Assessing the Antecedents of Customer Loyalty on Healthcare Insurance Products: Service Quality; Perceived Value Embedded Model

Purpose: This research aim to investigate the influence of service quality attributes towards customers’ loyalty on health insurance products. In addition, this research also tested the mediation role of perceived value in between service quality and customers’ loyalty on health insurance products. Design/methodology/approach: Based on the literature review, this research developed a conceptual model of customers loyalty embedded with service quality and perceived value. This research applied convenience sampling method. The study surveyed 342 healthcare insurance customers through self-administered questioner. Apart from assessing the reliability and validity of the constructs through confirmatory factor analysis, this research also used structural equation modelling (SEM) approach to test the proposed hypothesis. Findings: The results from the inferential statistics revealed that the healthcare insurance customers are highly influenced by service quality followed by the perceived value in reaching their loyalty towards a particular health insurance service provider. Research limitations/implications: The sample for this study is based on health insurance customers only and it is suggested that future studies enlarge the scope to include others type of customers of different insurance products.


Introduction
In financial services, especially the insurance industry, the financial performance is intimately attached to customers loyalty (Diacon & O'Brien, 2002).As the selling cost of an insurance policy is not recovered unless the policy is renewed (Zeithaml, Berry & Parasuraman, 1996), thus, customer loyalty is one of the most important determinant of economic success to the insurance firms (Mishra & Prasad, 2014;Moore & Santomero, 1999).In this regard, customer loyalty is an area of interest not only in academia but among marketing practitioners (Lovelock, 2008;Sagib & Zapan, 2014).
Although the relationship between customer satisfaction and customer loyalty has been researched in different industry at large, however the examination of the impact of the relationship between perceived value and loyalty has mostly been ignored by the health insurance perspective (Yang, Jun & Peterson, 2004).For this reason, understanding the nature of service qualities towards the customers' perceived value of healthcare insurance products becomes more critical, and hence, customer loyaltyhas also become a top priority issue by numerous researchers (Bloemer & Odekerken-Schroder, 2002;Chumpitaz & Paparoidamis, 2004;Dobre, Dragomir & Milovan-Ciuta, 2013;Iacobucci, Ostrom & Grayson, 1995;Nelson, Rust, Zahorik, Rose, Batalden & Siemanski, 1992;Zeithaml, 1988;Zeithaml & Bitner, 1996).
Commonly, the healthcare insurance service products are distributed through a very complicated insurance agency network (Grönroos, 1982).In this respect, building a distinctive and purposeful relationship between customers and the healthcare insurance providers will require a superior level of service qualities provided by insurance firms (Gera, 2011;Rahman, AbdelFattah & Mohamad, 2014;Wong, Tong & Wong, 2012).
Despite the growing body of knowledge on service quality, there is hardly any research that combined service quality with customer perceived value towards customer loyalty in one integrated model under the scope of healthcare insurance products.Thus, there is a substantial gap exists in testing the service quality and customer loyalty in the healthcare insurance industry.Hence, the key interest of this research paper is to examine the service quality attributes as suggested by Gronroos (1984) towards customer loyalty as well as testing the mediated effect of customer's perceived value in the proposed relationship.In order to answer the following questions: • How do the characteristics of service quality attributes (i.e.functional quality, technical quality and firm's image) influence the customer's perceived value in healthcare insurance service providers?
• Do customers perceive value play a mediating role between service quality attributes and customer loyalty?

Development of Construct
Empirical results indicated that service quality has a positive effect on customer's perceived value (Cronin, Brady & Hult, 2000).Furthermore, Gronroos (2007) described technical quality as the quality of what the consumer actually receives as a result of interaction with the service firms and thereby is considered as important in assessing the quality of service in determining customer's satisfaction.On the other hand, functional quality is the method of how customers obtain the technical outcome in terms of viewing the service that has been received.Hence, functional quality has been initially conceptualised in the GAP model that was proposed by Parasuraman, Zeithaml and Berry (1985).
Aside from technical quality and functional quality, firm's image was drawn as an initial decision for customers to determine the quality of service provided (Clow, Kurtz & Ozment, 1998;Faché, 2000;Mazursky & Jacoby, 1986).That's mean, when services are difficult to be evaluated, then firm image shown as an important factor that influencing the perception of customers towards the quality of service provided (Andreassen & Lindestad, 1998;LeBlanc & Nguyen, 2001).Previous studies established a positive relationship between the firm image and expectations in several service industries, such as catering, financial services, or travel agencies (Rodríguez del Bosque, San Martín, Collado & García de los Salmones, 2009;Clow, Kurtz & Ozment, 1997;Devlin, Gwynne & Ennew, 2002).A positive firm image develops cues that give source of the appearance of expertise that are capable of increasing a message's of persuasive effect (Schindler & Bickart, 2005).
As stated earlier, many studies have reported that there is a direct positive relationship between service quality and customer's loyalty (Bolton, 1998;Cronin & Taylor, 1992;Oliver, 1980).
According to the aforementioned discussion, service quality attributes, may also be related to customer's loyalty, mediated by customer's perceived value.Previous empirical studies indicated that there was a positive relationship between service quality and customer loyalty (Avkiran, 1994;Bitner, Booms & Tetreault, 1990;Johnston, 1997).In this regard, the current paper not merely adopted Gronroos (1984) model, but also expand the adopted model to include both customer perceived value and customer loyalty constructs in one integrated model for better understanding to the customers of healthcare insurance industry.Consequently, this paper attempt to examine three hypotheses that resulted from service quality attributes and customer loyalty through customer perceived value.

Functional Quality and Technical Quality
To be in the same line with other researchers like Dadfar, Brege andSemnani (2013), Sharabi andDavidow (2010) and Zeithaml and Bitner (2003), this paper employed functional quality measured through SERVQUAL items, in order to evaluate and assess the perception of customer's into specific dimensions, to include (reliability, responsiveness, assurance, empathy, and tangibility).In this research, we measured the service quality of health care insurance, by also investigating the technical quality terms through competence, reliability and responsiveness.Reliability is defined as the ability to perform a promised service consistency and precisely (Andaleeb & Conway, 2006).Apart from that, responsiveness are known as willingness to help customers by providing them with quick and prompt service (Zeithaml, Bitner & Gremler, 2006).In this regard, Gefen, Straub and Boudreau (2000) and Sandhu and Bala (2011) point out that both reliability and responsiveness ultimately, influence the customers' perceived service quality.Layton (1994) identifies knowledge; technical competence and technical will are the items to assess the technical quality.Rychen and Salganik (2003) described that competence is something that mobilize technology recourses.
Thus, to meet the existing prerequisites of Malaysian healthcare insurance customers', providers may require being complemented in responding the changes that occur in today's technology advancement (Jonason & Eliasson, 2001).

Firms Image
Corporate Image described as a company's image as how "customers see and perceive" the firm (Gronroos, 2007).Clow and Beisel (1995) and Mazursky and Jacoby (1986) added that the customer's image of particular service firm will have a direct effect on their attitude on future expectations.There is evidence that "image" is significantly related to perceptions of quality (Darden & Schwinghammer, 1985).Because, the firm image will precede customer evaluations as these assessments are components of the image (Mazursky & Jacoby, 1986).

Perceived Value
The concept of value was investigated by Zeithaml (1988).The researcher intends to identify two popular definitions of value by supporting literature.These are: • The value is the quality I get for the price I pay, and These two definitions have merged together and defined perceived value as the consumers' overall assessment based on the dimensions of utility theory.On the assumption that perceived product value and perceived service value are analogous, Bolton and Drew (1991) also carry the definition by Zeithaml (1988) to the service-oriented products.Rust and Oliver (1993) argue that it is likely that the value, like quality, is an encounter particular input to satisfaction.As stated earlier, many researchers have been studying service quality associated with customer satisfaction in life-insurance sector; however, what has not been tested is how the perceived value used as a mediating in between service quality and loyalty under the scope of the health insurance sector.

Customer's Loyalty
In this paper, we define customer loyalty as a composite construct by combining behavioural and attitudinal factors into one composite construct.Palmatier, Dant, Grewal and Evans (2006) view customer loyalty as a combined reflection of intentions, attitudes, and seller performance indicators.The loyalty concept in service provision in formulating marketing strategies is not a new phenomena (Wong et al., 2012).Apart from that, the role of managers in the service industry is to formulate strategies to raise the level of their customer loyalty that apparently lead service firm growth and foster business sustainability (Chen & Cheng, 2012;Keiningham & Aksoy, 2012).
Researchers also argued that customer loyalty increase firm's revenue and reduces the costs of customer acquisition and retention (Auka, 2012;Rapp, Beitelspacher, Schillewaert & Baker, 2012;Reichheld, 1993;Reichheld & Sasser, 1990).Thus, Krumay and Brandtweiner (2010), Mokhtar and Maiyaki (2011) and Wong et al. (2012) defined customer loyalty as the degree to which a customer exhibits repeat purchasing behaviour from a service provider, possesses a positive attitudinal disposition toward the provider, and considers using only this provider when a need for this service arises.The level of customer loyalty is measured by how willing a respondent is to recommend and say a good word-of-mouth about their respective healthcare insurance firms to their friends, relatives (Alok & Srivastava, 2013;Arasli et al., 2005;Elmayar, 2011;LeBlanc & Nguyen, 2001).

Conceptual Framework
Based on the above literature review; this research adopted the Gronroos service quality model for evaluating the service quality attributes (i.e.Functional quality, technical quality and firm's image) along with the customer's perceived value and customer loyalty.This research also investigates the role of customer's perceived value as a mediating construct in between the service quality attributes and customer loyalty.Figure 1 explains the proposed theoretical framework and proposed hypotheses to test in further in order to conclude the research.

Data Collection and Sampling
Since the purpose of this research is to investigate the influence of service quality attributes towards customers' loyalty on health insurance products, a structured questionnaire was developed to collect the needed information from the healthcare insurance customers from Klang Valley area in Malaysia.Questionnaires were self-administered and distributed by utilizing a convenient sampling method.We used a convenience sampling method to select the participants for the reason that it is considered an effortless approach to get respondents to participate in a study of this nature.Besides, prior related researchers have used the similar method in selecting participants.This sampling method is frequently used in marketing research.It is also cost effective as we can choose anyone to be a participant with ease.The sampling frame for conducting the principal component analysis (PCA) comprised with 342 healthcare insurance customers.
Initially, the measurement items of the studied variables were drawn from existing theories and instruments adapted based on the previous studies.The adapted items were pre-tested with 35 healthcare insurance customers.From the results of the interview, a number of questions were modified to suit the context of this study.Cronbach alpha was used to examine the reliability test that is show that all items were above the cut-off value 0.70 (Nunnally & Bernstein, 1994).Out of 450 distributed surveys, only 355 surveys were completed and returned.After, conducting the primarily screening test we confirmed that 342 surveys are valid for further analysis.

Construct Measures
The questionnaire has two parts.Part one includes the respondents' demographic information, e.g.: sex, age, job, educational level, marital status, race and the length of the period subscribing healthcare insurance.While Section two includes the indicators of the main elements of this study (service quality attributes, perceived value and customer loyalty).The items were developed from SERVQUAL elements to measure the functional quality variables.
Functional quality was measured through 16 items (reliability-four items, assurance -four items, tangibility-four items, and empathy-four items).Adapted from Kang and James, (2004), and Parasuraman, Zeithaml and Berry (1988).This scale also proved to be adequately reliable (α= 0.92).Technical quality was operationalized by using four items adapted from Duodu and Amankwah (2011) (α= 0.87).Firm's image was measured by four items adapted from Gurses and Kilic (2013).This scale also proved to be adequately reliable (α= 0.76).Perceived value was measured by four items adapted from (Sweeney & Soutar, 2001).This scale also proved to be adequately reliable with Cronbach α= 0.82.At last, the customer loyalty was measured by four items adapted from Aliza (2012) with α=0.91.All items were measured via 7-point Likert scale with rating scales categories ranging from 1= strongly disagree to 1=strongly agree.

Data Analysis Method
In this paper, the procedures to analysis the data on hand were systematized into two main steps.The first step was carried on confirmatory factor analysis (CFA).In this regard, researchers like Pett, Lackey and Sullivan (2003) and Thompson (2004) assured that using CFA provide the analysis with wide range of benefits to conclude; verifying the factor structure, confirm the convergent and divergent validity of the construct as well as CFA can support researcher to analyse and define the proposed model from a relatively significant of latent constructs to be presented by a set of points (Hair, Black, Babin & Anderson, 2010;Nunnally, 1978).In addition, examining the general fit of the proposed model can help researchers to test and answer the research questions.Consequently, factor constructs were also employed based on the maximum likelihood (Fish, 2005).To be sure of the model goodness of fit, several criteria as suggested by (Jöreskog & Sörbom, 1986), was also used to include, Chi-Square test, root mean square error of approximation (RMSEA), goodness-of-it index (GFI), adjusted goodness-of-fit index (AGFI), normed fit index (NFI), and comparative fit index (CFI).
On the other hand, the second step was conducting the Structural Equation Modeling (SEM) in order to test the previously mentioned hypotheses (Costello & Osborne, 2005).

Empirical Findings
After the data collection, Table 1 shows that 342 usable questionnaires were used for further analysis, it also showed that (192) respondents were Malay.Followed by (100) respondents were Chinese, and only 50 were Indian.In regard to respondent gender, 55.7 % was female, and 44.3 % were male.Based on the respondents' profile, it was found that most of the respondents were within 31 to 35 years of age categories (57.6%) followed by 26-30 years (37.7%),51 and over (4.7%).Apart from that, (78.5)% respondents have their insurance policy for more than five years.This analysis also found that most of the respondents (82.3%) have both life and health insurance scheme under various private, local and international companies.In contrast, more than half of the respondents worked for different local and international companies; followed by 18% are students and 27% of various government departments.The bulk of the respondents were at least a bachelor degree holder (63%) complied by a diploma (26%) and (11%) for higher education.The respondents were married (85%) followed by single (15%).

Analyzing the measurement model through (CFA)
We used CFA in order to investigate the generalizability of multidimensional measures of the constructs.This research obtained unidimensionality through recognize the items of each factor loading for individual latent construct.This paper retained the items under each construct that have a loading higher than 0.6.According to the analytical output, there is no item loaded below 0.60.As a result, all items of each construct were remained.
As mentioned earlier, by using three fitness-indexed to examine the fitness of the proposed model, the fitness categorized were absolute fit, incremental fit and parsimonious fit Table 2, highlight the indexed and the level of acceptance, were utilized in both default and revised model.From the results of the default model for functional quality, the modification indices for the covariance of measurement errors were 37.75 between Rel1 and Rel2,33.34 between Ass1 and Ass2,and 55.321 between Emp2 and Emp3.These three sets of measurement error were logically conceivable to be correlated.Based on observations, we deleted items Rel2, Ass2 and Emp2, which had almost the similar meaning.After deleting (Rel2, Ass2 and Emp2) items and testing, the revised measurement model was found fit for further analysis (see Table 2).
For technical quality, the modification indices for the covariance between the measurement errors of Tq1 and Tq2 were found to be 14.23.The correlation of these errors was logically possible, and, for this reason, the model was revised to incorporate this path (Table 2).After adding (Tq1 and Tq2) parameters, the measurement model fit indices of technical quality were found to fit adequately for further analysis.Among them, there were four measuring items for both constructs firm's image and consumers' perceived value.The fit indices for the default measurement model were found to fit adequately for these two constructs (see Table 2).
In addition, customers' loyalty modification indices showed that the covariance between CL3 and CL4 was 18.62.It is rationally conceivable that this pair of measurement errors is correlated as this research deleted one of the items (CL4); for this reason, the measurement model was revised.As a result, the amendment shows that all indices express an adequate fit of the model.
In addition to mentioned earlier, CFA was also used to verify the theoretical patterns of factor loadings on pre-specified constructs that represented the actual data.The model fit of the unidimensional construct, indicated by GFI, AGFI, RMSEA, CFI, NFI, TLI, resemble normal (see Table 2), which confirmed construct validity.The value of x2 was significant for unidimensional constructs with a sample size (N=342).However, the normed x2 was within the suggested guidelines (see Table 2).To determine the mediating role of customers' perceived value of functional quality, technical quality and firms image on customers' loyalty this research used a recommended method of testing the indirect effect through the bootstrapping technique (Bollen and Stine, 1990;Shrout and Bolger, 2002;Fritz et al. 2012;Kline, 2010;Baron and Kenny, 1986;Preacher and Hayes, 2010).fit.This research also concluded that all the proposed relationships in the hypotheses received favourable support.In addition, this study compared the magnitude of direct and indirect effect among functional and technical quality along with firms image on customers' loyalty.The total effect of technical quality was 0.64 with an indirect effect of 0.50 and direct impact of 0.14.For this reason, we concluded that the indirect effect of perceived value was more dominant than the direct effect between technical quality and customers' loyalty.Hence, the customer loyalty is significantly influenced by the overall technical quality of a particular operator where perceived value plays a strong mediating role.
Likewise the total effect of functional quality by was 0.75 with an indirect effect of 0.50 and direct impact of 0.25.We, for this reason, conclude that the indirect effect of perceived value was more dominant than the direct effect.Thus, customers' loyalty gain is significantly explained by the customers' perceived level of functional quality where customers' perceived value plays a strong mediating role.Finally, the total effect of firms was found to be 0.76 with an indirect effect of 0.30 and direct impact of 0.40.Consequently, we concluded that the direct effect of perceived value was more dominant than the indirect effect.Hence, customer loyalty is significantly influenced by the firm's image where customers' perceived value plays a partial mediating role.Thus, we can conclude that a strong mediating effect flows from FQ®CPV®CL; and TQ®CPV®CL compared to FI®CPV®CL.
The estimated results show that the reliability and validity were reliable in assessing the unique nature of the variance on each construct (TQ, FQ, FI, and CPV).In addition, the findings indicate that the majority of the customers' is aware of the existing services and the providers based on the technical, functional and their firm's image where the perceived value play an important role to reach the loyalty towards a healthcare insurance.Subsequently, this research also suggests that Malaysian healthcare insurance products need to concentrate more on uplifting their technical and functional qualities to gain more customers as well as retain the existing one.

Conclusion and Managerial Implication
This paper has provided empirical evidence which help practitioners of health care insurance products in Malaysia to be more focusing on the way of customers perceived the service quality as well as perceived values which in turns help on acquiring and maintaining customers.The results also showed that the measurement model was fit appropriately to investigate and comprehend customers loyalty toward the healthcare insurance industry in Malaysia.
In addition, customer's particularly in Malaysia are greatly concerned about the service quality and perceived value to reach loyalty towards healthcare insurance service providers which this finding consistent with Dick and Basu (1994) and Cronin et al., (2000).This finding is also consistent with the result found by Heskett, Jones, Loveman, Sasser and Schlesinger (1994) that loyalty is a direct result of customer perceived value and the quality of services obtained by their respective customers.On the other hand, the rivalry of the insurance industry in Malaysia is more intensified now than ever before.They need compete not only for the service quality that they are provided with maximizing to the medical network, but they also concern for acquiring a new customers and retention of old customers by direct and indirect price reduction.This finding is consistent with Zeithaml (1988), who asserted that in which goods and services can comprise any number of attributes; consumers typically infer quality from one or a few numbers of these attributes.In addition, Schlesinger and von der Schulenburg (1993), suggest that service quality is a factor upon which the customer can distinguish between identical insurance products.There are other characteristics of insurance that also influence the customer's choice.Such as, Leste and Wanderley (1997), reported in their study of the insurance industry that there are group of consumers, which takes the dimensions of a segment, who are particularly interested in the help and care that they receive from insurers and in the insurers' technical ability to provide information for the insurance products.In this regard, the health care insurance service providers should make such arrangements that would give extra value to the customers, in order to attain the customer loyalty.The service quality of the health insurance product is one the most important factor in increasing the overall customer loyalty.As a result, the healthcare insurance providers should be focused on the service quality attributes in order to sustain the loyalty of their quality seeking customers.The results of this study also suggest that the service provider should remain alert to take appropriate measures to enhance their service to strength the customer perceived value.In addition, this study suggests that reduced perceived value leads to decrease of customer loyalty.For this reason, the healthcare insurance providers in Malaysia should give prime focus on these elements to gain more long-term loyalty of their clients.
Although the perceived service quality has often been mentioned in consumer behavioural research, the intensity of the term "perceived service quality" has not been sufficiently examined in health insurance research.Thus, some important factors that affect perceived service quality have not been recognized or examined even in Malaysian health insurance sector.This research utilizes both technical quality and functional quality along with firm's image as suggested by Gronroos (1984) toward customer perceived value on customer's loyalty.As a result, the outcome of this research has highlighted the cues that lead to the assessments of each construct of the study.Therefore, this research successfully explains the contribution of the exogenous constructs in clarifying the customer loyalty from the perspective of health insurance consumers.

Study Limitation and Direction for Further Study
This research additionally contains some limitations that suggest a possibility for additional further research.The data originated from a convenience sample procedure.Further studies may include a more representative sampling structure.The constructs could be further explored and compared across services of different types of industry.This paper could not include Sabah and Sarawak states where a notable population of health care insurance customers also exists.The demographic characteristics, such as age, gender and race are often assumed to have a significant effect on customer intention to have healthcare insurance products.While this research used a wide variety of measures to minimize any possible halo effects, it would also be useful for future research to assess some of these variables more directly, such as a cross-country comparison to measure the relationship of overall satisfaction and behavioural intention on the consumers' perceptions of the service quality of a particular provider in the healthcare insurance context.In addition, this research assured that further investigations on this particular issue (healthcare insurance products) are required to investigate more with larger samples before generalization can be made.

Figure 1 .
Figure 1.Theoretical Framework for the antecedents of Customer loyalty on Healthcare Insurance

Table 1 .
Descriptive statistics of demographic variables

Table 2 .
Level of acceptance to test the fitness of the individual constructs

Table 3 .
Level of acceptance of the unidimensional construct
4.5.Testing HypothesisIn analyzing the results of the structural model, this study utilized standardized estimates for all hypothesized paths.This research found that the functional quality (FQ), technical quality( The full model was tested based on the measurement model previously validated through CFA in the study.The fit indices of the full model were χ²/df= 3.536 (χ²= 954.767/