes while being confident about future outcomes.Ref. [53] indicated that customer trust is generated from their experience, feeling of security, and confidence in a bank's ability to follow the rules and regulations and behave honestly.According to [54], trust is subjective and difficult to measure, as it is related to an individual's beliefs and is not based on hard facts.Previous studies regarding Islamic banks have found that customers of Islamic banks have used religion as the main criteria for selecting an Islamic bank [10,22].As religion is significant in influencing the decision-making processes of Muslim customers, this results in a relationship based on trust.Customers of Islamic banks believe (i.e., trust) that Islamic banking operates based on Shariah principles and behaves honestly [55,56].Although religious considerations are not the only factor influencing Muslim customers' decisions to patronise Islamic banks, it is still considered one of the most critical factors.Thus, an Islamic bank's philosophy is seen as an indicator of honesty and trustworthiness, enhancing the relationship between customers and the bank [13,57].Previous studies have found mixed results regarding the link between trust and other variables, such as the intention to purchase, customer satisfaction, and customer loyalty toward Islamic products and services.Some studies have found that trust is positively related to the satisfaction and loyalty of Muslim customers [58,59].However, other studies have found that trust influences customer loyalty indirectly rather than directly [11,60].

Customer Trust, Satisfaction, and Loyalty
Previous studies have defined trust as the belief that a service provider will fulfil the customer's needs and the provider avoiding any adverse actions that lead to undesired outcomes [61].Similarly, ref. [62] defined trust as the behaviour, feelings, thoughts, and emotions occurring as the customer perceives that a service provider will act in their best interest.According to the social exchange theory, trust is a prerequisite for any contract or relationship [63].Previous studies have shown that trust is an essential variable influencing bank customer loyalty [64,65].Ref. [66] found that higher levels of trust shown by a bank's customers lead to improved customer retention and loyalty for future transactions.This is because a high level of trust reduces the customer's search costs for bank services and increases the likelihood of customers recommending their bank to others [67].Therefore, trust is considered as a long-term determinant of loyalty [68][69][70], because it reduces the cost of adverse selection and moral hazard owing to asymmetric information.This is critical, especially in the service sector, as products are intangible and abstract [71].In addition, trust is critical in the case of high uncertainty [72].In the context of Islamic banks, ref. [13] defined trust to be an individual's moral obligation when performing or interacting with others in the community.Ref. [13] emphasised that trust is crucial to increase customer loyalty toward Islamic banks.As it is expected that Islamic banks should be Shariah-compliant, trust is automatically embedded into the relationship between the customer and the bank [73,74].Previous studies have indicated that customers patronise Islamic banks based on religious aspects [10,22].Consequently, customers might trust Islamic banks because of their Shariah-compliant services [13,75]; their religious duties; or fear of punishment hereafter, which could be referred to as captive customers [76].On the other hand, some studies have found that other factors, such as the quality of service provided [77], bank image, privacy, the facilities offered and proficient personnel [1], and family and friend's opinions [32] might influence a customer to deal with an Islamic bank.Therefore, the aspect of religion can have a significant impact on loyalty toward Islamic banks, especially in Muslim countries.
In addition, ref. [29] studied whether there is a different customer perception between Islamic and conventional banks and the impact of service quality on customer loyalty in Bangladesh.They found that Islamic banks scored higher than conventional banks in reputation, security, reliability, and responsiveness.Similarly, ref. [78] studied how trust and satisfaction influence customer loyalty of Islamic banks in Malaysia.In addition, they investigated the mediating effect of trust between service quality and customer loyalty.It was found that service quality and satisfaction positively influence loyalty.Moreover, trust was found to mediate the link between service quality and customer loyalty in Malaysian Islamic banks.
Researchers have indicated a strong link between customer satisfaction and trust, and that the former is an antecedent of trust [72,79].In other words, once a bank achieves a high level of customer satisfaction, this will increase the level of customer trust.In addition, it has been stated that trust mediates the link between satisfaction and loyalty [80].This is because the customer satisfaction level is combined with a high level of trust by customers who will not seek to compare the service provided by other banks.In this study, trust is studied as a mediator between satisfaction and loyalty.This has been supported by previous studies where trust was found to be a mediator between the level of satisfaction and customer loyalty [78,81].It is argued that trust impacts the value system of the client and service provider and, ultimately, the link between satisfaction and loyalty.Therefore, trust acts as an amplifier of the link between satisfaction and loyalty.
In turn, this will turn satisfied customers into loyal customers [82].Therefore, this study proposed the following hypotheses: Hypothesis 1 (H1).Customer trust positively influences customer loyalty.

Hypothesis 2 (H2).
Customer trust mediates the relationship between customer satisfaction and loyalty.

Customer Satisfaction
Customer satisfaction has been extensively studied in marketing, generally, and in the relationship marketing field, specifically [83,84].Satisfaction is generally defined as a measure of an organisation's capability to meet, or exceed, customer expectations [85].From a marketing point of view, satisfaction measures customer contentment with their prior experience with a product [86] or the continuous customer evaluation of consumption and purchase of a product or service [87].Customer satisfaction is considered one of the critical factors in explaining customer loyalty.In addition, customer satisfaction is one of the critical factors in the highly competitive banking industry [88].This is because a bank's ability to maintain a long-term relationship with its customers is profitable [84].A transaction that results in a satisfied customer creates a loyal customer through repeat purchases [89,90].High levels of bank customer satisfaction lead to better customer relationships with a bank and, therefore, higher levels of loyalty [91].This is important, given that banking services are homogenous and, therefore, gaining customer satisfaction is crucial [38], regardless of whether the bank is Islamic or otherwise.Thus, Islamic banks are competing with both conventional and other Islamic banks.This is because satisfaction is based on a comparison between the expected and actual experience [92].Previous studies have established a positive and significant link between customer satisfaction and bank loyalty [69,93].However, few studies have found a negative link between customer satisfaction and loyalty [94].
Customer satisfaction is one of the most heavily studied concepts in marketing literature.Satisfaction can be defined as the pleasure or disappointment felt from comparing the performance of a product or service with a customer's expectation [95].This means that, if the performance matches or exceeds the expectation, then satisfaction is achieved, while the opposite leads to dissatisfaction [96].Several researchers have suggested that one determinant of customer loyalty is customer satisfaction [93,97].Like any other competitive industry, banks produce products and services with negligible differences; hence, to increase their market share, they need to find ways to differentiate themselves from others [98,99].Such minor differences might be the keys to improving a bank customer's satisfaction levels and, after that, their loyalty.Satisfied customers would be less likely to leave their bank, thereby linking satisfaction and loyalty.Higher levels of loyalty to a bank would lead to higher profitability and market share [31].Although some studies have documented the positive link between customer satisfaction and customer loyalty (e.g., [100,101]), some studies have suggested that the link is not always direct [13].Thus, this study proposed the following hypothesis: Hypothesis 3 (H3).Customer satisfaction positively influences customer loyalty.

Religious Belief
Ref. [26] investigated whether Islamic banks customers' attitudes and purchase intentions were influenced by religiosity in Tunisia.They found that religiosity had three dimensions: religious beliefs, religious involvement, and fear of divine punishment.In addition, Souiden and Rani proposed direct relationships between three dimensions of religiosity and attitude, but not with purchase intention towards Islamic banks.However, there was an indirect relationship between religiosity and purchase intention through attitude.They indicated that, for Islamic banks to prevail, banks must communicate their ability to compete with conventional banks without losing or creating doubts about Islamic beliefs.Similarly, ref. [102] examined the links between religiosity, purchase attitude, and intention towards different Islamic financial products.Their finding suggested that the religiosity and purchase intention link was mediated by attitude.Ref. [103] looked at religiosity as a moderating effect between purchase intentions and attitudes toward Islamic and conventional life insurance.Their finding suggested that religiosity moderated the relationship between attitude towards life insurance and purchase intention.They concluded that religiosity could be used as a tool to segment the life insurance market.However, not all Muslims are convinced that Islamic banks are Shariah-compliant and truly follow Shariah principles [10].This means that religiosity is not as relevant in influencing the attitudes of Muslim customers [10].Thus, this study proposed the following hypothesis: Hypothesis 4 (H4).Religious obligation positively influences customer loyalty.

Attitude towards Shariah-Complaint Banking
Attitude is a construct that measures a person's interest, or disinterest, in a specific behaviour [104].It means that the more positive the attitude toward a behaviour, the higher the likelihood of engaging in that behaviour [105,106].In alignment with Islamic teaching, Islamic banks and their customers are guided by the Shariah principles, which differentiate between what is permissible and impermissible [107].These principles guide both the supply and demand behaviours of both parties.For the latter, these Shariah principles help individuals formulate their attitudes towards what is permissible and impermissible.For banks, these principles are considered as added value, as compared with conventional banks.Previous studies have indicated that one of the crucial factors in selecting Islamic banking services is the personal attitude towards Shariah-compliant banking [10,108].Therefore, this study proposed the following hypotheses: Hypothesis 5 (H5).Attitude towards Shariah-complaint banking positively influences customer loyalty.

Hypothesis 6 (H6).
Attitude towards Shariah-complaint banking mediates the relationship between religious obligation and loyalty.

Theoretical Model
This study's theoretical model was constructed based on a review of prior literature and empirical evidence.Besides, the Social Exchange Theory [63], the Theory of Planned Behaviour [109,110], and the Social Identity Theory [111] were used to conceptualise the relationships between customer satisfaction, religious orientation, trust, attitudes, and loyalty in the proposed model (see Figure 1).

Research Methodology
A survey method was adopted to collect the data for this study.The instrument for data collection comprised a measurement scale and the respondent's demographic profile.Measurement scales were borrowed from previous studies and modified slightly to reflect the present study's setting of Islamic banking.As the measurement scales were adapted from earlier studies, content validity was assured.We did the following to ensure the questionnaire is understood and the questions are unambiguous.First, the questions measuring each variable were modified to fit the Islamic context.Second, we sought the advice of two academia to comment on the face and content validity of the questionnaire.Third, we distributed the questionnaire to a pilot of 50 respondents and checked the reliability and validity of the responses.We found that Cronbach's alpha fulfils the cutoff point of 0.7 for each variable.Trust was measured using a two-item scale adapted from [10,47].A three-item scale and a seven-item scale were adopted from [81] to measure customer satisfaction and customer loyalty, respectively.Five items concerning attitude towards Islamic banks and religious obligation were derived from [112,113].All the measurement items were anchored using a five-point Likert scale.

Sampling and Data Collection
The present study employed a non-probability sampling technique, namely the purposive sampling technique, for data collection.This method was used to identify the desired target respondents.A researcher must set some criteria for the respondents (i.e., purposes) when deciding on the respondents.In this study, the researchers decided that the respondents must have had some prior experience in dealing with Islamic banking services.Therefore, respondents with accounts at Islamic banks were sought.As the sampling unit of this study consisted of respondents who were currently working and had an account with an Islamic bank in the UAE, people with accounts at conventional banks were excluded.According to Bankscope database 2020, the UAE only has 10 Islamic banks, out of a total of 41 banks.Owing to availability and accessibility, only Islamic bank branches in the three largest cities of the UAE, namely, Dubai, Abu Dhabi, and Sharjah, were targeted to collect data (these banks included Sharjah Islamic Bank, Abu Dhabi Islamic Bank, Dubai Islamic Bank, and Emirates Islamic Bank.)Moreover, regarding the population size, these three states represented 43% of the UAE population.The research assistant distributed the survey questionnaire among clients of the chosen banks in a face-to-face manner.The respondents were assured that all the collected information would be reported in an aggregate manner and used solely for academic purposes.The survey was conducted between November 2018 and February 2019.
A total of 550 questionnaires were distributed to clients of the sampled Islamic banks.Four hundred and ninety-six (496) questionnaires were returned, yielding around a 90% return rate.Of the 496 returned questionnaires, 80 were largely incomplete (i.e., more than 50% of the questions were unanswered); thus, they were not considered usable.Overall, a total of 416 responses were found to be suitable for the analysis.

Common Method Bias
We used Harman's single factor method to check whether common method bias exists in the data.We performed exploratory factor analysis for all the questions and found that all the items explained 38% of the variance, below the 40% threshold [114,115].

Research Model
The theoretical model (Figure 1) of this study consisted of both direct and indirect effects concerning customer loyalty towards Islamic banks.Equations ( 1) and ( 2) represent the mathematical expressions of the four direct effects and two indirect effects, respectively, which were examined in this study.
In these equations, LOY is customer loyalty; SAT is customer satisfaction; RO is religious obligation; TR is customer trust; ATT is attitude towards Islamic banks; β 0 − β 1 , β 0 − β 4 are the coefficients of the direct and indirect links, respectively; and u i and v i are the error terms.
This study tested the hypotheses by applying the partial least squares structural equation modelling (PLS-SEM) method, using the SmartPLS version 3.0 software package.PLS-SEM was chosen as the overall objective of this study was to predict the level of customer loyalty toward Islamic banks, and PLS-SEM can analyse a complex model containing multiple constructs [116].Ref. [116] indicated that PLS-SEM is a better technique to estimate complex models that include various constructs, indicators, and paths without having to worry about distributional assumptions.Therefore, PLS-SEM is designed to estimate and predict the causal links in the structural model [117,118].In addition, PLS-SEM estimates the coefficient of proxy measuring a latent variable that is measured by multiple items.PLS-SEM overcomes the small sample size problem.Moreover, PLS-SEM can be applied for reflective as well as formative latent variables [119].
Following standard procedures, the measurement model was assessed as the first step, followed by structural model analysis as the second step.Essentially, the first step determined the reliability and validity of the measures, followed by structural model analysis in the second step.

Respondents' Profile
Table 1 reports that the sample of this study comprised 55.5% females and 44.5% males.The results show that the participants were primarily young, as 54.1% of the total respondents were aged between 20 and 30 years, followed by the age group between 31 and 40 years old (31%), between 41 and 50 years old (11.1%), and older than 50 years (3.8%).The respondents were mostly married (52.6% of the respondents), compared with single (45%).A total of 64.7% of the respondents had attained a bachelor's degree, followed by about 13.9% who held a master's degree, and approximately 7.5% who were PhD holders.Concerning the respondents' nationalities, 64.9% of them were non-UAE citizens compared with only 35.1% who were UAE citizens.Approximately 27.9% of the participants' monthly incomes were above AED10,000, followed by 26.9% who had a monthly income between AED5001 and 7000, 23.1% had an income below AED5000, and 22.1% had a monthly income between AED7001 and 10,000.Besides, 34.4% of the respondents have work experience of less than one year, followed by those with work experience of between 1 and 5 years (29.6%),those with more than 10 years of work experience (18.3%), and those with between 6 and 10 years of work experience (17.8%).All the respondents were Muslim.

Descriptive Statistics
The results of the item descriptive statistics suggest that the mean responses range from as low as 3.15 and 3.57 (See Table 2).At the same time, the standard deviation indicates an acceptable level of variability.Lastly, the skewness and kurtosis of each item point to a normal distribution, as all the items' skewness and kurtosis are within +1 and −1.
Table 3 exhibits the results of the descriptive statistics, such as the means and standard deviations of the variables used in this study.The attitude towards Islamic banks scored the highest mean, followed by religious obligation, customer satisfaction, and customer loyalty; finally the lowest mean was calculated for customer trust.In addition, there was some variability found in the responses, as the standard deviations ranged from 0.757 to 0.915.LOY is the customer loyalty of Islamic bank, SAT is customer satisfaction, TR is customer trust, ATT is customer attitude towards Islamic bank, and RO is religious obligation towards Islamic banks.

Measurement Model Analysis
The reliability and convergent validity of the measurement model were assessed using composite reliability (CR), factor loadings, and the average variance extracted (AVE).When assessing the internal reliability of constructs, Cronbach's alpha has often been criticised, as it tends to underestimate the reliability of constructs [119].Instead, CR is a more highly regarded index when examining the reliability of latent variables [116].All the constructs had CR values greater than 0.70.Individual item reliability with loading was greater than 0.70, which demonstrated a satisfactory level of construct reliability.Owing to factor loadings below 0.70, two items were deleted from the trust, one item from religious obligation and two items from customer satisfaction (Table 4).The convergent validity was examined using the AVE.The results exhibited an acceptable level of convergent validity, as the AVEs of all the constructs were above 0.50 (Table 4).Next, the discriminant validity was examined using the ref.[120] criterion.The results in Table 5 signify that the square root of the AVE exceeded the inter-correlations between the factors, demonstrating that the discriminant validity of all the constructs was achieved.Figure 2 represents the measurement model.LOY is the customer loyalty of Islamic bank, SAT is customer satisfaction, TR is customer trust, ATT is customer attitude towards Islamic banks, and RO is religious obligation towards Islamic banks.

Structural Model Analysis
PLS-SEM techniques were applied to test the hypotheses of this study.According to standard procedures, the nonparametric bootstrapping technique, with 5000 resamples, was applied to test the hypotheses (Figure 2).The results in Table 6 signify that the RO, ATT, TR, and SAT were positive and significant in influencing customer loyalty toward Islamic banks (β RO = 0.104, p < 0.05, β ATT = 0.149, p < 0.01 β TR = 0.150, p < 0.01 β SAT = 0.525, p < 0.001).Therefore, hypotheses H1, H3, H4, and H5 were supported.These findings explained that the level of loyalty toward Islamic banks in the UAE was significantly determined by favorable assessments regarding, customer satisfaction, trust, attitude, and religious obligation.The mediating effect was also examined using 5000 bootstrapping subsamples.Table 6 shows that both indirect relationships in the theoretical model were significantly positive.More precisely, with an indirect effect of 0.099, ATT significantly mediated between RO and LOY (β RO ' = 0.099, p < 0.01), indicating that hypothesis H6 was supported.Additionally, this explains that RO had a significant positive effect on LOY, both directly and indirectly, via ATT.Therefore, religious orientation and attitude were considered as important determinants of customer loyalty.In addition, hypothesis H2 was also supported (β SAT ' = 0.108, p < 0.001), illustrating that TR significantly mediated the relationship between SAT and LOY.These results also signified that the SAT had both direct and indirect effects, via TR, on LOY.
After examining the hypotheses of this study, the predictive accuracy of the theoretical model was also assessed, using the amount of variance explained by the antecedents.The result in Figure 3 shows that 64% of customer loyalty (R 2 = 0.638) was explained by the determinants, such as SAT, RO, ATT, and TR.The magnitude of R 2 signified that the level of customer loyalty toward Islamic banks was adequately predicted by the manifest indicators of the latent variable.Figure 3 represents the results of the PLS regression.

Discussions
To attain this study's objective of examining the direct and indirect effects of SAT, RO, ATT, and TR on the LOY of Islamic banks, the theoretical model of this study was analysed using PLS-SEM methods.The results were calculated using the SmartPLS 3.0 software package.The findings in Table 6 demonstrate that all the hypotheses are supported.In the setting of Islamic banks, satisfaction (SAT), as compared with religious obligation (RO), was found to be a stronger predictor of customer loyalty (LOY).This result aligned with several previous studies, namely, refs.[86,121,122].Therefore, the management teams of Islamic banks should realise that they can no longer rely solely on religious factors to maintain customer loyalty.They must ensure high levels of customer satisfaction by providing desirable services to their clients [108].This indicates that Islamic banks have been competing actively with other Islamic banks and conventional banks to create satisfied customers.Besides, as about 85% of the respondents were highly educated, the decisionmakers of Islamic banks should strategise themselves to ensure higher levels of customer satisfaction.Therefore, it is evident that Islamic banks need to exceed customer expectations when offering their services.Moreover, they must ensure that customers are delighted with the online products and services provided by Islamic banks in the UAE.
RO was found to have a significant positive direct effect on LOY.Past studies have also found that RO is one factor that drove consumers to choose one bank above another [13,22,75].Religious obligations or duties can influence an individual to become loyal to a specific organisation.Accordingly, religious obligation, though not the most decisive factor, was one of the crucial variables that explained customer loyalty toward Islamic banks.Therefore, Islamic banks should not solely emphasise profit maximisation, but should also comply with Islamic laws regarding business dealings.Besides, one of the Shariah goals is for businesses, especially Islamic banks, to achieve social justice and welfare for society.In addition, Islamic banks should prioritise their "interest-free" motto and emphasise avoidance of other prohibited elements, such as ambiguity or fraud, when providing goods and services to their customers.
Next, customer trust was positive and significant with customer loyalty toward Islamic banks.Trust is a crucial element in the success of Islamic banks.This is true because, nowadays, most banks are moving towards offering online banking services, implying that banks must be at the forefront in protecting customer information to maintain loyalty [123].It has been reported that customer trust plays a significant role in competition among banks, as it is easy for customers to move from one bank to another at almost no cost [124,125].Therefore, banks should always keep customers' information private and confidential.
In addition, banks must be proactive in protecting customer's personal information and ensure that they have secure online banking systems.
The relationship between attitude towards an Islamic bank (ATT) and customer loyalty (LOY) was positive, significant, and more robust than RO.This finding was also in line with previous research.According to various scholars, customer loyalty depends on the attitude rooted towards products or services [126][127][128].A favourable and positive attitude toward Islamic banks is vital to facilitate loyalty.For example, the essential aspects for clients when choosing an Islamic bank are being different from conventional banks, as well as being beneficial and instilling positive values.Hence, Islamic banks should focus on developing a favourable attitude toward their target consumers.
Moving on to the mediating effect of ATT and TR, the results exhibited a weak, but significant mediating effect of ATT toward Islamic banks between RO and LOY and TR between SAT and LOY.Specifically, the direct effect between RO and LOY and the indirect effect via ATT were almost the same (i.e., β RO = 0.104 and β RO ' = 0.099).This finding signified that RO and ATT were equally crucial to the customers of Islamic banks in the UAE, whereas the direct effect between SAT and LOY was much stronger than the indirect effect via TR.Thus, assuring customer satisfaction alone would be a winning strategy for Islamic banks compared with the other determinants of customer loyalty.

Conclusions, Implications, and Limitations
The present study aimed to investigate two objectives.The first is to test the direct links of customer satisfaction, trust, attitude, and religious obligations on customer loyalty.The second objective is to test the indirect links of attitude and trust on the links between religious obligation and customer satisfaction with customer's loyalty.Using a sample of working people who patronised Islamic banking services in the UAE, the following was found: religious obligation, customer trust, attitude towards Islamic banks, and customer satisfaction significantly and positively influenced customer loyalty toward Islamic banks.It was found that the most vital variable was customer satisfaction, followed by customer trust, attitude towards Islamic banks, and lastly religious obligation.In terms of the indirect links, or the mediating effect of attitude and trust, it was found that both variables mediated the relationship between religious obligation and customer satisfaction, respectively.This result suggested that there were direct and indirect links to customer loyalty in the Islamic banks in the UAE.
This study has the following theoretical implications.First, the study investigated the mediating effect of trust and attitude on customer loyalty and found positive mediating effects.This means that, although the level of satisfaction and religious obligations can improve customer loyalty, the trust and attitude fortify this relationship further.This in fact implies that Islamic banks need to emphasize more the trust and attitude dimensions.This is because Islamic banks do not face the same issues encountered by conventional banks.Second, the results suggest that trust and attitude act as partial mediators between religious obligation and customer satisfaction towards customer loyalty.Therefore, trust and attitude strengthen the links between satisfaction and religious obligation and customer loyalty.Hence, being a satisfied customer with a high level of trust will eventually improve longterm loyalty with Islamic banks.Similarly, religious obligation is crucial in determining loyalty, but a positive attitude will lead to better loyalty.
The following managerial implications can be derived from the results.First, given that satisfaction is crucial in determining loyalty in Islamic banks, the managers and decision makers must ensure that customers receive better services than those provided by competitors.In this sense, Islamic banks should give the customers the experience that leads to repeat service use to fortify loyalty.In addition, looking at how religious obligations influence loyalty, managers need to understand that they do what they preach.In other words, the floating rumours that Islamic banks are only Islamic by name must be amended in the customer's mind.This will ease the long-term relationship with the Islamic banks.In addition, the right, positive, and healthy attitude about Islamic banks is imprinted in the customers' minds to improve loyalty.This implies that managers must provide customers with all the details about how Shariah-compliant the products and services are to strengthen the attitude towards Islamic banks.Similarly, trust can be targeted by managers to improve loyalty through establishing the belief that the bank's promises are true and deliverable.
This study was not without limitations.First, considering that the study setting was only in one economy, namely, the UAE, the findings might not be generalisable to other countries.Therefore, future studies should expand their sample to cover more than one country and then compare the results.Second, additional religious variables such as intrinsic and extrinsic religiosity and fear of punishment should be taken into consideration in determining the level of customer loyalty toward Islamic banks.Lastly, future studies may benefit if their research were carried out using longitudinal data.Each of the limitations listed here can be addressed in future research.For example, samples can be collected from multiple Muslim countries to understand whether religious obligation has a similar effect across the Muslim countries.

Table 2 .
Descriptive statistics for items.

Table 3 .
Descriptive statistics for variables.

Table 4 .
Result of CFA for the measurement model.

Table 5 .
Correlation and discriminant validity of constructs.

Table 6 .
Results of the PLS regression., ** significant at 1% and 5%, respectively.LOY is the customer loyalty of Islamic bank, SAT is customer satisfaction, TR is customer trust, ATT is customer attitude towards Islamic banks, and RO is religious obligation towards Islamic banks. *