Examining the Key Factors Affecting e-Service Quality of Small Online Apparel Businesses in Malaysia

e-Service quality (eSQ) is increasingly recognized as an important aspect, as well as the key to determining the competitive advantage and factor in the long-term retention of firms operating online. This study, therefore, is aimed at identifying the key determinants of eSQ among the small online apparel businesses in Malaysia. This study used a cross-sectional design, and data were collected from 765 customers who purchased apparel online at the point-of-purchase. Findings of this study indicate that “product information quality,” “website design,” “security and privacy,” and “expected consumer service” are the key determinants of eSQ among small online apparel businesses in Malaysia. The implication for the owner-managers of the apparel businesses in Malaysia is that they must be aware of the significance of the key eSQ indicators while designing their businesses, to attract and retain customers.


Introduction
e-Commerce has changed the traditional ways of doing business and has created many new forms of businesses such as selling apparel online. To understand how online consumers behave, several studies have been conducted on selling apparel online in recent years (Chae, Black, & Heitmeyer, 2006;Li & Zhang, 2006). As noted by Parasuraman, Zeithaml, and Malhotra (2005), the key to sustain in the online marketplace and gain competitive advantage is to provide a better e-service quality (eSQ). This is why recent studies have been focused on how to conceptualize, measure, and manage service quality and its impact in the virtual environment (e.g., Gregory, Wang, & DiPietro, 2010;Guo et al., 2012).
According to Parasuraman et al. (2005), eSQ is "the extent to which a website facilitates efficient and effective shopping, purchasing and delivery" (p. 135). This definition shows that eSQ does not solely rely on the pre-purchase phase (product information, ease of use, ordering information, and personal information protection) but extends to the post-purchase phase as well (delivery, customer support, fulfillment, and return policy). The nature of the online environment is different from the traditional retail environment. Hence, Ladhari (2010) classified it into four factors: convenience and efficiency, safety and confidentiality, absence of face-to face contact, and production of service quality.
The current rapid growth of e-commerce and business to consumers reveal the convincing advantages of shopping online. The future of e-commerce seems to be bright for Malaysia. The impact of e-commerce growth has led customers to switch from their traditional purchasing behavior to an online purchasing behavior especially in buying apparel. Park and Lee (2009) stated that apparel is now being sold via the Internet; in 2006, 10% of all apparel purchased in the United States occurred online. A study by Ha and Stoel (2014) found that apparel is often associated with experience because customers can evaluate its quality after purchasing and using it. However, customers take a risk when purchasing apparel online because they cannot try it on to check whether it fits, or its texture or color in natural light (Kim, Kim, & Lennon, 2006). Due to the need for an adequate amount of product information to make a purchasing decision, online shoppers largely depend on information available on the web. Many customers prefer to purchase apparel online due to the lack of confidence-purchasing online means that they do not have to go to a changing room to try on the clothes. According to Parasuraman et al. (p. 217, 2005), eSQ is the extent "to which a web site facilitates efficient and effective shopping, purchasing and delivery." This definition shows that eSQ does not solely rely on the prepurchase phase (product information, ease of use, ordering information, and personal information protection), but it also extends to the post-purchase phase (delivery, customer support, fulfillment, and return policy). The nature of the online environment is different from the traditional retail context. Hence, Ladhari (2010) classified it into four factors: convenience and efficiency, safety and confidentiality, absence of face-to face contact, and production of service quality. Wu (2011) also added that eSQ did not affect customer satisfaction directly but had an indirect effect, or in other words, customer satisfaction played a mediating role in the relationships among e-service qualities. In this study, the researcher believes that eSQ should be measured at the point-of-purchase in understanding online purchases. Point-of-purchase is measured as the time when customers purchase a product whereas post-purchase is measured after they have received the product. Many previous studies combine items at prepurchase and post-purchase for instance service delivery and customer service. However, as this study is measuring at the point-of-purchase, expected service delivery and expected customer service are differentiated from service delivery and customer service at the post-purchase stage. This suggests that these elements play a different role at every stage, and the research was planned on this basis. Expected service delivery and expected customer service are measured when customers place their orders. This is in contrast to service delivery and customer service at the post-purchase stage, which happens after customers have received their products. This study aims to identify the key determinants of eSQ from a wide range of determinants used and suggested in earlier studies, to provide a better understanding of eSQ for small online apparels business in Malaysia.

Literature Review
Empirical research is needed to shed light on the detailed determinants of eSQ and its impact on the perception of customers in the online environment. This is to reflect the limitations of the existing literature to provide a conceptual model for eSQ. This study proposes that eSQ consists of general belief, information quality, merchandise attributes, transaction capability and payment, security and privacy, expected delivery, and expected customer service.

Theoretical Foundation
The theory used for this study is the Theory of Reasoned Action (TRA) formulated by Ajzen and Fishbein in 1980. Accordingly, TRA advocates that a person's behavior toward an object is formed by the intention to perform that behavior. This theory is widely used in explaining a person's behavior. Ajzen and Fishbein (1980) added that intention is the best predictor of behavior as intention is the cognitive representation of readiness of a person to perform a given behavior, and this behavior is determined by three factors, which are attitude toward a specific behavior, the subjective norms, and the specific behavioral control. This study thus postulates that customer' purchase intentions provide an acceptable surrogate for actual online purchase behavior. In other words, their (customers') satisfaction with the eSQ provided by the website leads to their intention to purchase online from the specific website. Indeed, many previous studies (e.g., Lee & Lin, 2011;Pookulangara, Hawley, & Xiao, 2011;Weisberg, Te'eni, & Arman, 2011) in the area of online business research have applied this theory.

General Belief About Online Shopping
According to Ajzen and Fishbein (1980), belief is an attitude, which refers to a person's favorable or unfavorable evaluation of an object. In other words, it corresponds with the information he or she has about the object. The object of belief can be a person, a group of people, an institution, behavior, a policy or an event, and the connected attribute may be any object, trait, property, quality, characteristic, outcome, or event. The study conducted by Verhagen and Dolen (2011) introduced a model and showed how beliefs about functional convenience (e.g., online store merchandise attractiveness and ease of use) and representational delight (e.g., enjoyment and website communication style) were related to online impulse buying. These finding are supported by a previous study conducted by Madhvaram and Laverie (2004) who suggested that website elements (e.g., merchandise attractiveness) would likely lead to impulse buying.

Information Quality, Website Design, and Merchandise Attributes
Consumers use a variety of methods to gather information. They seek information from trusted and respected personal sources such as family, friends, and peers; they use the Internet to compare service offerings and search for independent reviews and ratings; they rely on firms with a good reputation; and they look for guarantees and warranties (Wirtz, Chew, & Lovelock, 2012). Sufficient information on a website will encourage customers to shop online (Kuo & Chen, 2011). Coker (2013) added that Information quality is due to the absence of tangibility on the Internet; the key driver of online consumer traffic would be the available subject matter that is relevant to consumer needs or goals. Besides, another factor that affects customer satisfaction is the website design. According to Gregory et al. (2010), the design of the website is associated with consumer perception of the organization and order at the site. In addition to having sufficient information and a good website design, Guo, Ling, and Liu (2012) believes that merchandise attributes play a significant role in whether customers are satisfied or dissatisfied with their shopping experiences.

Transaction Capability, Response, and Security/ Privacy
Transaction capability and payment is an important factor when customers consider online purchases. The majority of B2C trades are paid by credit card at about 60% and the next most popular payment method is Electronic Funds Transfer at about 30% (Turban, King, & Lang, 2010). There are many types of online payments, such as credit card, payment by cash and telegraphic, that are provided by most of the online retailers; however, when choosing the type of payment, sometimes customers consider the convenience of payment together with the issues of payment security (Guo et al., 2012). It is generally accepted that website speed is of major importance for a successful website. Internet user's patience levels are constantly decreasing and the majority of visitors will leave a website if it does not load within 3 s (Jorg, 2013). Many researchers find that security and privacy are indeed important for online sales in recent years (Akinci, Atilgan-Inan, & Aksoy, 2010;Cho & Menor, 2010;Hsu, Hung, & Tang, 2012;Ladhari, 2010). Security system is one of the most important issues.

Expected Service Delivery and Expected Customer Service
Expected service delivery and expected customer service are measured when customers place their orders in contrast to service delivery and customer service at the post-purchase stage, which happens after customers have received their products. Besides having a standard delivery, an e-retailer can offer express delivery and cash on delivery (COD) based on customers' preferences on how the product is to be delivered. In addition, the price stated for delivery charges has to be reasonable and without any hidden costs. Meanwhile, expected customer service is more than policies, procedures, and responsiveness in handling complaints. It also refers to the return policy, customer's right to change, and customer's problem solving. At this stage too, customers are offered many interactive communications with the company. For instance, customers can use the chat room, email, telephone, bulletin board, and feedback forms to communicate with the company. According to Teo, Oh, Liu, and Wei (2003), interactivity features have a significant relationship with the consumers' perceived value. Indeed, Zeithaml (1988) added that customer value has been recognized as the key for businesses to succeed.

Research Methodology
This study is aimed at identifying the key determinants of eSQ among the small online apparel businesses in Malaysia from a selected list of determinants, namely (a) general belief, (b) information quality, (c) website design, (d) merchandise attributes, (e) transaction capability, (f) security/ privacy, (g) expected service delivery, and (h) expected customer service. This study used a cross-sectional design, and simple random sampling method was used to collect quantitative data from individuals who purchased apparel online.

Research Instrument and Facilitating Software
Online survey using the SurveyMonkey platform was chosen to collect the data for this study. The questions were developed using a 7-point Likert-type scale ranging from 1 (strongly disagree) to 7 (strongly agree). The questionnaire design was straightforward and comprehensible, avoiding ambiguities and estimations, and misleading and presumptive questions.
Items for "general belief" about online shopping were adopted from Hirst and Ashwin (2009), Rajamma and Neeley (2005), and Swinyard and Smith (2003). "Information quality" of purchase website was operationalized by the following variables: accuracy, comprehensibility, completeness, and relevance. These items were adapted from previous studies by Muyllea, Moenaert, and Despontin (2004) and Jeong, Oh, and Gregorie (2003). The "website design" construct was operationalized by website structure, navigation, combination of colors, and ease of use. These items were adopted from Kim and Stoel (2004), Muyllea et al. (2004), and Jeong et al. (2003). "Merchandise attributes" were divided into aspects of purchase incentive and product variety, and adapted from Xia, Mengqiau, Fang, and Peihong (2008) and Szymanski and Hise (2000). "Security and privacy" issues were considered and adapted from Wolfinbarger and Gilly (2003).
Items from Kim and Stoel (2004) and Xia et al. (2008) were adapted for "transaction capability and payment." The interviews revealed that to satisfy their customers, the owners of the online apparel businesses offered customers a minimum deposit requirement; provided many options for payment such as post office remittance, online payment, and COD; and also various options in receiving proof of payment (e.g., email, SMS, etc.). These added values are incorporated into the questionnaire to appropriately measure the "transaction capability and payment" construct. "Expected service delivery" was adapted from Rossiter (2002) and three items were developed and added by the researcher to refine the definition of this construct based on responses from the indepth interviews. The interviews revealed that the major delivery activities and expectations specific to online apparel retailers in Malaysia were that consumers expect a standard price for a delivery service, expect online apparel retailers to provide clear terms and conditions for delivery service, and also expect to be rewarded through this service for purchasing more items. These activities were incorporated into the questionnaire to appropriately measure the "expectation delivery service" construct. The interview's findings were also used to refine the "expected customer service" questions offered by Wolfinbarger and Gilly (2003) in their study that was carried out in the United States. For instance, the findings from Question 2 were extracted into four questions in the survey. Based on the interview's finding, online business retailers offered customers a refund if they were not satisfied with the purchase, guaranteed an exchange if customers were not satisfied with the product, provided online tracking for products, and customers could contact them through many options such as the chat room, email, and phone. These findings guided respondents in the Malaysian context in understanding the item being referred to.

Sample Selection and Data Collection
The target population for this study was defined as individuals who purchased apparel online at the point-of-purchase through Blogspots and websites in Malaysia. To establish the sample frame, a list of Blogspots and websites selling apparel online was obtained from the Internet. The respondents were obtained from the owners of the online shops. The respondents were approached by the researcher for permission to email the questionnaire's online link to them and the researcher assured them that their particulars would be kept confidential and that their particulars would be deleted once the survey has been completed. Random sampling was used as this research sought to generalize the results obtained as much as possible from the Internet users who purchased online. In this phase of the study, 836 online customers were surveyed at the point-of-purchase. However, the actual pointof-purchase sample with complete and usable information for this study was 765 online apparel customers (after deleting 71 cases due to incomplete responses, missing data, and outliers).

Respondents' Demographic Characteristics
A complete data were collected from 765 customers who purchased apparel online at the point-of-purchase through Blogspots and websites to examine the factors affecting eSQ. As noted in

Reliability Analysis
Reliability test was performed to test the internal consistency of general belief, information quality, website design, merchandise attributes, transaction and payment, security and privacy, expected delivery service, and expected customer service. The higher the Cronbach's alpha value, the more reliable the measurement instrument. The results indicate that the measurement instrument for the study for each factor is reliable since the Cronbach's alpha values range from 0.801 to 0.949, which exceed 0.6 as the recommended value.

Factor Analysis: Principal Component Analysis
Interpretive adjectives for the Kaiser-Meyer-Olkin (KMO) measure of sampling adequacy are 0.90 is marvelous, 0.80 is meritorious, 0.70 is middling, 0.60 is mediocre, 0.50 is miserable, and below 0.50 is unacceptable. The value of the KMO measure of sampling adequacy for this set of variables, as noted in Table 2, is 0.947, which would be labeled as "marvelous." As the KMO measure of sampling adequacy meets the minimum criteria, it implies that it is no longer vital to examine the anti-image correlation matrix. It is proven by the adequate values attained for all variables in measuring the sampling adequacy (MSA) where the overall value is greater than 0.8 and the minimum value of the antiimage correlation is 0.903, whereas the maximum value is 0.972. Therefore, the factor analysis is appropriate for these data. Bartlett's measure tests the null hypothesis that the original correlation matrix is an identity matrix. The factor analysis needs to establish the relationships between variables and if the R-matrix were an identity matrix then, all the correlation coefficient would be zero. Therefore, the factor analysis is appropriate as there are some relationships between the variables in this study where the Bartlett's Test of Sphericity is shown to be significant (p< .05).
There are 42 items used to identify the key determinants of eSQ. However, after conducting the factor analysis, seven of these items were deleted as the respective factor loadings were lower than 0.4. Table 3 shows the eigenvalues associated with each linear component (factor) before extraction, after extraction, and after rotation. Before extraction, 35 linear components were within the data set. The eigenvalues associated with each factor represent the variance explained by that particular linear component. Table 3 also shows the eigenvalue in terms of the percentage of variance explained. Factor 1 explains 42.972% of total variance whereas subsequent factors explain only small amounts of variance. Subsequently, this study also extracted all factors with eigenvalues greater than 1 (Extraction Sums of Squared Loadings), and remained with eight factors. The values in this part of Table 3 are the same as the values before extraction except that the values for the discarded factors are ignored (the table is blank after the eighth factor). The final part shows the Rotation Sums of Squared Loadings. Rotation has the effect of optimizing the factor structure and one consequence for these data is that the relative importance of the eight factors is equalized. Before rotation, Factor 1 accounted for considerably more variance than the remaining seven (42.97% compared with 8.44%, 6.33%, 4.84%, 3.65%, 3.06%, 2.72%, and 2.25%); however, after rotation, it accounts for only 17.98% of variance compared with 11.39%, 9.83%, 8.27%, 7.80%, 6.72%, 6.40%, and 5.88%, respectively. Moreover, Table 4 shows the communalities after extraction. Principal components analysis works on the initial assumption that all variance is common, in which the values of communalities are all 1 before extraction. Therefore, the communalities after extraction reflect the common variance in the data structure. The study indicates that 71% of the variance associated with Question 1 (variable 1) is common variance. In other words, the amount of variance in each variable that can be explained by the retained factors is represented by the communalities after extraction. All the values after extraction is acceptable because the sample size exceeds 250 and the average communality is greater than 0.6. The average communalities can be found by adding them up and dividing them by the number of communalities (25.992/35 = 0.743). Therefore, the study can retain all the factors with eigenvalues of above 1 (Kaiser's criterion). Table 5 shows the factor loadings that resulted from the Varimax rotation. These eight rotated factors are just as good as the initial factors in explaining and reproducing the observed correlation matrix. The study that involved 35 variables is now represented by the eight rotated factors namely as general belief, information quality, website design, merchandise attributes, transaction and payment, security and privacy, expected delivery service, and expected customer service.
As noted in Table 5 on the Rotated Component Matrix, the result shows the value for each component. The remaining items for General Belief are three, and Item 2, "I believe shopping online is compatible with my lifestyle," contributes the highest item, which is 0.804. In Information Quality, the remaining items for this factor are five, and Item 7 (see Table  5) was recorded with the highest item score at 0.791. Item 15 in website design was recorded with the highest item score at 0.835. Merchandise attributes factor has four items left and the highest item score was recorded for Item 20 at 0.815. Meanwhile, in the transaction capability and payment factor, the remaining items are four, and Item 27 was recorded with the highest item score at 0.682, followed by security and privacy where the items left are four and the highest item score is recorded for Item 31 (Table 5). Another two factors left are Expected Delivery Service and Expected Customer Service. The remaining items for Expected Delivery Service are four and the highest item score is 0.691 for Item 34, and Item 41 for Expected Customer Service was recorded with the highest item score. Among the eight factors, Website Design had the highest factor loading at 0.794, followed by Information Quality at 0.771.

Discussion
The researcher is interested in perceived eSQ at the point-ofpurchase, instead of the pre-sale stage, to capture respondents' true evaluation of eSQ factors influencing the choice of an online apparel transaction. Responses obtained at the pre-sale stage tend to reflect attitudes of respondents prior to their purchase actions and the value judgments provided by these respondents may be provisional and differ from their evaluation at the point-of-purchase. Therefore, to minimize potential deviation and to ensure an appropriate representation of actual experience, this study focused on the point-of-purchase stage. Based on the findings, the eight factors (general belief, information quality, merchandise attributes, website design, transaction capability and payment, security and privacy, expected delivery service, and expected customer service) are identified as factors of eSQ. Findings of this study showed that "I believe shopping online is compatible with my lifestyle" had the highest factor loading among the relevant indicators in general belief. As Malaysian Internet users have increased in number from 16 million in 2008 to 18.9 million in 2012, this provides a strong indication that the Malaysian people are pleased with the advantages of online shopping. For instance, Yee and Seong (2009) indicated that online Malaysian shoppers believe that the online medium can offer convenient purchases, affordable alternatives and discounts compared with shopping in traditional stores. Through this study, it is indicated that customers stress the importance of having reliable information on the website. This finding is in contrast to Moliner, Sánchez, Rodríguez, and Callarisa (2006) who found that accuracy of information was the most important element of delivering information in online businesses. In addition, customers prefer to use websites, which use a good color combination. This is supported by a previous study by Zeithaml, Parasuraman, and Malhotra (2002), who stated that customers would be attracted to a website, which has good graphics such as good color combinations. The results indicate that customers like to shop most at websites, which provide extra incentives to purchasing using discount offers and the collection of bonus points. This is supported by Chae et al. (2006), who listed incentive programs like cumulative discounts or rebates based on the purchase amount, saved shopping lists, point-and incentive-based premiums, and gift programs to enhance the experience of the online shopper. This result also shows that customers feel comfortable with websites, which provide many options of payment methods. This is consistent with the previous study by Xia et al. (2008) who found that convenient payment mechanisms raised the degree of satisfaction among customers. Besides, customers will use the website if they feel secure in doing transaction with it. This is in contrast to previous studies by Kim et al. (2006), Xia et al. (2008), and Carlson and O'Cass (2010) who found that websites that provide clearer and understandable terms and conditions have the highest item score for security and privacy factor. Meanwhile, in expected delivery service, the finding is in contrast to a previous study by Wang and Huarng (2002) who stated that customers would look out for websites or Blogspots, which offer a standard price for the delivery service. Similarly, online customers in China prefer shopping at websites that offer low delivery costs (Fu, Huang, & Wei, 2007). As expected, customers want their questions to be answered immediately and this has been proven by the highest factor loading of 0.800 in statistical value recorded for the item, "The business seems ready and willing to respond to customer needs." Likewise, online shoppers may be intolerant and, if there is no good response to their questions, this will lead to dissatisfaction with the website and they will turn to other competing websites (Flavian & Guinaliu, 2006). Among the eight factors in the Malaysian context, website design is the main factor that customers really look for and as such, it needs to be well maintained. This is due to the highest factor loading, which is 0.794 (Table 5).

Managerial Implication
From a managerial perspective, this study highlights the importance of providing good quality service, particularly in selling apparel online. According to Cronin and Tylor (1992), service providers should know how to set their objectives clearly: Whether to have customers satisfied with the business or to deliver the maximum level of perceived service quality. Therefore, this study helps marketing practitioners by providing models that enable them to better understand their consumers on what is the best e-service attraction for them, and how to react with them in the online environment. They can also increase their levels of awareness on eSQ, its antecedents, consequences, and mediators. As this study corresponds with previous studies, which demonstrate that service quality has a significant impact on behavioral intentions and satisfaction, managers should take this into consideration for maintaining and improving their competitive advantage. According to Olorunniwo, Hsu, and Udo (2006), managers should plan operation strategies that focus on the dimensions of service quality that boost customer satisfaction, which can in turn lead to positive behavioral intentions. In other words, if customers are satisfied with the service, they will become loyal, revisit, and recommend to others. Furthermore, managers can use this framework to create benchmarks against which they can evaluate their own website. By doing so, they can identify their competitive strengths and weaknesses from the customers' viewpoints regarding the eSQ at the time customers make their purchase (point-of-purchase). This study also suggests that online marketers should have sales personnel who are ready and willing to respond to customers' needs due to the absence of a sales person.

Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding
The author(s) received no financial support for the research and/or authorship of this article.