Impacts of service quality on future intentions of loan purchase under the mediating of borrowers’ bank image and moderating Fatwa and Sharia supervisory board reputation

The purpose of this study is to test the moderating role of fatwa and Sharia supervisory board unit reputation in the relationship between service quality dimensions and borrowers’ bank image. It will also assess the mediating role of borrowers’ bank image in the relationship between service quality dimensions and future purchase intentions of loans. Questionnaire surveys are used to collect data by convenience sampling from the loan borrowers in Saudi Arabian Islamic banks. Structural equation modeling was performed to analyze 370 usable responses. The study finds that service quality dimensions have varying effects on loans future purchase intentions, the borrowers’ bank image fully and partially mediates the relationship between service quality dimensions and loans future purchase intentions (LFPI), and the fatwa and Sharia supervisory board reputation moderating the influence of service quality responsiveness on borrowers’ bank image. Bank image and fatwa and Sharia supervisory board reputation influences on future purchase intention were assessed by personal loan borrowers, and future studies could be conducted on other bank loan borrowers (corporate) according to their difference in the goal of loaning. The bank managers should be aware of the fatwa board's reputation influences on borrowers’ bank image and loan future purchase intentions. This research adds to the existing body of reputation and image management literature in the Islamic banking context by comprehending how specified service quality dimensions can improve the borrowers’ bank image and increase loan purchase intentions. The study is the first attempt to discover empirical support to the role of fatwa and Sharia supervisory board reputation in increasing the bank’s image purchase intentions at Islamic banks.


Introduction
*Today, Islamic banking services are facing intense competition from conventional banking, so they try to provide the best service quality and use the fatwa and Sharia supervisory board unit reputation to create a strong image in their loan borrowers as a powerful determinant of the market differentiating.
The relationship between service quality and banks' performance could be investigated through different measurements in various sectors. A number of studies have shown the importance of service quality for improving the corporate image and intentions of customers to share their positive experiences of the company (Ladhari, 2009;Carrillat et al., 2007;Kang and James, 2004;Vyas and Raitani, 2014).
Future purchase intentions are considered in this study as economic' behavioral intentions, which refer to the customers' behavioral reactions in the financial aspect (Anderson and Mittal, 2000;Maxham and Netemeyer, 2003). Therefore, the loan's future purchase intentions (LFPI) creates the likelihood that borrowers will purchase loans in the future, drawing on his/her financial needs. The types of loans provided by the Saudi Arabian banks are land finance, home finance, credit card, auto finance, and personal loans. When loan borrowers received high service quality offers delivered from the bank, they take a good image about service providers that in turn lead to purchase intentions.
The fatwa and Sharia supervisory board reputation (FSSBR) can be defined as a collective unanimous perception of stakeholders based on the known information about board members' personal past knowledge. The bank's fatwa and Sharia supervisory board is a management unit acquired by Saudi banks that provides the advisory opinions and issues fatwas on the customers' questions regarding each Islamic banking product and service offered by the bank needing clarifications or stating the Sharia judgment about. The board with a positive reputation sends a good signal to the borrowers that the bank is implementing Sharia principles.
Over the last decade, a body of knowledge has developed that examines why individuals and organizations may wish to create positive reputations (Hall et al., 2009;Zinko et al., 2012;Zinko, 2013). However, to date, no existing theory is there, investigating how the fatwa and Sharia supervisory board reputation established by the bank strengthens the service quality and influences the borrower's bank image growth as a moderator variable, and increases loans purchase intentions in the Islamic banking context in Saudi Arabia (SA); specifically, as the mediating effect of the borrowers' bank image on the relationship between service quality and future purchase intentions by structural equation modeling (SEM) according to loans borrowers viewpoints in the SA banking sector.
Borrowers' bank image (BBI) is an aggregate picture of the bank that reflected borrowers' attitudes and experiences with the bank. Loans Service Quality (LSQ) is excellent at providing the best service offerings that attain or surpass the borrowers' expectations and to be easily able to assess the services. From the widely known service quality models/scales, three instruments are used to test loan borrowers' assessment of service quality. They are as follows: Tangibility, safety/security, and responsiveness, which have been used as a sample for assessment. The tangibility of service-the equipment and technology, appearance of employees, appealing materials and physical facilities. Safety/security refers to feeling secure and the perceived safety of dealing in relation to banking transactions (Ennew et al., 2013). Responsiveness in the service quality literature refers to a firm's willingness to help customers and its ability to provide prompt service (Kotler, 2009). Therefore, the research questions are as follows:  How do the bank loan borrowers assess the LSQ?  To what extend does BBI influence the service quality and are willing to purchase or continue doing loans offered by the current bank?
 How to manage the bank FSSBR as a resource competitive advantage in the banks' sector?
To reach the answer to these questions, the study carried out an empirical investigation. The result of this research will add new insights in terms of theory and practice for scholars and practitioners in the mentioned areas of relationships. The present study is organized into the following parts: The underpinning theories, the literature review and hypothesis development, methodology, data analysis, and discussion.

The underpinning theories
The study draws on resource-based theory (RBT), the theory of dynamic capabilities (TDC), and theory of planned behavior (TPB) as underpinning theories to integrate and test the influences of the service quality, Sharia fatwa and supervisory board reputation and bank image on loans future purchase intentions; the literature does not wholly disclose concerns allied with these dimensions, contributing and utilizing to the triumph in banking performance yet. Barney's (1991) RBT emphasized that a company has a package of assets and abilities. In addition, its abilities refer to the skill of the firm to exploit its resource or capacity to mix and match various resources by managerial methods to produce anticipated outcomes (Parahald and Hamel, 1990). The RBT is used in this study to facilitate the assessment of the constructs as recourses that have significant influences on bank performance.
The TDC-based on resources and capabilities explain the importance of intangible assets for the competitive success of the company (Villena and Souto-Pérez, 2016). The TDC helps in explaining how to use the internal capabilities that are responsible for responding to changes in the environment (Zajac et al., 2000). This theory generates the innovation and development of more new solutions abilities at the workplace.
The TPB is regarded as one of the most influential and well-supported social psychological theories for predicting human behaviors (Ajzen, 1985). Further, the TPB can provide useful insight into the processes that translate positive attitudes or evaluations into purchase intentions and purchase behavior (Smith et al., 2007). Hence, the TPH helps this study by providing a possible theoretical mechanism links between fatwa and Sharia board reputation, such that borrowers' bank image and loans purchase intentions can be established. According to TPB, the intention is viewed as an immediate antecedent of actual behavior (Lin and Chen, 2011). Fig. 1 is describing the study's framework. The service quality dimensions are considered in different studies to be a critical measure of banks' performance (Lenka et al., 2009;Yoo and Park, 2007;Al-Hawari, 2006). Recently, Tandon et al. (2017) argued that high service quality leads to favorable repurchase intention and vice versa.
A positive relationship between corporate image and behavioral intention has been found in various industry sectors, including hotels (Christou, 2003), travel (Richard and Zhang, 2012), and a gaming company (Wu, 2014). This recognition leads to conclude that the image of a bank's service has an impact on their actual behaviors to purchase or repurchase and encourages customers to stick with the service provider.
Future purchase intentions are considered as post-purchase intentions have been frequently used as a foundation to predict consumers' future behaviors (Wierenga, 2008;Kuo and Wu, 2012). Consumers tend to purchase services or products that they believe are aligned with their own selfimage (Sirgy et al., 2000;Malär et al., 2011). Because of the importance of image in service firms, services are first required and then produced. Once a customer perceives consistency between himself or herself and a company, his or her behavioral intentions might remain high (Jeong and Jang, 2017).
A future intention to purchase or repeat loans depended on borrowers' need for loans, reinvestment returns, and loyalty behavior. A good borrowers' bank image facilitated the loan purchase intentions decision-making. Accordingly, the following propositions are formulated: H1. Service quality (H1a safety/security, H1b tangibility, and H1c responsiveness) will have a positive influence on loan purchase intentions. H2. Service quality (H2a safety/security, H2b tangibility, and H2c responsiveness) will have a positive influence on borrowers' bank image. H3: Bank image has a positive influence on loan repurchase intentions. H4. Bank image will mediate the positive influence between service quality (H4a safety/security, H4b tangibility, and H4c responsiveness) and loan repurchase intentions.

The moderating influence of FSSBR between service quality and borrowers' bank image
It is worth mentioning that the bank's fatwa and Sharia supervisory board is an independent management unit. It consists of three or more respectful jurists, world-renowned and prominent, with extensive knowledge and the widest experience in the field of Sharia; eminent and distinguished Sharia scholars who are eligible for fatwas in the field of financial institutions transactions and Islamic banking in particular. Their fatwas are issued by agreed collective jurisprudence, and they provide annually published reports confirming that the Islamic precepts, principles, and rules are being encompassed correctly. The matter of the thing to the borrower is substantiating that there are not any kinds of interest ("Rabba") and unlawful transactions in the bank's loans. This stage of uncertainty motivates the borrowers to place the fatwa board reputation that is created by such a bank at the center of loan future purchase intentions assessments when applying for loans from the target bank. Hence, the fatwa board represents shield protection in front of banks transactions offers to customers by reassuring better transparency and confidence on a bank's financial transactions that raise the awareness about Islamic concepts of banking services and finding or suggesting the possible Sharia alternatives solutions for management finance problems inconsistent with provisions of Islamic Sharia. Thus, when the reputation of the fatwa board in a bank is good, borrowers may be convinced and trust in the loan services and may be tending to do business or repurchase.
Reputation is linked to impression management (Sosik and Jung, 2003). A fatwa member's reputations' salient characteristics are generally built on a person's renowned knowledge/scholarship of the verdict of Holy Quran, Sunnah Hadiths, and "Ijma" related to financial transactions, in particular, requiring the competence to look at Sharia legality, honesty (deeds and words), the trustworthiness of the person, godliness, and it depends on how they are respected by Muslims and ascetic people.
Service quality was found to be highly correlated with the image (Sajtos et al., 2015). The service quality is found in Islamic countries, which impacted customer satisfaction in the banking sector in Saudi Arabia (Sanjuq, 2014) and in the Turkish banking industry (Yavas et al., 1997).
A good corporate image as an intangible asset leads to positive customer interactions and differentiation between firms. Park and Park (2017) showed that the airline service quality positively influenced the brand image and did not directly influence reuse intention. Hence, besides the service quality attributes, Muslim customers consider fatwa and Sharia supervisory board reputation highly when they assess the quality of the banks' services, specifically their loans services. Thus, the borrowers' bank image is strengthened by the bank's service quality and fatwas board reputation as these factors help shape valuable corporate image in the customers' mind and optimize their decisions when applying for loans.
In spite of the fact that previous studies have proposed that there is a positive correlation between service quality and corporate image, this relation's impact may be contingent on other factors, such as fatwa board reputation. Therefore, this study hypothesizes such a connection below: H5. Sharia fatwa and supervisory board reputation will moderate the positive relationship between service quality (H5a safety/security, H5b tangibility, and H5c responsiveness) and bank image, such that the relationship will be stronger when Sharia fatwa and supervisory board reputation is higher.

Questionnaire design and measures
The study adopted the questionnaire survey method, one of the most common methods used for social research, which involves a series of processes to collect the data (Neuman, 2006). All of the questionnaire items are adopted from previous studies, and the pre-test is performed by seven bank consultants. The questionnaire comprises five parts questions. The first part is the participant demographic characteristics, which utilized the ordinal and nominal scale; the second part is the loans service quality tangibility, safety/security, and responsiveness; the third part is the borrowers' perceptions of Sharia fatwa and supervisory board reputation; the fourth part is the borrowers' bank image, and the last part is the loans future purchase intentions.
All items were conducted on a Likert scale of (1) strongly disagree to (5) strongly agree. Basically, descriptive statistics, unengaged responses check, and structural equation modeling (SEM) can be used to estimate the proposed research model because they are more powerful methods that can deal with a large number of endogenous and exogenous variables. Moreover, the latent and observed variables as well as multicollinearity, following Anderson and Gerbing's (1988) suggestions, are used. In addition, principal factor analysis (PFA), exploratory factor analysis (EFA), Kaiser Meyer Olkin (KMO) of sampling adequacy, Bartlett's Sphericity test, common bias method, and confirmatory factor analysis (CFA) was performed. The CFA will be conducted to examine whether a theoretical relationship between items and their hypothesized factors were supported by the data and to test construction reliability and validity. Further, to examine the internal consistency and reliability of each construction, reliability tests, including Cronbach's α statistics and Composite Reliability (CR), was performed in this study. AMOS 22.0 statistical packages were used to analyze the data collected.

Population and sampling techniques
The target population for the study is the banks' borrowers in the Saudi Arabian banking sector. Personal bank borrowers will be chosen to represent the population of the study. Consequently, the target population scatters in this study will be used for the convenience sampling method through voluntary participation. Specifically, for the sampling techniques, convenience sampling was performed for the three months' target, which relates to the bank borrowers who visited the bank during this period. In this study, the sample size was 450, and the research model contained 6 constructs. It is necessary to say that the sample size is suitable according to Hair et al. (2010), who noted that an SEM model with five or fewer constructs, and each construct with three or more items, can be adequately estimated with samples as small as 100-150. A sample size of 150 is also usually sufficient for a convergent and proper solution (Anderson and Gerbing, 1988). Therefore, it was fit to conduct SEM to investigate the research hypotheses.
To deal with the issue of loan future purchase intentions in the context of the banking industry, the bank's personal borrowers in the Riyadh Area in 2018 were chosen as the study sample. The questionnaire was distributed in the Arabic language, and after the data was collected, it was translated into English. The process of translation was completed by the two experts who agreed on the consistency of the translation to avoid a shortage of translation and meanings. In addition, the trial study test was employed by the consultants to ensure the reliability and validity of the items' measurements and to minimize the loss of meaning.

Measurements
Criterion procedures for new measurement development were performed wherever it was required, and items of measures were adapted from related literature. The service quality scale comprises three primary variables: tangibility (4item), safety/security (4-item), and responsiveness (4-item), which are conducted from models (Parasuraman et al., 1988;Chen et al., 2010). The service quality is used as an independent variable.
The fatwa and Sharia supervisory board reputation items were developed and validated based on Hochwarter et al. (2007) and Ponzi et al. (2011) studies. This variable was measured using seven items and conducted as a moderating variable.
The borrowers' bank image was measured with five items modified from Dowling (1986) and Bloemer et al. (1998) and used as a mediating variable.
Following the studies by Lam and Hsu (2006), Jahn and Kunz (2012), Bian and Forsythe (2012), and Kuo and Wu (2012), loans' future purchase intentions were modified and measured using five items as dependent variables. The items of measurement are presented in Table 2.

Data collection
The data collection took place in the SA banking industry by means of a self-administrated survey questionnaire. The study began on March 01, 2018, and ended on May 31, 2018. The data collection covered only the banks' personal borrowers, not the banks' other customers, because the data from them were expected to be a reasonable and acceptable matching for the study goals, and the bank provides two types of loan personal and corporate. Each participant was chosen randomly from the borrowers who visited the bank's branch at that period and accepted to fill out the questionnaire voluntarily based on their banking experiences. All the visited banks owned fatwa and sharia board, namely National Commercial Bank, Alrajhi Bank, Riyadh Bank, Albilad Bank, and Alinma bank. Table 1 shows the basic demographic characteristics of the participant. Out of a total of 450 questionnaires distributed by the banks' employees, a total of 370 valid responses are obtained, yielding an overall response rate of 82%. This high response rate due to the questionnaires ware given one by one, and participants in this country still prefer face-to-face interactions. Participants between 41 and 50 years of age represent 29 percent of the sample; those between 31 and 40 years represent 26 percent; those who are 30 and less represent 24 percent, and those who are 51 and more represent 21 percent. About 77 percent of the participants are male, followed by 23 percent female. Of the sample, 50 percent of the participants are graduates; followed by 27 percent with postgraduate education; followed by 23 percent are others.
The majority of the participants-55 percenthave banking experiences between 3 and 6 years, followed by 25 percent who have 7 years' experience and more, and at least 20 percent who have less than 3 years' experiences. 53 percent of the participants have a yearly salary as per Saudi Arabia Riyal (SR) of 86000 and more, followed by 28 percent with 73000-85000 SR, then 12 percent with 60000-72000 SR, and the least 7 percent with less than 60000 SR.

Data screening and exploratory factor analysis
To check unengaged responses, (STDEV.P) is calculated, and outliers are checked. Therefore, no outliers were found on the dataset, and everything in the dataset was logical. Regarding normal distribution, skewness and kurtosis observed fairly normal distribution for our indicators of latent factor and for all other variables (categories variables) in terms of skewness. However, we observed mild kurtosis for our variable. These kurtosis values ranged from benign to 3.
While this does violate strict rules of normality, it is within the more relaxed rules suggested by Sposito et al. (1983), who recommended 3.3 as the upper threshold for normality.
Exploratory factor analysis was used on random questionnaire items using Varimax rotated to assess whether the chosen models could display the expected number of constructs (six). The final test outcomes were shown in Table 2. Kaiser Meyer Olkin (KMO) of Sampling Adequacy and Bartlett's sphericity test were conducted for validity (KMO=0.788, w2=2908.434, p=0.000). These values indicate that our sampling adequacy was acceptable compared with 50% of the benchmark (Leech et al., 2005), and significant Bartlett's test as recommended by Williams et al. (2012) when the pvalue is less than 0, 05, which implies that the study sample data are appropriate for factor analysis. The analysis applied a factor weight of .50 as the minimum cut-off (Hinkin, 1995). Any item exhibiting cross-loading over .40 was dropped (Hair et al., 2010). As a result, eleven non-significant items were dropped one by one when they repeated the process after following Anderson and Gerbing's (1988) guidelines. The analyses resulted in six factors. The six factors then explored from the exploratory factor analysis (EFA) as shown in Table 2.
The Cronbach's alpha coefficient values of all the extracted factors are shown in Table 2. Cronbach's alpha for the four-item responsiveness variable was 0.826, for the three-item fatwa and supervisory board reputation variable was 0.841, for the threeitem loans, future purchase intentions were 0.789, for the three-item bank image variable was 0.776, for the three-item safety/security variable was 0.712 and for the two-item tangibility variable was 0.766, which exceed the suggested threshold of 0.7 (Nunnally, 1978) and the average variance extracted (AVE) for each construct above the desirable level of 0.50 (Bagozzi and Yi, 1991), which shows strong internal consistency and reliability. Table 2 also showed that factor 1 is manifested by dimensions that capture the responsiveness, while factor 2 captures fatwa and Sharia supervisory board reputation measure items, and factor 3 captures loans future purchase intentions measure items. Factor 4 captures safety/security measure items, and factor 5 captures tangibility measure items.

Confirmatory factor analysis
To further confirm the factor structure of the proposed research model, a repetition of confirmatory factor analyses is required. To achieve an acceptable model fit, as shown in Table 3.

Common method bias
Common method bias (CMB) tested by variance explained values for each factor. Among the total variance explained that equal (62.35) factor one as the biggest factor counted only 15.16%, which does not show any risk of CMB the study's framework data (Podsakoff et al., 2003). Furthermore, the variance inflation factor VIF values that ranged between 1.134 to 1.702 sign no multicollinearity exists in this study.  (Hair et al., 2010;Iacobucci and Churchill, 2010). For more accurate reliability, the composite reliabilities CR are calculated. Composite reliability provides a measure of the internal consistency and homogeneity of the items comprising a scale (Churchill 1979). Table 4 shows the CR for safety/security and loans future purchase intentions are less than 0.70, as recommended by Nunnally (1978), but still acceptable according to Tseng et al. (2006), who suggested that composite reliability should be greater than 0.6. Other constructs exceed 0.7, indicating good convergent validity. Therefore, we can deduce that the internal consistency and reliability were acceptable.

Validity, reliability, means, and standard deviations
The AVE displays the mean-variance extracted for the model items loading on a construct. All the AVE values exceed 0.5, confirming the adequate discriminant validity of the constructs (Hair et al., 2010) with the exception of safety/security. Also, the convergent validity of the constructs was supported. In addition, the maximum shared variance MSV that reflects the highest correlation coefficient between latent constructs and average shared variance ASV that describes the mean of squared correlation is used to check discriminant validity. Thus, the values of MSV and values of AVE are less than AVE values, which shows the accepted discriminant validity. Based on these outcomes, the proposed study measurement model was found to be a valid and reliable model to be conducted in the Saudi Arabian bank sector in terms of the observed items' discriminant validity, convergent validity, and internal consistency. Table 4 also shows the Means, standard deviations and the correlations of the study constructs. It indicated that safety/security (mean=3.86, Std= 0.68), tangibility (mean=3.66, Std= 0.77), responsiveness (mean= 3.52, Std= 0.86), fatwa and Sharia board reputation (mean= 3.84, Std=0.67), borrower's bank image (mean= 3.80, Std= 0.75), loan future purchase (mean= 3.36, Std= 0.73), that above the assumed test mean which equals (3.00). Further, under the (p-value<0.01 and p-value<0.05) overall constructs showed significant simple correlation between each other that encourage us to take more comprehensive test for hypothesized relationships in the next section. Fig. 3 shows full model hypotheses testing. Table 5 shows the regression weights for direct effect, and Fig. 3 shows the statistical outcomes of the structural path model. This study established a structural equation model to test the direct effect of research hypotheses with a maximum likelihood. Table 5 notes good fit indices for overall fits of the estimated model of 2 = 142.483 (p= 0.000), df=54, 2 /df (NFI)= 2.639, AGFI= 0.931, CFI= 0.944, SRMR= 0.034, RMSEA= 0.065, above the recommended threshold. Table 5 depicts the results of the hypotheses testing. With the exception of the paths from responsiveness to LFPI (H1c) and tangibility to LFPI (H1b), all hypothesized relationships were significant and supported. As summarized in Table 5, safety/security was found to have a positive influence on LFPI (β= 0.253, p<0.05), which supports H1a.

Direct effects
The evidence for that β= the probability of getting a critical ratio (CR) as large as 2.8 in absolute value is 0.005. In other words, the regression weight for safety/security in the prediction of BBI is significantly different from zero at the 0.01 level. Also, safety/security was found to have a positive influence on BBI (β= 0.186, p<0.01), H2a was therefore supported in this study. The evidence for that β= the probability of getting a critical ratio as large as 2.479 in absolute value is 0.013. In other words, the regression weight for safety/security in the prediction of LFPI is significantly different from zero at the 0.05 level.
Tangibility was found to have a positive impact on BBI (β= 0.184, p<0.01), which lends support for H2b. The evidence for that β= the probability of getting a critical ratio as large as 1.618 in absolute value is 0.106. In other words, the regression weight for responsiveness in the prediction of LFPI is not significantly different from zero at the 0.05 level. On the other hand, tangibility does not have a positive impact on LFPI (H1b: β=−.011, p>0.10), which not supports H1b. The evidence for that β=the probability of getting a critical ratio as large as 3.051 in absolute value is 0.002. In other words, the regression weight for tangibility in the prediction of BBI is significantly different from zero at the 0.01 level.

Fig. 3: Full model hypotheses testing
Responsiveness is positively related to BBI (β= 0.356, p<0.001), which provides support for H2c. The evidence for that β=the probability of getting a critical ratio as large as 0.19 in absolute value is 0.849. In other words, the regression weight for tangibility in the prediction of purchase intention is not significantly different from zero at the 0.05 level. However, there was a lack of support for significant positive relationships between responsiveness and LFPI (H1c: β=−0.112, p>0.10). Hence, H1c is not corroborated. The evidence for that β=the probability of getting a critical ratio as large as 5.758 in absolute value is less than 0.001. In other words, the regression weight for responsiveness in the prediction of BBI is significantly different from zero at the 0.001 level. Finally, BBI was found to have a positive impact on LFPI (β= 0.202, p<0.05), so H3 was supported. The evidence for that β=the probability of getting a critical ratio as large as 2.008 in absolute value is .045. In other words, the regression weight for bank image in the prediction of LFPI is significantly different from zero at the 0.05 level. The underlined values are critical ratios exceeding 1.96 at the 0.05 level of significance; S.E. is an estimate of the standard error of the covariance; CR is the critical ratio obtained by dividing the covariance estimate by its standard error; **p<0.05; ***p <0.01

The mediating effect of bank image
The role of the BBI as a mediator was assessed in three hypotheses (H4a, H4b, and H4c). Preacher and Hayes (2004) and Zhao et al. (2010) suggested that the full mediation occurs if (1) An insignificant relationship exists between the independent variables (safety/security, tangibility, and responsiveness) and dependent variable (LFPI) and (2) A significant relationship exists in the indirect path through the mediator (BBI). Partial mediation occurs if (1) A significant relationship exists between the independent variables (safety/security, tangibility, and responsiveness) and dependent variable (LFPI) and (2) A significant relationship exists in the indirect path through the mediator (BBI). Thus, the mediation that can occur even though the criteria of Baron and Kenny (1986) were not met.
For the test of H4a, H4b, and H4c, in the mediation analysis, a linear regression analysis procedure described by Preacher and Hayes (2004) and Zhao et al. (2010) was performed.
Hypothesis H4a was partially supported based on the SEM analysis as illustrated in Fig. 3. Safety/security had a statistically significant direct impact on LFPI and a significant indirect impact on LFPI when BBI entered between them relationship. Beta estimates 0.038, P=0.033, the indirect (mediated) effect of safety/security on LFPI is 0.038. That is, due to the indirect (mediated) effect of safety/security on LFPI, when safety/security goes up by 1, LFPI goes up by 0.038. This is in addition to any direct (unmediated) effect that safety/security may have on LFPI. Generally, according to the findings presented in Table 6, it can be concluded that there is a partial mediating BBI between LSQ dimension safety/security and the LFPI relationship.
On the other hand, hypothesis H4b was fully supported based on SEM analysis path coefficient decomposition, as shown in Fig. 3. SEM analysis showed that tangibility does not directly impact LFPI, while it has statistically significant indirect impacts only through other predictor variables borrowers bank image P value=0.04, as it can be seen in Table 6. In Table 6, Beta is the indirect (mediated) effect of tangibility on LFPI is 0.037. That is, due to the indirect (mediated) effect of tangibility on LFPI, when tangibility goes up by 1, LFPI goes up by 0.037. This finding indicates that LSQ tangibility alone does not explain how significantly LFPI is existed, showing that it is necessary to enter another variable to completely explain how a good BBI leads to positive LFPI.
Further from the observation of the SEM model depicted in Fig. 3, it implies responsiveness has no direct effect on LFPI but has an indirect negative effect on LFPI with BBI as a significant full mediating variable because Table 6 shows P value=0.045. More noticeably, responsiveness plays a fully mediating role in predicting the impact of responsiveness on LFPI. Also, Beta estimates 0.072 the indirect (mediated) effect of responsiveness on LFPI is 0.072. That is, due to the indirect (mediated) effect of responsiveness on LFPI, when responsiveness goes up by 1, LFPI goes up by 0.072. This result concludes that BBI has a fully mediating role in explaining the relationship between responsiveness and LFPI, which lends support for H4c. Therefore, these two full mediating variables help explain the idea that LFPI is driven by LSQ tangibility and responsiveness when creating a good bank image that is perceived by the bank's customers.

The moderating effect of fatwa and Sharia supervisory board reputation
To test the moderating effects of FSSBR on BBI (H5a, H5b, H5c), the study adopted approaches by Ping (1995), Baron and Kenny (1986), and Sharma et al. (1981) to moderate structural equation modeling MSEM, using the three steps described by Cortina et al. (2001). This is because moderated regression analysis limits the researcher to investigating one dependent variable at a time and can lead to both the loss of statistical power as the reliability decreases (Aiken et al., 1991), biased coefficient estimates (Ping, 1995), and all the variables were latent. These three steps are shown in Fig. 4. Aside from the χ2 statistic= 261.016, the results show a good model fit (X 2 /df = 2.807; df= 93; NFI= 0.90; CFI= 0.924; RMSEA= 0.069; SRMR= .038). Fig. 4 shows the results of the analysis.
H5a proposed that FSSBR will moderate the positive relationship between LSQ safety/security and BBI, such that the relationship will be stronger when FSSBR is higher.
The findings showed in Fig. 4 display that the interaction coefficient for FSSBR and LSQ safety/security was not significant (β= -0.06, p>0.10). Thus, hypothesis H5a was not supported.
H5b predicted FSSBR would moderate the positive relationship between LSQ tangibility and BBI such that the relationship will be stronger when FSSBR is higher. The results from Fig. 4 show that the interaction coefficient for FSSBR and service quality tangibility is not significant (β=-0.08, p>0.10). Thus, hypothesis H5b was not supported.
Moreover, H5c proposed that FSSBR will moderate the positive relationship between LSQ responsiveness and BBI, such that the relationship will be stronger when FSSBR is higher. Fig. 4 shows that the interaction coefficient for FSSBR and LSQ responsiveness is significant (β=0.36, p<0.01). This supports Hypothesis H5c.
On the other hand, if there is a significant interaction, we want to make an interaction plot to see the degree of occurrence at different levels of this variable or in order to illustrate the nature of the moderator effect. Also, if this significant, then at different FSSBR, LSQ responsiveness will not be the same and vice versa. For further support to the moderated relationship, one medium above and below the moderator simple slopes test is used. Fig.  5 plots the interaction, which shows that the relationship between LSQ responsiveness and BBI is stronger when FSSBR is high (B=0.199, t=5.16, p<0.001). Fig. 5 plots the interaction and shows that the positive relationship between LSQ responsiveness and BBI strengthens when FSSBR is high. The simple slopes test provides further support. The relationship is differed from 0 at low levels of FSSBR compared with from 0 at high levels. This result indicates that banks that possess high FSSBR show a positive impact of LSQ responsiveness on BBI at a high range of LSQ responsiveness. Table 7 shows the results of hierarchical regression analysis.

Discussion
The findings of this study showed that the main study's questions were answered. As the purpose of this research was to disclose how to relate loans service quality LSQ (safety/security, tangibility, and responsiveness) with LFPI, how borrowers' bank image BBI mediates the relationships between LSQ and LFPI, and how fatwa and Sharia supervisory board reputation FSSBR moderates the relationships between LSQ and BBI. It was deduced that all these five relationships could influence LFPI directly or indirectly in the following ways: First, the findings indicate that among the study, LSQ dimensions safety/security alone has a direct significant positive influence on LFPI and becomes the first priority to the borrowers compared to tangibility and responsiveness. Additionally, the bank needs to provide safe and secure transactions when dealing with loan borrowers. This finding apparently shows that borrowers did not concern themselves with items of LSQ tangibility during the loan purchasing process. In this sense, the role of tangibility might change as the BBI arise and assess as a mediator between tangibility and LFPI.
Second, the findings provided positively significant and solid support for the three hypotheses of LSQ (safety/security, tangibility, and responsiveness) with BBI. These direct positive influences may be justified by a critical role of LSQ dimensions in perceiving the brand or corporate image since the customers are assessed service provider products or services offers bases greatly on these dimensions.
Third, the findings demonstrated that LSQ safety/security has both direct and indirect influences on LFPI, and it has a direct influence on BBI according to the structural equation modeling. While loan borrowers may perceive that they are impacted by LSQ (tangibility and responsiveness), these variables do not, in fact, appear to determine their loaning behavior. Another possible explanation is that loan borrowers might not consider LSQ tangibility and responsiveness when deciding to purchase new loans or may do not play a critical role in repurchase intentions since some variables (e.g.. , fatwas' opinions and corporate image, etc.) need to be considered. Further, a possible explanation for these findings is that our participants might have experienced more about these LSQ from the previous transactions.
Fourth, the findings showed the mediating role of BBI influencing LSQ safety/security, tangibility, and responsiveness effect LFPI. Moreover, BBI partially mediated the relationship between the LSQ safety/security LFPI. Because safety/security has direct and indirect relations with BBI and LFPI. The major reason for that is that loan borrowers think about both LSQ (safety/security) and bank image before taking the loan decisions.
Furthermore, BBI acts as a complete mediator in the relationship between LSQ tangibility and LFPI. These findings highlight the significant role of BBI in influencing the future using the behavior of borrowers. Also, BBI works as a complete mediator in the relationship between LSQ responsiveness and LFPI. These findings proved the importance of BBI in influencing the future purchasing behavior of borrowers or encouraging them to repeat the loan.
Finally, the findings support only the suggested moderating influence of FSSBR on the relationship between LSQ responsiveness and BBI. This means that FSSBR positively moderates the influence of LSQ responsiveness on BBI. The result showed that as FSSBR levels raised, the rising influence of responsiveness strengthened the influence of BBI in submitting LFPI. On the other hand, the FSSBR did not exert a moderating influence on the relationships between LSQ safety/security and tangibility and BBI. These findings confirmed that the influence of FSSBR might not work even more efficiently and effectively when LSQ safety/security and tangibility were assessed by loan borrowers.

Fig. 5: Interactions between LSQ responsiveness and fatwa
and Sharia supervisory board reputation on borrowers

Managerial implications
The study has important implications for Islamic bank managers. Accordingly, the Islamic bank managers' should increase the level of bank's service quality and image for loan performance growth with specific attention on the fatwa and Sharia supervisory board member's reputation or influence where they are receiving required fatwa information and perceiving the image. Appropriate Islamic loan products design, including transparency fatwas for the best understanding of loan products and flexible borrowing for loan methods and relevant terms, encourage or keep the current customers and attracting potential ones to loan from these banks. Further, the fatwa and Sharia supervisory board can facilitate banks' top managers to produce more customized Islamic services and increase the ability to reduce the cost of attracting and maintaining customers that resulted in profit growth. Also, this study finding can help bank managers boost the loaning service quality strategy by borrowers' bank image and fatwa and Sharia supervisory reputation, which leads to differentiate the loans services offered by the banks. Saudi Arabian Monetary Authority central bank of Saudi Arabian can have performed the findings of this study by setting some policies in the Islamic banks to improve the level of service quality and strengthen the borrower's trust in Sharia and Fatwa board reputation.

Social implications
This study developed and validated fatwa and Sharia supervisory board reputation that has an important competitive edge for Islamic banks and other Islamic financial institutions. Moreover, it provides an empirical model for understanding the relationship between loan service quality, borrowers' bank image, and fatwa and Sharia supervisory board reputation and loans future purchase intentions in the Islamic bank sector. Further, it shows how to conduct them as intangible resources or performance progressing capabilities to ingrain trust, which might end in increasing the loans' future purchase intentions and attractiveness of new Muslims' borrowers. In society, the Sharia and Fatwa board members play an important role in convincing the individuals and changing the behaviors of the borrowers to apply for a loan.

Limitations and future research
As the study is single exploratory research, it is not free of potential limitations, which may catalyze further future research into this topic. The first limitation is that the study focused on three service quality constructs, which are not exhaustive. Future studies could widen test how other service quality constructs affection and explore how can leverage the relationships between banks' image and loan borrowers purchase intentions. Also, it would be interesting to examine which constructs of service quality are most influential in the direct and indirect relationship between banks' image and future intentions.
A second limitation of the study involves a relatively small sample and examines one sector and single service industry in one country. To enable the generalizability of the study findings, more studies are needed across various industries and geographic regions. Also, consumer behavior and tastes are varying in a different context. Large sample size can provide a clear picture of interesting observations.
A third limitation pertains to the measure of fatwa and Sharia supervisory reputation. The study develops a valid and fairly reliable reputation measure for fatwa Sharia and the supervisory board. However, others can consider this endeavor as a first attempt to develop and validate more measures.
Board Reputation strengthens the positive relationship between Responsiveness and Banking_Image. Lastly, although the study till date in this field has tended to concentrate largely on fatwa and Sharia supervisory board reputation, loans service quality, bank image, and loans future purchase intentions, specifically, the study exploring the moderating role of fatwa and Sharia supervisory board reputation in loans service quality-bank image-loans future intentions framework, further researchers are invited to test the role of other pertinent moderators (e.g., borrower income and age) in this relationship or controlling them. Additionally, as fatwa and Sharia supervisory board reputation are relatively new in the banking management literature at present, much can be explored by testing other probable predictors of fatwa and Sharia supervisory boards.

Conclusion
The study is a quantitative study to investigate the influence of loan service quality tangibility, safety/security, responsiveness, bank image, and fatwa and Sharia supervisory board reputation on loans future purchase intentions among banks' borrower's perceptions in Saudi Arabia's banking sector. The empirical findings supported the most assumed research hypotheses in a positive and significant way except H1b, H1c, H5a, and H5b. This means that banks need to select besides the service quality, the best fatwa, and Sharia supervisory board members, who are differentiated and create a better bank image for customers and work out attractive retention strategies for competitive advantages since loaners reuse the banks' loans services, based on fatwa and Sharia supervisory member's reputation and bank images, which are perceived and joined with service quality. The present work represents the first attempt to analyze the fatwa and Sharia supervisory board reputation as a source of competitive advantages, increased the understanding of this phenomenon, and how it can help direct or indirect loan providers to be more customer-focused. It further advances the growing body of knowledge in these areas.