Determinants of Customer – Perceived Service Quality in the Albanian Banking Industry

The purpose of this paper is to identify the determinants of customer service quality perceptions and to develop an instrument to measure the level of service quality in Albanian banking industry. For this purpose we use the BSQ instrument and evaluate whether it can be applied in Albanian banks context. The BSQ items are evaluated and new item are added after an extensive and in depth analysis of the literature review published on bank service quality measurements. As a result of this analysis, a modified version of BSQ was created as a measurement scale of bank service quality in Albania. Data were collected through customer surveys conducted in two major cities in Albania, Tirana (the capital city) and Korca. A structured questionnaire was used in face to face interview with 352 retail bank customers. The questionnaire was formulated using BSQ items and items find from the literature review. The collected data were analyzed through tools such as factor analysis and one way ANOVA. This analysis helped to provide the validity and reliability of the modified instrument. The proposed scale comprises 28 items which spans four dimensions: responsiveness and informing, reliability and security, commodities, effective access. The research findings confirm that the BSQ instrument needs to be modified to the cultural context of Albania. Given the limited research done in measuring bank service quality in Albania this study is a significant contribution to the literature. The instrument developed in this study is important to bank managers for evaluating, controlling and improving service quality.


Introduction
In the last two decades is reported a large number of research on service quality.There are several reasons for the growing interest in studying service quality.High service quality results in retain existing customer and acquire new ones, reduce costs, enhance corporate image, generate positive world of mouth recommendation and improve profitability (Berry, 1989;Cronin, 2000;Cronroos,1984;Hurley & Estelami,1998;Keaveny,1995;Boulding,1993;Shemwell, 1998;Kim, 2009;Kang & James, 2004;Reicheld & Sasser,1990;Rust and Zahorik, 1993).Also, high service quality create a competitive edge for firms and differentiate them in the market.In order to use service quality as a mean in achieving a competitive edge first is necessary to correctly identify the determinants of service quality.In this regard, Parasuraman, Zeithaml and Berry (1985) pointed out that customer rely on a set of criteria or dimension to evaluate and determine the service quality of any business.

Due to cultural and environmental factors consumers of services in different countries have different perception of what service quality is.
Understanding what service quality is, identify the factors and dimensions that form the construct, conceptualize and operationalize models to measure it are issues extensively researched.Although the extant literature has done a credible job in advancing the understanding of the service quality construct, the vast majority of this research has been based in western and American markets (Yaprak, 2003).There is no evidence of studying service quality measurements issues in Albania.
Therefore, the purpose of this paper is to build a service quality measurement instrument for banks in Albania.Consequently, examining its item and factorial structure, asses its reliability and validity and deepen the understanding of service quality determinants.For this purpose we use a modified BSQ instrument.Thus, the survey questionnaire used in this study included the scale items from BSQ model as well as additional items from reviewing the bank service quality literature.In order to determine the relevance of existing instrument to another population the data were analysed using similar techniques used in the BSQ study.
The study's finding will be to great importance to both academic and managers.Given the limited research done in measuring bank service quality in Albania this study is a significant contribution to the literature.The instrument developed in this study is important to bank managers for evaluating, controlling and improving service quality.
The research finding confirms that the BSQ instrument needs to be modified to the cultural context of Albania.The proposed scale for measuring service quality in Albanian banking system comprises 28 items which spans four dimensions: responsiveness and informing, reliability and security, commodities, effective access.
The paper is structured as follows: first, we give a background of the Albanian banking industry, present a literature review of service quality definition and measurements models in general and in banking industry in particular; second we explain the research methodology used in this study; third, results from data analysis are presented and fourth, a discussion of results is provided and insights for service quality of Albanian banking industry are highlighted.

Albanian banking industry
Services continue to represent the main share in Albanian economy for year 2013, by 44.6 % of GDP (Albanian Statistical Institution).The financial intermediation sector is perhaps the most important sector of the service industry.It contributes around 2.2% of total GDP in Albania (Albanian Statistical Institution, 2013).These statistics provide a clear picture of the importance of services for Albania economy.Since services are a significant part of the economy it is important to examine the sector in more detail and generate methods and strategies to help its growth.
Albanian banking system is a universal banking system and represent the dominant sector within financial system, managing 94% of the total financial asset (Financial System Stability Assessment -IMF, 2014).Albania has 135 savings and loan associations but these only account for around 0.5% of the financial system (Partners for Financial Stability, 2011).
In Albania develop their activity 16 private commercial banks which are mainly foreign owned (13 out of 16 are foreignowned).
During these two decades, Albania has made important changes towards the creation of a modern banking system.Banks have change their ways of conducting business relying more on technology in service delivery.A lot of technological improvements are made by banks in Albania.This can be proved by the increased number of POS and ATM in the last ten years (from155 POS in 2004 to 6500 in 2014 and from 93 ATM in 2004 to 820 in 2014), the increased number of banks that offer mobile banking and internet banking since 2005 when this services where introduced (12 banks out of 16 offer internet banking, and 7 offer mobile banking) (Albanian Association of Banks, 2014).Even that the necessary technological infrastructure exists in Albania, the usage of electronic payment system is significantly lower compared with other countries in the region (Beqiri, 2014).
Banks competition is intense and has been strengthened with the introduction of foreign-owned banks as they introduced a diversity of products and services and a high-level technology (Ceca, 2008).Intensification of the competition also came as a result of deregulation, competition from other financial institution and free capital movement.Increased competition imposes banks to choose superior strategy and tactics in order to succeed.Until recently, Albanian banks tried to maintain their competitive standing via aggressive promotion and extensive branch networks (Lleshanaku.A, 2008).According to Lleshanaku (2008) the bank managers interviewed in Albania firmly believe that in order to succeed in the future emphasis must be put on delivering excellent service to customers.

Service quality
The literature on service quality is divided in two concerns: the definition of service quality and the measurement of service quality.
Definition of service quality: Service quality is an overall judgment of a service provider and the outcome of the comparison between customers' perceptions and their expectations (Zeithaml, 1988).Service quality can be defined as meeting the needs and expectations of the customer (Smith, 1988).Although unanimity in the definition of service quality remains elusive the general consensus in the marketing literature is that service quality must be considered from the customer's perspective (Cardozo, 1965;Fiegenbaum, 1982;Schembri & Sandberg, 2002).The arguments that support this point: 1) Customers are the ones who ultimately evaluate the level of service delivered by a service provider (Gronroos, 1984; Zeithaml, 1990;Reeves, 1994).2)Customer's perception of the subjective factors of service quality can better capture what customer want, in comparison with standard set by service providers that may, or may not be accurate (Takeuchi, 1983;Reeves, 1994).3) Unlike goods quality, service quality has no objective measures, so firms must rely on customer's perceptions (Karatepe, 2005).A firm in order to compete successfully must have an understanding of customer perception of the quality and the way service quality is influenced (Gronroos, 1984).
Measurement of service quality: Consensus as to a universal measure of service quality has not been reached in the literature, and studies that have examined construct measurement issues continue to receive research attention (Roberts, Varki and Brodie, 2003).Questions as to which (or what number of) factors best define the service quality construct (Llosa, 1998; Carman, 1990) and which scale variables should be included (Richard, 1993), continue to be contested.
The SERVQUAL model developed by Parasuram (1985Parasuram ( , 1988) ) is the most cited and applied model for measuring service quality.This measurement of service quality is based on the comparison of the expectations and performance perceptions of customers.Cronin and Taylor (1992) offer a theorical justification for discarding the expectation portion of SERVQUAL and including only the performance perceptions in the scale (what they term as SERVPERF).
This study adopts the performance only approach for its empirical part for two reasons: First, it is emphasizing that using perceptions scores as in SERVPERF is likely to be more effective and accurate in identifying the overall assessment of service quality (McAlexander, 1994;Cronin, 1992).Also, using only perceptions scores provide superior predictive and convergent validity Cronin and Taylor (1992), especially in the banking sector (McDougall, 1994).Second SERVPERF is more appropriate in measuring service quality in service situation where customers have difficulty in forming precise preconsumption expectations (Mick, 1998;Zeithaml, 2002).Transition economies have relatively short period of experience with free market service situations, so it is reasonable to believe transition economy customers would have more difficulty in forming precise pre-consumption expectations required for using gap measurement instrument (Mick, 1998;Zeithaml, 2002).Consequently, as Albania is a transition economy it is more appropriate to use a scale instrument that measures service performance.On these grounds, this study adopts the performance perceptions approach for its empirical part.
Despite wide application of SERVQUAL, it has been often criticized by researchers (Karatepe, 2011;Chi Cui, 2003;Avkiran, 1994).They suggest using the model considering industry, culture and country specific features.
To provide a lasting solution to the problem of the unsuccessful measurability of SERVQUAL, researchers agreed on a possible modification of the SERVQUAL model to suit the specific service settings (

Research instrument
In this study we decided to rely on BSQ instrument because it is a standard scale, designed specifically for measuring service quality in banking industry (Petridou,2007;Glaveli, 2006); cover all the dimensions of the marketing mix in contrast with SERVQUAL model who concentrate only on service/product dimension; was proved to be more reliable than SERVQUAL model.
Existing research models and measures may be inapplicable to other cultures unless the dimensionality of scales and reliability are demonstrated to exist (Douglas, 1983;Hui, 1985).As reviewed in the literature, when BSQ instrument was used in other culture/country both the validity of the factor structure, as well as the stability of the scale items is questioned (Petridou, 2007).Hence, a generic instrument for measurement of service quality or even one specifically developed for banking industry may not be applicable in its original form.
This paper examines the degree to which a structural model of bank service quality has a similar factor to that of existing BSQ model (Bahia, 2000).In order to determine the relevance of an existing scale to another population the methods of measuring constructs and analyzing data were the same as in the original BSQ study.
The literature review revealed that researchers value that service quality is not a generic construct and have created service quality measurement instrument that are customized according to industry and culture under investigation.In addition to the original BSQ scale, 11 relevant items from literature review and results of the interviews with bank experts were also adapted and incorporated.The literature review provided the theorical basis for this research and helped in deciding which items to include in the questionnaire.The items were modified, taking into consideration the specific contextual industry, society characteristics and the opinions of bank quality experts.The original items were translated and modified into Albanian, and their content validity (wording and meaning) was checked carefully by two Albanian experts.A pre-test was then conducted with a group of bank customers and minor changes were made to the scale to ensure that questions were no repetitive.
A non-probability sample of Albanian bank customers was deemed the most appropriate sampling method for this study due to an identified hesitancy by customers within transition economies to complete customer surveys in general (Kinsey, 1994).Face to face contact is the most effective method to collect data by questionnaires in less-developed countries (Kumar, 2000).
In determining the size sample the rule of thumb of a minimum sample size of ten times the number of free model parameters (Kline, 1998) was followed.
The researcher selected banks' branches within large cities (Tirana -capital city) and also within small city (Korça) in order to support possible generalization for data collection process within Albania.It was decided that it would be best to interview as they exited the bank branch as the last perception about the level of service quality experienced would be still fresh in their minds.This survey took place during the period of January -April, 2015.
The questionnaire was composed of statements and rating scale.It consisted of three sections.In the first section the respondent were asked to provide their demographic information.The second section was composed of 42 service quality statements.For every statement, we asked the respondents to indicate their level of agreement with the statement using a seven point Likert scale ranging from "1 -strongly disagree" to "7 -strongly agree" based on their perceptions.Finally, the last section was composed of questions asking the respondent to gauge the overall service quality of their bank, whether they were satisfied with the bank' services, whether they will recommend the bank to a friend and whether they have ever had problems with the banks' services.
A total of 380 customers from 13 different banks have been approached from whom 352 correctly completed questionnaires have been obtained.The final structure of the sample is in accordance with the market shares of banks in Albania.
Factor analysis was used to identify the underlying structure of service quality in the Albanian banking industry.For reducing and summarizing the data, a principal component analysis with varimax rotation was performed to keep the factors as independent of each other as possible.To achieve these tasks SPSS statistical software programs was employed.After that, reliability and validity of the scale was checked.
The data was checked in terms of appropriateness for factor analysis (test of sphericity and sampling adequacy Kaiser-Meyer-Olkin (KMO) and Bartlett's scores).In this research study, the outcomes of the Bartlett's test confirmed that the sample (χ 2 = 11900.511;p< 0.001) is suitable for factor analysis.
A principal factor analysis was performed, followed by an orthogonal rotation (VARIMAX).According to Hair (2006), factors with eigenvalues greater than 1 and actors loading with values equal to or greater than 0.5 were retained.The result was a scale composed of 28 items loading in four dimensions or factors.These factor structure explain 66.09 % of the total variance.Table 1 shows the four retained factors and their factor loading.

Reliability Analysis
As Nunnally (1978) suggested, reliability should be examined whenever new measures are developed.Cronbach's alpha is the most widely used method to evaluate measurement instrument reliability.Referring to Nunnally (1978) the minimum accepted level of Cronbanch' alpha in social sciences is 0.7.Table 1 provides an overview of the reliability scores.As can be seen from this table, all the alpha coefficients were above the required level 0.7 ranged from 0.838 to 0.94.

Convergent validity
According to Peter' s (1981) definition, the convergent validity of a construct is proven when the construct which is used to measure perceived service quality in banking industry is measured by two different instruments, both measures must converge.As suggested by Bahia (2000) the relationship between scale scores and a customer evaluation about the overall service quality (OSQ) was examined.To evaluate the OSQ the respondent were asked to choose between the following answers categories: (1) very good quality (2) rather good quality and (3) rather bad quality or very bad quality.The one way analysis and Student-Newman-Keuls (SNK) procedure showed that this association was statistically significant (see table 2).This reflects a good convergent validity of our scale.

Nomological validity
According to Peter' s (1981), the nomological validity of a construct could be proven only if it was possible to empirically validate associations between the focal construct (i.e.perceived quality in banking sector) and other constructs to which it is supposed to be related theoretically.As suggested by Bahia (2000) the associations between scale scores and satisfaction, recommendation and problems with the banks were examined.The respondents were asked to choose one of the following categories (1) very satisfying banking services, (2) services rather satisfying and (3) services rather dissatisfying or very dissatisfying in order to answer to the question whether they were satisfied with the bank' services.The respondents were asked to choose one of the following categories (1) frequently or sometimes, (2) rarely and (3) never in order to answer to the question whether they have ever had problem with the bank' services.The respondents were asked to choose one of the following categories (1) yes, (2) rather yes and rather no or no in order to answer to the question whether they will recommend the bank to a friend.Once again, one way analysis and Student-Newman-Keuls (SNK) procedure showed that this association was statistically significant (see table 3,4,5).This indicates a very good nomological validity of our scale.

Conclusions
The present study strives to uncover the critical determinants of service quality in Albanian banking industry and consequently provides an instrument to measure bank customer perceptions of service quality.The results showed that service quality could be conceptualized and measured as a four-dimensional construct consisting of (1) Responsiveness and Informing, (2) Reliability and Security, (3) Commodities and (4) Effective Access.The 28 items making up the instrument are indicators of quality as reported by bank customers since out of the 42 items originally proposed, customers, through their responses choose 28 items in influencing their perceptions of service quality.The 28-items of the proposed instrument consist of 24 items from BSQ model and 3 new items derived from the literature review.The added items were specific for Albanian banking industry.The resulting model differs from the original six dimensional BSQ instrument leading to the conclusion that a simple adoption of bank service quality instrument as it is will provide no reliable results.
(1) Responsiveness and informing -responsiveness include willingness to respond to customer inquiries, needs and incentiveness.Informing include giving information, explain the service (give financial advice) and price (explaining direct and indirect service fees).Informing is seen as a customer right, a bank duty and responsibility but at the same time as benevolence to provide fair information and fast procedure.
Our results showed that Responsiveness and Informing is the most important dimension of service quality followed by Reliability and Security, Commodities and Effective Access.It is very important for Albanian customer to have better information on the offering (good explanations of service fees), term and condition due to poor information flow in Albania as a developing country.Albanian customer evaluate as more important price as a lack or delay of information and discard price in its monetary aspect from the BSQ model.
In high uncertainty avoidance societies like Albania the possibility of failure has to be reduced by handling the problem, therefore responsiveness is very important (Furrer, 2000).Fair treatment is an important factor in Albanian banking system due to negative experiences like frauds and financial systems crisis (Grazhdani, 2015).In collectivist societies like Albania because insisting or negotiating is embarrassing and do not promote harmony (Winsted, 1997) customers expect that employees have to do their best to understand the specific needs of the client (table 1 -BSQ 37 item) and show that they have its best interest at heart.Also, because of collectivism, organization-initiated recovery (Pasongsukarm, 2001; Muller, 2003) should be emphasized.
As Albania is a high uncertainty avoidance society customers prefer to have everything in control therefore constant notification about the recovery process and handling the problem (BSQ 20 -table 1) is essential (Mattila, 2004).So, the item: "bank informs you every time that a better solution appears for a problem" was expected to be a determinant of service quality in Albanian banking industry.Furthermore, for Albanian collectivist customers courteous, formal and empathetic manner recovery behaviors are more significant that compensation (Mattila, 2004).
(2) Reliability and security -include reliability, accuracy, integrity and security elements.This dimension is also very important for Albanian banking customers explaining 20.47% of the total variance.In order to reduce the insecurity associated with the development of Albanian banking industry, manager should insist to deliver exactly the way it is promised.Bank managers in Albania have to improve the communication between advertising departments and employees who come into contact with customers and have to avoid exaggeration in advertisements.
(3) Commodities -include physical facilities, appearance of personnel and environment, physical representations of the service (such as bank statement) and the complete gamut of services.So, for Albanian customers "the complete gamut of services" means a form of commodity.The lack of confidence of confidence might be also "responsible" for the relative importance of the tangible element of service quality in Albanian banking industry.By stressing the bank's tangible elements of the service the credibility and good image of a firm are highlighted (Grazhdani, 2015).
(4) Effective Access -convenient location of service facilities, waiting time to receive the services and effectiveness of service access (no service interruption).In accordance with Glaveli (2006) this dimension is the least important, because of the former political situation in Albania, queuing is something expected.
The service quality measurement instrument developed in this study appears to be reliable, as supported by the internal consistency measured by Cronbanch's alpha and valid as supported by convergent and nomological validity testing.Thus this research represents the first known study to develop and test bank service quality construct within Albania.Perhaps the most important conclusions here is that it offers valuable information on bank customers' perceived evaluations that will help bank managers to make decisions as to what is important for customers and what quality parameters they need to consider to keep their customers satisfied.However, this study should be considered merely as a starting point in this area and the instrument developed here is not an end in itself.
The major limitation of this study lies in the fact that the scale construction is based on BSQ model, the existing service quality literature and on bank managers opinions these inputs might not have captured completely the service quality construct for the banking industry.In subsequent replications, primary qualitative research with bank customer should be performed in order to identify and describe as accurately as possible quality determinants for the banking context in Albania.
This research will provide valuable advice for both academics and practitioners in the Albanian banking industry.This paper will help managers to understand and appreciate the most important dimensions of service quality and thus it guides them towards the required improvements.Bank managers can employ the service quality instrument to identify distinct customer clusters or segments with varying perceptions about service quality.

Table 2 . The relationship between BSQ scores and overall service quality
Student-Newman-Keuls a,b rather bad quality or very bad quality28 3.3227

Table 4 . The relationship between BSQ scores and Problems with the bank ANOVA Bank Service Quality(BSQ)
Student-Newman-Keuls technique Bank Service Quality(BSQ)Problems with the bank N Subset for alpha = 0.05