Impact of Service Quality On Customer Satisfaction in Malaysia Air Lines [PDF]

Zitiervorschau

Journal of Air Transport Management 67 (2018) 169–180

Contents lists available at ScienceDirect

Journal of Air Transport Management journal homepage: www.elsevier.com/locate/jairtraman

Impact of service quality on customer satisfaction in Malaysia airlines: A PLS-SEM approach

T

Muhammad Shoaib Farooqa,∗, Maimoona Salamb, Alain Fayollec, Norizan Jaafard, Kartinah Ayuppd a

Institute of Business and Management, University of Engineering and Technology, Lahore, Pakistan Faculty of Computer Science and Information Technology, Universiti Malaysia Sarawak, Malaysia c Department of Strategy and Organization, EMLYON Business School, Ecully, France d Faculty of Economics and Business, Universiti Malaysia Sarawak, Malaysia b

A R T I C L E I N F O

A B S T R A C T

Keywords: Quality Service quality Malaysia airlines Airline industry Customer satisfaction IPMA PLS-SEM AIRQUAL scale

This study is aimed to assess the quality of service provided by Malaysia Airlines and its impact on overall customer satisfaction. This study employed a convenience sampling method for collecting data from 460 respondents using a self-administered questionnaire, designed on five dimensions of AIRQUAL scale. Moreover, variance based structural equation modelling (PLS-SEM) was used for testing the proposed structural model. Findings of this study revealed that all five dimensions of AIRQUAL scale i.e. airline tangibles; terminal tangibles; personnel services; empathy and image have a positive, direct and significant impact on customer satisfaction of Malaysia Airlines. This study investigated the impact of service quality dimensions on customer satisfaction in Malaysia Airlines. Due to limited resources and time constraints this study involves respondents from Malaysia Airlines only; for that reason a comparative analysis of findings with other airlines was not possible; therefore it is considered a limitation of this study. Moreover, importance-performance map analysis (IPMA) was also performed for exploring the importance of various dimensions of service quality. Findings indicate that airlines should focus on all dimensions of service quality, with special focus on personnel services and image for enhancing their customer satisfaction. It is expected that findings of this study will help airlines to understand the role of various dimensions of service quality for enhancing their customer satisfaction.

1. Introduction

Airlines and Malaysian Airline System (MAS), which is now known as Malaysia Airlines (Clarke, 2004; O'Connell and Williams, 2005). As the nation's only flag-ship carrier Malaysia Airlines enjoyed the monopoly status in domestic market since 1974–2000. However, with the liberalization of Malaysian domestic airline industry in early 2000, Air Asia have changed the face of Malaysian air travel industry (Hankins, 2016; Radovic-Markovic et al., 2017). Resulting in an intense competition between its incumbent Malaysia Airlines and other budget airlines such as Air Asia and Malindo Air (Ong and Tan, 2010). Despite holding the recognition of being a leading airline in and to Asia from World Travel Awards (2010-11, 2013) these days Malaysia Airlines struggles to cut cost to compete with its rival low-cost airlines (Hankins, 2016; Ong and Tan, 2010; Radovic-Markovic et al., 2017). Since decades airline industry is facing challenges in terms of profitability and customer satisfaction all over the world (Forgas et al., 2012). However, Malaysia Airlines’ financial troubles started exacerbating in 2014, after the loss of flight MH-370 and flight MH-17. Both

With the rapid advancements in competitive business environment, customer expectations and demands are also increasing, leading to a situation where many companies – especially airlines – find it difficult to retain their customers (Ali et al., 2015). Moreover failure to recognize true needs and wants of customers is also a barrier to providing high quality services (Izogo and Ogba, 2015). Today's competitive market situation have forced airlines to focus on cost reduction for achieving efficient business operations; however while doing so the element of service quality and customer satisfaction is often compromised (Boetsch et al., 2011). Malaysia Airlines which was formerly known as Malaysian Airlines System (MAS) enjoys the prestige of being Malaysia's only national flag carrier. Company began as Malayan Airways Limited and flew its first commercial flight in 1947 (Zaid, 1995). However after Singapore's expulsion from Malaysia in 1972, airline's assets were divided in Singapore ∗

Corresponding author. E-mail addresses: [email protected] (M.S. Farooq), [email protected] (M. Salam), [email protected] (A. Fayolle), [email protected] (N. Jaafar), [email protected] (K. Ayupp). https://doi.org/10.1016/j.jairtraman.2017.12.008 Received 6 September 2017; Received in revised form 26 December 2017; Accepted 27 December 2017 0969-6997/ © 2017 Elsevier Ltd. All rights reserved.

Journal of Air Transport Management 67 (2018) 169–180

M.S. Farooq et al.

accidents were less than five months apart, and left a terrible impact on the overall goodwill of company, which led to its renationalization (Hankins, 2016; LeHardy and Moore, 2014). An unbiased analysis of current situation reveals a harsh fact that, after these two accidents many passengers still lack confidence in Malaysia Airlines in terms of its service quality, reliability and value for money (Smith and Marks, 2014). Better service quality is a critical factor which can be useful for distinguishing and improving organization's performance in the era of intense competition (Namukasa, 2013; Ong and Tan, 2010). Pertaining to the subjective nature of service quality, its dimensions and measurement issues have been investigated by various recent studies (e.g. Farooq et al., 2009; Izogo and Ogba, 2015; Shabbir et al., 2016). Conceptual and empirical relationship between service quality and customer satisfaction have received substantial attention from researchers, turning it into one of the core marketing instruments (Gustafsson et al., 1999; Zaid, 1995). Although measurement of service quality has received a great deal of attention, yet service quality of airline industry is still unexplored and requires a thorough investigation (Ali et al., 2015; Farooq et al., 2017; Park et al., 2005). Unique nature of airline service industry, which is substantially different from other service industries, warrants further studies to explore the service quality of Malaysia Airlines and its impact on overall customer satisfaction (Farooq et al., 2017; Radovic-Markovic et al., 2017). Therefore, this study is aimed to assess the perceptions of Malaysia Airlines' passengers, regarding the service quality and resulting customer satisfaction using the AIRQUAL scale. For the ease of readers this paper is organized in five sections, starting with the introduction and background of this study. Section two provides a precise yet comprehensive insight of literature on service quality and customer satisfaction, along with its application in various airlines and description of scales used to measure service quality in airline industry. Next section three explains the research methodology and section four elaborates the survey results and findings of this study. Last section concludes this study followed by discussion of practical implications and limitations of this study.

(8) security, (9) competence and (10) courtesy. Same model was latter simplified and named as SERVQUAL by Parasuraman et al. (1988) reducing it to five dimensions i.e. (1) tangibles, (2) reliability, (3) responsiveness, (4) assurance and (5) empathy. The SERVQUAL scale has been widely recognized by academicians, researchers and practitioners in various fields and different countries (Butt and Run, 2010; Farooq, 2016; Lee-Ross, 2008). SERVQUAL offers a comprehensive measurement scale with practical implications for customers' perceived service quality (Parasuraman et al., 1994). It is worth mentioning that, although SERVQUAL has been widely accepted and adopted by various scholars (e.g. Gilbert and Wong, 2003; Lee-Ross, 2008; Samen et al., 2013); yet it has also faced criticism by some scholars (e.g. Buttle, 1996; Cronin and Taylor, 1992; Robledo, 2001) because it only involves comparison of perceived quality of service received and customers’ expected service quality. In this regard Wu and Ko (2013) assert that SERVQUAL offers some general guidelines for assessment of service quality by incorporating its few dimensions and contexts; however service quality dimensions are ought to be scrutinized and examined discretely for incorporating various industry-specific issues. Moreover Park et al. (2005) argue that particular industry-specific operations and issues which are exclusive for airline industry (e.g. online ticketing, check-in, luggage allowance, boarding service and on-board facilities) distinguish airline industry from those of other service oriented industries. Various scholars (e.g. Chang and Yeh, 2002; Cunningham et al., 2004; Farooq et al., 2017; Namukasa, 2013; Radovic-Markovic et al., 2017; Wu and Cheng, 2013) have suggested that customers’ expectations in the field of airline industry are formed at the “moment-of-truth” by interacting with the reservation department, telephonic communication, ticketing experience, baggage handling system, flight schedule and service of cabin crew members. Therefore Park et al. (2005) assert that only five dimensions of SERVQUAL scale are not suitable for measuring all dimensions of service quality in airline industry as they do not involve industry specific (i.e. airline industry) aspects of service quality. Due to massive criticism on the application of SERVQUAL scale, various scholars have used and recommended another service quality measurement scale, which is developed by Cronin and Taylor (1992) and is named as SERVPERF. According to Cronin and Taylor (1994) SERVPERF scale is mainly designed to focus on customers' perceptions about performance of service providers, to assess the actual service quality received. Although some researchers have used this scale for assessing service quality in airline industry; yet there are number of critiques reporting its inability to capture all dimensions of airlines' service quality (Ali et al., 2015; Farooq et al., 2017; Ostrowski et al., 1993). Moreover some scholars (e.g. Cunningham et al., 2004) have also criticised the generic nature of SERVPERF and argue that, too much generic nature of this scale makes it difficult to capture industryspecific dimensions of airline industry, which is crucial for understanding passenger's perception of service quality. Therefore, various scholars have proposed different models for exploring the dimensions of service quality with specific reference to airline industry (e.g. Chang and Yeh, 2002; Gourdin, 1988; Ostrowski et al., 1993; Truitt and Haynes, 1994). One of the models presented by Gourdin (1988) describes airline service quality with three distinct dimensions; i.e. price, safety, and timeliness of flights. Likewise, airline service quality model presented by Ostrowski et al. (1993) involves comfort of seats, food, and timeliness of flights. Whereas Truitt and Haynes (1994) suggested to use cleanliness of seats, check-in process, timeliness of flights, food and beverages, and customer complaints handling system, as dimensions of airline service quality. However Chang and Yeh (2002) suggested a revised version of five dimensions of service quality presented by Parasuraman et al. (1988) which include tangibility, responsiveness, reliability, empathy and assurance. Further Park et al. (2005) also analysed airline service quality by involving only few dimensions of service quality, which are reliability of customer service, convenience of accessibility, and quality of in-flight services.

2. Literature review 2.1. Service quality According to Parasuraman et al. (1988, p. 13) service quality refers to the “function of [the] difference between [the] service expected and [the] customer's perceptions of the actual service delivered”. In recent past service quality have received an intense attention from researchers in the field of service marketing and business development (e.g. Aagja and Garg, 2010; Farooq et al., 2009; Qin et al., 2010; Samen et al., 2013; Shabbir et al., 2016). Moreover, a considerable attention has been given to its conceptualization and measurement scales as well (Akter et al., 2013; Cristobal et al., 2007; Farooq et al., 2009). Specifically, element of service quality has been extensively explored in various industries such as mobile banking, health management, telecommunication, online education, hoteling and tourism etc. (Abdullah et al., 2011; Farooq et al., 2017; Izogo and Ogba, 2015; Samen et al., 2013). According to Tsoukatos and Mastrojianni (2010) customers compare actual service delivery with their own expectations, which are shaped by their prior experience, memories and/or word of mouth. This comparison helps to determine customers' perceived service quality (Parasuraman et al., 1988). Moreover in this regard Zeithaml et al. (1996) assert that better understanding of customers' perceived service quality is significantly important for enhancing customer satisfaction by delivering quality services. In order to measure service quality, Parasuraman et al. (1985) proposed a comprehensive model comprising ten dimensions of service quality i.e. (1) tangibles, (2) reliability, (3) responsiveness, (4) understanding the customers, (5) access, (6) communication, (7) credibility, 170

Journal of Air Transport Management 67 (2018) 169–180

M.S. Farooq et al.

quality of Malaysia Airlines. Here is a brief description of all five dimensions involved in AIRQUAL scale:

Moreover, a recent study on Ugandan airline industry by Namukasa (2013) categorized airline service quality in three areas, which include pre-flight service quality, in-flight service quality, and post-flight service quality. In order to measure pre-flight service quality, he used responsiveness and accessibility of discounts; further in-flight service quality was measured using courtesy of staff, tangibles and luggage handling; whereas post-flight service quality was measured using timeliness of flight and frequent flyers schemes (Namukasa, 2013). Findings of his study confirmed that all three areas i.e. pre-flight service quality, in-flight service quality and post-flight service quality are equally important and have significant positive effect on customer satisfaction in Ugandan airline industry (Namukasa, 2013). In another recent study Wu and Cheng (2013) classified airline service quality into four main dimensions i.e. physical environment quality, interaction quality, access quality and outcomes quality. These four dimensions were further divided in eleven sub-dimensions which include, cleanliness, problem solving skills, overall conduct, staff expertise, comfort, safety and security, tangibles, waiting time, convenience, valence and access to information (Wu and Cheng, 2013). Findings of their study revealed that conceptual and theoretical understanding of airline service quality and customer satisfaction is still at infancy stage (Farooq, 2016; Wu and Cheng, 2013). Moreover, inconsistency of measurement scales, diverse analytical methodologies and different dimensions used for assessing service quality of airline industry, have made it difficult to perform a cross-study analysis for drawing a meaningful conclusion. An overview of main dimensions of service quality, which are most frequently used by previous studies for assessing service quality of airline industry, is summarized in Table 1. In order to resolve these inconsistencies Ekiz et al. (2006) presented a comprehensive model, named AIRQUAL for assessing airline service quality. This AIRQUAL model comprises of five dimensions, i.e. airline tangibles, terminal tangibles, personnel services, empathy and image. Later another study by Nadiri et al. (2008) also validated AIRQUAL scale by using it to assess impact of airline service quality on customer loyalty of North Cyprus Airlines; followed by Ali et al. (2015) who used this scale for assessing service quality in Pakistan International Airlines (PIA). However, they called for further studies to explore different contexts of customer satisfaction and airline service quality, so that AIRQUAL scale can be generalized and validated in a broader context. Therefore, this study adapts AIRQUAL scale for analysing service

2.2. Airline tangibles Airline tangibles refers to the tangible cues which are associated with service quality of an airline (Ekiz et al., 2006). It is considered as one of the most important dimensions of service quality in airline industry (Farooq, 2016; Gudmundsson, 1998). According to Ali et al. (2015) airline tangibles refers to the overall condition of aircrafts; which involves the quality of interior and exterior equipments; quality of catering service; comfortable seats and cleanliness. 2.3. Terminal tangibles According to Ariffin and Yahaya (2013) terminal tangibles is one of the most visible indicators of an airline's service quality. According to them, terminal tangibles have a direct influence on airline's overall image building. Terminal tangibles implicate the quality of services which are available at terminal (Ekiz et al., 2006). These services comprise of effective sign boards; friendly security and control system, good air-conditioning system on terminal; clean toilets and information counters for guiding passengers (Ali et al., 2015; Wu and Cheng, 2013). 2.4. Personnel services Personnel services is an important dimension of airline's service quality (Nadiri et al., 2008). It refers to the quality of service provided by airline's staff (i.e. their attitude and behaviour towards customer service) and flight attendants (Boetsch et al., 2011; Ekiz et al., 2006). Moreover, personnel services also encompass an error free ticketing service; responsiveness of aircraft crew members; personal care and helping attitude (Namukasa, 2013). 2.5. Empathy Benefits of emotional intelligence and empathy are an open secret; therefore empathy is considered as an integral part of service quality in any business (Humphrey, 2013; Radovic-Markovic et al., 2017). Various studies have reported that empathy have a direct effect on repeat

Table 1 Airline service quality dimensions. Sr. No.

Author(s) and Year

Dimensions of Service Quality

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Li et al. (2017) Kos Koklic et al. (2017) Sandada and Matibiri (2016) Ali et al. (2015) Leong et al. (2015) Suki (2014) Archana and Subha (2012) Boetsch et al. (2011) Saha and Theingi (2009) An and Noh (2009) Teichert et al. (2008) Babbar and Koufteros (2008) Aydin and Pakdil (2008) Tiernan et al. (2008) Nadiri et al. (2008) Park (2007)

17 18

Liou and Tzeng (2007) Shaw (2007)

19 20 21 22

Ekiz et al. (2006) Park et al. (2005) Kozak et al. (2003) Tsaur et al. (2002)

Tangibles, reliability, responsiveness, assurance, empathy Airline tangibles, quality of personnel Reliability and customer service, convenience and accessibility, inflight service Airline tangibles, terminal tangibles, personnel quality, empathy, image, customer satisfaction Tangibles, reliability, responsiveness, assurance, empathy Terminal tangible, empathy, word of mouth, airline tangibles In-flight service, in-flight digital services, airline back office services and operations Airline brand, price, sleep comfort Tangibles, schedule, flight attendants, ground staff Responsiveness, empathy, food quality, alcoholic beverage, non-alcoholic beverage, and reliability Flight schedule, total fare, frequent flyer package, flexibility, catering, punctuality, ground services Level of concern and civility, listening and understanding, individual attention, cheerfulness, courtesy, friendliness Tangibles, reliability, responsiveness, assurance, empathy On-time performance, overbooking, mishandled baggage, customer complaints Airline tangibles, terminal tangibles, empathy, personnel In-flight service, reservation-related service, airport service, reliability, employee service, flight availability, perceived price, passenger satisfaction, perceived value, airline image Employees' service, safety and reliability, on-board service, on-time performance, schedule, frequent flyer package Frequency and timing, punctuality, airport location and access, ticketing flexibility/seat accessibility, frequent flyer benefits, in-flight services, airport services Airline tangibles, terminal tangibles, image, personnel services, empathy Reliability and customer service, in-flight service, convenience and accessibility Terminal tangibles, reliability, responsiveness, empathy Safety, courtesy of staff, seat comfort

171

Journal of Air Transport Management 67 (2018) 169–180

M.S. Farooq et al.

2.8. Service quality and customer satisfaction

buying behaviour (e.g. Chang and Yeh, 2002; Ekiz et al., 2006; Farooq et al., 2009; Humphrey, 2013). In airline industry empathy is reflected by hassle and problem free service; which involves careful luggage handling, courteous ticketing service and thoughtful compensation plans in case of any loss or hazards (Ekiz et al., 2006; Farooq, 2016).

Service quality have been considered a strong antecedent and predictor of customer satisfaction (McDougall and Levesque, 2000). A recent study by Saha and Theingi (2009) investigated the relationship between airline service quality and customer satisfaction; findings of their study revealed a strong connection between perceived service quality and customer satisfaction. A satisfied customer is most likely to remain loyal to airline company; whereas an unsatisfied customer have a likelihood of switching to any other service provider (Ali et al., 2015; Archana and Subha, 2012; Gudmundsson and Lechner, 2006). It is pertinent to mention that, despite a general consensus on the basic definition of the concept of perceived service quality and customer satisfaction; the causal relationship between both constructs have remained controversial (Al-alak, 2014; Bansal et al., 2004). Various scholars (e.g. Cronin and Taylor, 1992; Oliver, 1997; Parasuraman et al., 1988) have suggested perceived service quality to be the antecedent of customer satisfaction; whereas others (e.g. Andreassen and Lindestad, 1998; Bitner, 1990; Bolton and Drew, 1991) consider customer satisfaction as an antecedent of perceived service quality. In order to resolve this discrepancy Han et al. (2008) investigated the role of perceived service quality as an antecedent of customer satisfaction in a number of different industries, including banking sector, hospitals, information technology, education sector, beauty salons and airline companies. A recent study by Ali et al. (2015) have also used similar notion for investigating customer satisfaction and perceived service quality in Pakistan International Airlines. Therefore, this study also embraces first school of thought and hypothesises that, perceived service quality of airline companies will have a significant effect on its customer satisfaction. As mentioned earlier this study employs AIRQUAL scale, which was developed by Ekiz et al. (2006) to overcome the shortcomings of other existing service quality scales in terms of airlines industry. This AIRQUAL scale encompasses five dimensions i.e. airline tangibles, terminal tangibles, personnel services, empathy and image, which are also posed in Fig. 1 showing proposed research framework of this study. Developing on the base of logical relationships derived from aforementioned literature review; which provided support for perceived service quality being an antecedent of customer satisfaction in airlines industry, this study proposes following five hypotheses:

2.6. Image Life cycle of airlines is not different from other general firms (Gudmundsson, 1998). Therefore, airlines also have to be very conscious about their image, goodwill and brand recognition value (Nadiri et al., 2008; Radovic-Markovic et al., 2017). In order to maintain their good image, airlines have to bring various promotional offers and frequent flyer programs (Gudmundsson et al., 2002; Radovic-Markovic et al., 2017). According to Ekiz et al. (2006) an airline's image comprises of its overall perception; value for money; promotional offers and goodwill. 2.7. Customer satisfaction According to Kotler and Caslione (2009) satisfaction refers “to a person's feeling of pleasure or disappointment resulting from comparing a product's performance in relation to his or her expectations”. Customer satisfaction have remained a key focus area in many social and behavioural studies (e.g. Chen et al., 2012; Farooq et al., 2009, 2010). The concept of customer satisfaction is generally based on the notion that a business must satisfy its customers in order to be sustainable and profitable (Farooq, 2016; Izogo and Ogba, 2015; Radovic-Markovic et al., 2017). According to Westbrook and Oliver (1991) customer satisfaction is defined as positive post-consumption feelings. Instead of exploring cognitive outcomes, customer satisfaction is considered to be an effective measure of usefulness of a product or service availed by customers (Berezina et al., 2012). Moreover, various scholars (e.g. Cronin and Taylor, 1992; Farooq et al., 2009; Seth et al., 2005) have agreed that re-purchase intention is strongly associated with customer satisfaction. Likewise customer satisfaction serves as a major element for customer retention (Abdullah et al., 2011; Farooq et al., 2016). Further, Park et al. (2005) note that customer satisfaction leads to a positive and favourable word-of-mouth; which is widely acknowledged as a critical source of indirect marketing for brand building. A number of studies (e.g. Farooq and Radovic-Markovic, 2017a; Jun et al., 2004; Prayag, 2007; Shin and Elliott, 2001) have reported that satisfied customers can improve the profitability of organizations, by helping them to expand their business through new referral customers and repeat business from existing customers. Pertaining to the complex nature of human behaviour and perceptions, the phenomenon of customer satisfaction has remained underresearched in various industries (e.g. Ali et al., 2015; Farooq et al., 2009; Izogo and Ogba, 2015; Shabbir et al., 2016). Moreover, due to the subjective nature of customer satisfaction, complete understanding of its determinants have remained somewhat elusive (Farooq et al., 2017; Gudmundsson et al., 2002; Qin et al., 2010). Specifically, for businesses which are operating in service sector it is far more challenging to achieve and maintain customer satisfaction (Li et al., 2017). For instance, nature of some services is multi-layered and extremely complex; due to multiple service encounters involved in the whole process (Farooq and Radovic-Markovic, 2016; Han and Ryu, 2012). According to Archana and Subha (2012) customer satisfaction in airline industry is influenced by multi-dimensional service quality, which involves preflight services, in-flight services, baggage handling, in-flight digital services, and post-flight services. Likewise another recent study by Ali et al. (2015) also reported a positive relationship between customer satisfaction, loyalty and repeat purchase intentions. Therefore in airline industry customer satisfaction is a very critical element, for ensuring a sustainable business and long term relationship with customers (Ali et al., 2015; Archana and Subha, 2012; Wu and Cheng, 2013).

H1. Perceived quality of airline tangibles have positive and significant direct effect on customer satisfaction

Airline Tangibles H1 Terminal Tangibles Personnel Services

H2 H3 H4

Empathy H5 Airline Image Fig. 1. Conceptual model.

172

Customer Satisfaction

Journal of Air Transport Management 67 (2018) 169–180

M.S. Farooq et al.

H2. Perceived quality of terminal tangibles have positive and significant direct effect on customer satisfaction

ensuring quality and rigour of any study. For this purpose, Hair et al. (2017, p. 20) have suggested the use of 10 times rule, which was proposed by Barclay et al. (1995) for determining minimum sample size in a PLS-SEM analysis. This rule states that minimum sample should be “10 times the largest number of structural paths directed at a particular construct in structural model”. Structural model of this study involves six constructs (i.e. five independent and one dependent variable) and according to this 10 times rule criterion, our minimum sample size should be 50 respondents. However, we have adopted a more rigorous criterion proposed by Westland (2010). Moreover, sample size of this study was determined on the base of review of previous related studies and suggestions of different scholars (e.g. Ali et al., 2015; An and Noh, 2009; Archana and Subha, 2012; Farooq and Radovic-Markovic, 2017b). A self-administered survey questionnaire was used for data collection. Using a convenience sampling approach, 750 questionnaires were distributed in waiting lounges of Kuala Lumpur International Airport and Kuching International Airport. A total of 460 responses were received back, which indicates an overall response rate of 61.33%. This response rate is very close to a recent study by Ali et al. (2015), who reported 58% response rate while investigating service quality of Pakistan International Airlines.

H3. Perceived quality of personnel services have a positive and significant direct effect on customer satisfaction H4. Perceived empathy have a positive and significant direct effect on customer satisfaction H5. Perceived airline image have a positive and significant direct effect on customer satisfaction 3. Research methodology 3.1. Research instrument/operationalization of constructs A survey instrument was adopted from Westbrook and Oliver (1991) and Ekiz et al. (2006) for data collection from passengers of Malaysia Airlines. Final questionnaire comprised of total 33 items, out of which six items belonged to airline tangibles (AT) i.e. (1) aircrafts were equipped with latest and modern technology (2) quality of catering service was good (3) quality of air-conditioning in the planes was good (4) interior of aircraft was well maintained (5) seats were clean and comfortable (6) cleanliness of the plane toilets was well maintained. Seven items belonged to terminal tangibles (TT) i.e. (1) quality of air-conditioning on airport terminals (2) information counter was readily available to assist me (3) number of shops were adequate for my needs (4) airport had effective sign boards (5) security and control system was friendly and reliable (6) adequate number of trolleys were available on airport (7) cleanliness of the airport toilets was well maintained. Further, seven items belonged to personnel services (PS) i.e. (1) airline staff was well dressed (2) workers were well aware of their duties (3) ticketing and reservation service was error free (4) flight attendants were well behaved and had a good attitude (5) whenever I required airline personnel answered my questions (6) flight attendants were providing equal personal care to everyone (7) workers were willing to extend their help to everyone on plane. Next six items belonged to empathy (E) i.e. (1) departures and arrivals are usually on time (2) transportation between city and airport is not a problem (3) in case of loss or hazard compensation plans are clearly communicated (4) my luggage is handled very carefully (5) ticketing office staff is very courteous (6) number of flights is enough to satisfy passengers’ demands. Three items belonged to image (I) i.e. (1) availability of seats and promotional offers are very much appealing to me (2) ticket prices are worth the services I received (3) Malaysia Airlines bears a good brand image. And last four items belonged to customer satisfaction (CS) i.e. (1) i am happy for my decision to choose Malaysia Airlines (2) my choice of Malaysia Airlines was a wise decision (3) i did the right thing to choose Malaysia Airlines as a service provider (4) i am satisfied, and my experience with Malaysia Airlines was very enjoyable. A seven point Likert-type scale was used to enhance the redundancy and sanctity of this study, as advised by (Farooq, 2016). Moreover, in order to validate the questionnaire, a pilot study was also conducted, which involved 50 respondents who had recently travelled with Malaysia Airlines. Although some minor changes were made in the sentence structure of final questionnaire, but overall findings of pilot study established the reliability and validity of questionnaire used for data collection.

3.3. Analytical methods Data was analysed using IBM SPSS Statistics version 24.0 and SmartPLS version 3.2.7 (Ringle et al., 2017). Variance based PLS-SEM approach was adopted, because it can handle both types of measurement models (i.e. reflective and formative models) which are involved in the proposed model of this study. However, CB-SEM/AMOS can typically handle only reflective models. In a similar recent study Farooq et al. (2017) have also used PLS-SEM for validating UTAUT3 model (i.e. extended version of unified theory of acceptance and use of technology). Moreover, this choice of PLS-SEM was made on the base of its ability to estimate causal relationships among all latent constructs simultaneously, while dealing with measurement errors in the structural model (Farooq, 2016; Hair et al., 2017). Furthermore, our study is explanatory in nature; therefore, PLS-SEM is a best fit for this study (Farooq and Radovic-Markovic, 2017b). Considering the guidelines suggested by Hair et al. (2017) measurement models were evaluated separately before the evaluation of structural model. Furthermore, in order to ascertain the data quality and consistency of structural model, several tests (e.g. common-method variance bias test, non-response bias test and data screening for missing values etc.) were also performed along with other validity and reliability checks, before performing PLSSEM analysis.

4. Data analysis 4.1. Common-method variance bias test In order to determine the possible presence of common-method variance bias among variables, this study employs Harman (1976) onefactor test. Researchers observed the guidelines and approach of Podsakoff et al. (2003) for conducting Harman (1976) one-factor test. For this purpose, all items of measurement scale were entered into a principal component analysis with varimax rotation, so that any signs of single factor could be identified from factor analysis. The results extracted six different factors from 33 items of measurement constructs (i.e. Airline Tangibles; Terminal Tangibles; Personnel Services; Empathy; Image and Customer Satisfaction) and rotation converged in 7 iterations. On the base of these results, it is determined that this study do not have any problem of common-method variance bias.

3.2. Sample design and data collection This study is aimed to investigate the role of service quality in determining customer satisfaction of Malaysia Airlines. In order to achieve this objective, the target population for this study was identified as all passengers who have travelled with Malaysia Airlines in last three months. Determining a right sample size is very crucial for 173

Journal of Air Transport Management 67 (2018) 169–180

M.S. Farooq et al.

4.2. Non-response bias test

Table 2 Passengers’ demographic Attributes.

This study employs extrapolation method, for testing non-response bias. Extrapolation method is most commonly used technique which involves comparison of early and late respondents for possible difference in demographics and mean values of other key constructs (Armstrong and Overton, 1977). For this purpose, an independent sample t-test was performed for comparing the responses of first 50 and last 50 questionnaires. Findings of independent sample t-test revealed that there was no significant 0.05 level difference in the mean values of both groups (i.e. first 50 respondents vs last 50 respondents). Thus, on the base of the findings of independent sample t-test, it was concluded that there was no substantial difference in the responses of both groups; and hence, non-response bias is not a problem for this study.

Attributes

Distribution

Frequency

%

Gender

Male Female Under 20 years 21-30 years 31-40 years 41-50 years Above 50 years Malaysian Indian Chinese Others School High school Bachelor's degree Master's degree Other Vacations Business Education Visiting friends and family Medical

212 248 37 216 129 46 32 147 80 137 96 28 110 166 129 27 248 55 78 69 10

46% 54% 8% 47% 28% 10% 7% 32% 17% 30% 21% 6% 24% 36% 28% 6% 54% 12% 17% 15% 2%

Age

Ethnic Background

Education level

4.3. Data screening and pre-analysis As part of preparation for data analysis, a thorough screening process was conducted. Data was tested for any possible statistical error of normality, outliers, missing values and demographic characteristics. Although there were very few missing values, however missing values were handled using widely recommended approach of mean replacement. This option is a built-in feature of SmartPLS, and replaces missing data points with the mean values of all data points of the same indicator (Hair et al., 2017). One of the most sought after benefit of mean replacement method is that; this approach does not alter our sample size (unlike list-wise deletion and pair-wise deletion); and at the same time mean values of all the variables remain unchanged (Hair et al., 2017; Radovic-Markovic et al., 2017). Further, data analysis and discussion of research findings begins with the brief description of demographic attributes of respondents in terms of their age, gender and education level; moreover, nature of their travel plan is also categorized by asking their purpose of visit. Out of total 460 respondents 46% were male, whereas 54% were female. Majority of the respondents (almost 47%) were aged between 21 and 30 years. Ethnic background information revealed that, majority of respondents (nearly 32%) were Malaysians, followed by 30% Chinese and only 17% were from Indian ethnicity. Furthermore, 36% of the respondents had a bachelor degree, 28% had a master's degree and 24% had a high school certificate. Regarding the nature of their trip, 17% respondents mentioned that they were traveling for educational purpose and 54% were traveling for leisure and tourism purposes. Complete details about the respondents' demographic attributes are listed in Table 2.

Purpose of Visit

discriminant validity. This section is aimed to discuss the evaluation of measurement models (outer models) starting with the assessment of reflective measurement models. 4.5. Analysis of reflective measurement models Considering the guidelines of Hair et al. (2017) and Henseler et al. (2009) constructs with reflective measurement models (i.e. Airline Tangibles; Terminal Tangibles; Personnel Services; Empathy and Image) were separately analysed. In order to evaluate the reflective measurement models, all constructs were assessed for their reliability and validity. Results revealed that, all constructs have a fairly acceptable factor loading value, ranging between 0.70 and 0.90. Further all constructs were assessed for their composite reliability (CR) and Cronbach's alpha values, which were higher than 0.70 critical level suggested by Cohen (1988). Average variance extracted (AVE) value of all constructs was also higher than the critical value of 0.50 suggested by Hair et al. (2017). Complete results of validity and reliability of all constructs are presented in Table 3. Moreover, Fornell-Larcker criterion was used to assess the discriminant validity, which is presented in Table 4. Bold values in Table 4 show the square-root of AVE, which is higher than the estimated correlation values, thus demonstrating the discriminant validity of constructs involved in the proposed measurement models (Farooq et al., 2016; Hair et al., 2017). Overall, these results satisfy all requirements for establishing the validity and reliability of reflective measurement models. Moreover, HTMT ratio of correlations was also calculated, which is suggested by Henseler et al. (2015) as a modern tool for analysing discriminant validity of constructs involved in measurement models. As a rule of thumb, an HTMT value greater than 0.85 indicates a potential problem of discriminant validity (Hair et al., 2017). For this study, all

4.4. Analysis of measurement model Conceptual model of this study involves both types of measurement models, i.e. formative measurement models as well as reflective measurement models. Out of six total variables, one variable (i.e. Customer Satisfaction) has a formative measurement model and five variables (i.e. Airline Tangibles; Terminal Tangibles; Personnel Services; Empathy and Image) have reflective measurement models. Statistical evaluation criteria for reflective measurement models is different from formative measurement models (Hair et al., 2017). In case of formative measurement models, concept of internal consistency is inappropriate (Chin, 1998), because items of formative measurement scale are likely to represent an independent cause and are not necessarily highly correlated with each other (Hair et al., 2017). Whereas items of reflective measurement models need to be correlated, and they should depict significant outer loading values (Hair et al., 2017). For the purpose of this study, both reflective and formative measurement models were evaluated separately. Considering the guidelines of Hair et al. (2017) all reflective measurement models were analysed for reliability and validity of constructs; whereas formative measurement model (i.e. Customer Satisfaction) was analysed for its convergent validity and

Table 3 Validity and reliability of latent constructs.

174

Latent Constructs

AVE

Composite Reliability

Cronbach's Alpha

Airline Tangibles (AT) Terminal Tangibles (TT) Personnel Services (PS) Empathy (E) Image (I)

0.589 0.647 0.638 0.687 0.63

0.907 0.941 0.938 0.942 0.843

0.865 0.883 0.901 0.851 0.799

Journal of Air Transport Management 67 (2018) 169–180

M.S. Farooq et al.

are likely to represent an independent cause for underlying latent construct, thus formative indicators do not have high correlation among measurement scale items. Moreover, method of convergent validity calculation is also different for formative measurement models (Chin, 1998; Hair et al., 2017). As mentioned in the previous section, this study involves one formative measurement model (i.e. Customer Satisfaction). In order to establish convergent validity the magnitude of path coefficient (correlation) between formative constructs i.e. CSformative → CSreflective was assessed. As a rule of thumb correlation value between Yformative → Yreflective should be 0.80 or higher for determining convergent validity of formative constructs (Chin, 1998; Hair et al., 2017). Results demonstrate that path coefficient values between CSformative → CSreflective is higher than the threshold of 0.80, which fulfils the criteria described by (Chin, 1998). Thus, it is established that our formative measurement model (i.e. Customer Satisfaction) have a sufficient degree of convergent validity. Further, outer weights (relative importance) of formative indicators were also assessed for establishing the relative importance of indicators for their underlying latent construct. A complete list of outer weights of all items involved in the measurement of formative model of customer satisfaction is provided in Table 6. Considering the guidelines of (Hair et al., 2017; Henseler et al., 2016) these outer weight values were assessed for their significance also. Findings depict that all indicators of formative measurement model have significant and positive outer weight values. It proves that all indicators of formative measurement model have also met the criteria for establishing their relevance and significance. On the base of above discussion, suitability of formative constructs is established and overall assessment of reflective and formative measurement models demonstrates acceptable results to proceed with the evaluation of structural model. Hereafter discussion continues with the assessment of structural model (inner model) in the next section.

Table 4 Discriminant validity (Fornell-Larcker criterion). Latent Constructs

1

2

3

4

5

Airline Tangibles (AT) Terminal Tangibles (TT) Personnel Services (PS) Empathy (E) Image (I)

0.767 0.612 0.678 0.679 0.620

0.804 0.787 0.655 0.423

0.798 0.629 0.420

0.828 0.614

0.793

∗

Values in the bold are Square root of AVE.

Table 5 Cross loadings among reflective measurement scale items. Items

AT

TT

PS

E

I

AT-1 AT-2 AT-3 AT-4 AT-5 AT-6 TT-1 TT-2 TT-3 TT-4 TT-5 TT-6 TT-7 PS-1 PS-2 PS-3 PS-4 PS-5 PS-6 PS-7 E-1 E-2 E-3 E-4 E-5 E-6 I-1 I-2 I-3

0.791 0.760 0.729 0.745 0.841 0.771 0.299 0.274 0.465 0.385 0.156 0.188 0.255 0.250 0.244 0.238 0.232 0.226 0.221 0.215 0.209 0.465 0.385 0.191 0.186 0.180 0.174 0.465 0.385

0.120 0.281 0.129 0.123 0.117 0.373 0.753 0.832 0.821 0.800 0.793 0.850 0.826 0.232 0.226 0.220 0.476 0.396 0.202 0.197 0.191 0.185 0.476 0.396 0.167 0.162 0.476 0.396 0.217

0.310 0.285 0.476 0.396 0.167 0.199 0.266 0.261 0.255 0.249 0.243 0.237 0.232 0.760 0.745 0.851 0.809 0.839 0.770 0.859 0.189 0.221 0.288 0.283 0.277 0.271 0.265 0.259 0.254

0.373 0.293 0.207 0.182 0.373 0.293 0.064 0.096 0.163 0.158 0.152 0.146 0.140 0.134 0.099 0.094 0.088 0.082 0.373 0.343 0.775 0.762 0.891 0.844 0.901 0.835 0.418 0.224 0.219

0.184 0.179 0.173 0.167 0.423 0.343 0.149 0.144 0.138 0.132 0.423 0.343 0.114 0.109 0.423 0.423 0.343 0.114 0.146 0.213 0.208 0.202 0.196 0.190 0.142 0.257 0.739 0.799 0.859

4.7. Analysis of structural model Structural model was assessed for overall explanatory power of constructs through R2 value, predictive relevance through Q2 value and path coefficient β-values. Findings of structural model are presented in Fig. 2. These results indicate that proposed model have 52.1% of explanatory power for customer satisfaction with R2 = 0.521. Moreover, it is found that relationship between airline tangibles and customer satisfaction (β = 0.548; t-value = 4.842; p = .000) is positive and significant, providing support for H1. Similarly, H2 which is relationship between terminal tangibles and customer satisfaction (β = 0.442; tvalue = 3.251; p = .000) is also supported. Likewise, proposed relation between personnel services and customer satisfaction (β = 0.606; tvalue = 5.754; p = .000) is also significant, thus H3 is supported. Further, a strong relationship of (β = 0.485; t-value = 3.788; p = .000) between empathy and customer satisfaction provides support for H4. Lastly, findings of SEM analysis support H5 indicating a strong and positive direct relationship (β = 0.626; t-value = 5.891; p = .000) between image and customer satisfaction. A summarized overview of these findings is presented in Table 7.

Note: Bold values are loadings for items, which are above the recommended value of 0.5.

HTMT values were well below the threshold level of 0.85, which indicates that there is no issue of discriminant validity. Another test for discriminant validity of reflective measurement models was performed by evaluating all cross-loading values of reflective constructs’ indicators. As a rule of thumb, indicators of reflective measurement models should have highest loading on their own underlying latent construct, as compared to other constructs involved in the structural model (Farooq et al., 2017; Hair et al., 2017). Complete list of crossloading values of all indicators involved in the constructs of reflective measurement models is presented in Table 5. As per the findings presented in Table 5 all indicators (measurement scale items) of reflective measurement models have a higher loading on their respective underlying latent construct, as compared to loading on any other construct involved in the model. Hence, these findings meet the cross loadings evaluation criteria and provide a satisfactory evidence for discriminant validity of the reflective measurement models. Now, discussion continues with the assessment of formative measurement model (i.e. Customer Satisfaction) involved in this study.

Table 6 Outer weights of items involved in formative constructs. Outer Weights

CS_01 - > CS CS _02 - > CS CS _03 - > CS CS _04 - > CS

4.6. Analysis of formative measurement models Evaluation process of formative constructs is different from reflective constructs (Chin, 2010; Hair et al., 2017; Henseler et al., 2009). The logic behind this notion is that, all formative measurement models

0.4355 0.2620 0.3276 0.2811

** ** ** **

T Statistics (|O/STDEV|)

P Values

7.9126 4.5213 3.4251 3.1092

.0000 .0000 .0004 .0020

Note: *P < .05, **P < .01, CS = Customer Satisfaction.

175

Journal of Air Transport Management 67 (2018) 169–180

M.S. Farooq et al.

model. As a rule of thumb, if a Q2 value is larger than zero, it suggests that latent exogenous constructs involved in the structural model possess predictive relevance for latent endogenous constructs (Chin, 2010; Hair et al., 2017). The Q2 value of our model is 0.386; which supports the underlying assumption of this study, that the endogenous construct (i.e. Customer Satisfaction) involved in this study have strong predictive relevance. Moreover, every construct was assessed for possible collinearity issue. Findings revealed that collinearity is not an issue for our study. Hence, overall predictive relevance for our proposed structural model is achieved. Now discussion continues with the analysis of importance-performance map analysis (IPMA); followed by assessment of Goodness of Fit (GoF) value in next section.

Airline Tangibles 0.548* Terminal Tangibles Personnel Services

0.442* 0.606*

R2 = 0.521 Customer Satisfaction

0.485* Empathy

4.8. Importance-performance map analysis (IPMA)

0.626* Importance-performance map analysis (IPMA; also known as importance performance matrix analysis or priority map analysis) is a very useful analytical tool in PLS-SEM; which graphically extends the standard path coefficient estimates in a more practical approach (Ringle and Sarstedt, 2016). More precisely, IPMA presents a contrast of importance (i.e. total effect of predecessor constructs in predicting a target construct) and performance (i.e. average latent variable scores). According to Ringle and Sarstedt (2016) goal of IPMA is to identify predecessors which have a relatively low performance but high importance for the target constructs. A one-unit point increase in the performance of predecessor construct, will increase the performance of target construct, by the total effect size (i.e. importance) of the same predecessor construct (Ringle and Sarstedt, 2016; Schloderer et al., 2014). In our case, Customer Satisfaction is a target construct, which is predicted by five predecessors (i.e. Airline Tangibles; Terminal Tangibles; Personnel Services; Empathy and Image); refer to Fig. 1. We have performed IPMA for this study and result is presented in Fig. 3. Looking at the lower right area of the importance performance map; it is depicted that “image” have highest importance score i.e. 0.626; if Malaysia Airlines increases its image performance by one unit point; its’ overall customer satisfaction will increase by 0.626 (ceteris paribus). Moreover, our findings have revealed that Malaysia Airlines have lowest performance on Airline Tangibles and Terminal Tangibles i.e. 58.256 and 51.857 respectively; which means that there is a great room for improvement in these areas. For the ease of readers, a complete list of importance-performance values is provided in Table 8.

Airline Image

* p