ºÚÁÏÍø

ºÚÁÏÍø Scientific Reports

Your Research - Your Rights

Customer Satisfaction Measurement in Airline Services Using Servqual

Research Article ºÚÁÏÍø
Elif Degirmenci*, Huseyin Basligil, Ahmet Bolat and Yavuz Ozdemir
Yildiz Technical University Department of Industrial Engineering
*Corresponding authors: Elif Degirmenci
Department of Industrial Engineering, Yildiz Technical University Barbaros Street, Yildiz, Istanbul 34349, Turkey,
E-mail: elifdegirmenci@thy.com
 
Received December 08, 2012; Published August 11, 2012
 
Citation: Degirmenci E, Basligil H, Bolat A, Ozdemir Y (2012) Customer Satisfaction Measurement in Airline Services Using Servqual. 1: 294. doi:10.4172/scientificreports.294
 
Copyright: © 2012 Degirmenci E, et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
 
Abstract
 
Recently, airline companies have realized the importance of satisfied customers to find a place for themselves in this competitive world and initiated many projects to measure service quality and satisfy the customers by improving service quality. Based on this, in this study, in order to evaluate customer satisfaction at Turkish Airlines, the factors affecting customer’s experience were analyzed using weighted SERVQUAL methodology. In addition, the gap between Turkish Airline’s current service quality and 5-star service quality defined by SKYTRAX (the most accepted airline quality rating organization that uses evaluations of airline customers’ from all over the world) was measured. The analysis and the results were extended by constructing 5 hypotheses.
 
This research varies in many ways from the studies in the literature. In determining the factors affecting customer’s experience, unlike the studies in the literature, SKYTRAX customer satisfaction criteria were considered. Factor analysis grouped the questions included in the survey into 6 factors (dimensions): ground handling, employees, in-flight services, e-commerce, image and empathy. The results suggested that image dimension has the highest customer satisfaction level; employees and empathy dimensions followed the image. e-commerce has the lowest satisfaction level; in-flight services and ground handling service followed that. Another result is that meals and passenger transferring services have the highest impact on customer satisfaction. Furthermore, as expected none of service quality dimensions are around the desired 5-star customer satisfaction level defined by SKYTRAX.
 
Keywords
 
SERVQUAL; Customer satisfaction; Service quality; CRM; Loyalty; Airline Service Quality.
 
Introduction
 
In the aviation sector, investments and volume of business are increasing each day, and depending on this, competition increasingly deepens. Airline companies must be financially strong to find a place for themselves and survive in this deeply competitive world since in today's world, almost all of the hundreds of studies conducted under the heading “the most powerful companies" are already related to the financial power. Financial power is about the product sold. The sale of a product is possible only when customer satisfaction is met. Also companies have realized that they have to listen to consumers to survive in a competitive market [1] and airline companies also have become aware of the importance of having happy customers [2]. The pre-requisite of the customer satisfaction is understanding and knowing what they want. This imperative gave birth to the concept "CRM (Customer Relationship Management)". This concept is about customer satisfaction.
 
Customer satisfaction, after a step leads to the concept of customer loyalty. According to Oliver [3], loyalty is the highest level of satisfaction. According to Kramer [4], the more the customer satisfaction is, the lower the transaction costs are and the higher the fidelity is. Sirdeshmukh et al. [5] suggest that if the value increases offered to the customer by an airline after it loyalty comes. Cronin and Steven [6] and Butcher et al. [7] showed that the quality of service affects the tendency to purchase again as a result of increased satisfaction. Based on these, in our study service quality at Turkish Airlines was measured using weighted SERVQUAL method.
 
This study differs from previous SERVQUAL studies in some significant respects: First, it is the first application that its criteria were constituted based on the SKYTRAX criteria. Second, all the expectations were accepted as 5 (the highest rating level) thereby, it is aimed to show that the distance which should be taken to achieve the best in all the areas. Third, the study identified a new factor e-commerce which has not been considered before although it is very important nowadays. And a new model called “the spiral of loyalty” was developed in this study. This model is related to degree of customer loyalty. In addition to these, sample size in this study is large compared to the sample sizes of previous studies and nationality distribution of this study is broadly symmetrical. Thus, it can be said that this study is valid on global scope.
 
Concepts and a New Model: Spiral of Loyalty
 
In our study, we developed a new model called "The spiral of loyalty". This model aims to show that the customer approaches to the company’s value center gradually and thus aims to identify the degree of customer loyalty. Figure 1 show that a customer, who cannot be seen in the spiral of loyalty, is approaching to the company’s value center gradually after entering the spiral and the spiral ends with loyal customers. In addition, loyal customers fall into a ring at their endstage. This ring is a secure area where all customer types are required to be included, and advanced to the area so that this would explain the aim of this model.
 
Figure 1: Spiral of Loyalty.
 
According to Dinçer [8], service is benefit or satisfaction offered to the market activity or by depending on the sale of goods. According to Goetsch and Stanley [9] service is to perform work for someone else.
 
Quality means compliance with requirements [10]. The service quality model by Grönroos holds that the quality of a service should be divided into two: technical quality and functional quality. Technical quality is related with the quality of main products; functional quality is related with service purchasing processes that reflect the experiences of service procurement processes [11]. Quality is defined by customers and perceived quality is the direction and degree of difference between customer expectations and perceptions [12,13]. In other words service quality is the measure of the difference between customers’ expectations and the perceived ones [12,13].
 
Servqual in Airline Services
 
In literature, there are various studies measuring the quality of airline service. SERVQUAL method is a popular approach to this. Most of these studies aim to show the relationships between service quality and related issues. Surovitskikh and Lubbe [14] classified airline service quality in terms of three items: consistency of service, reliability of service, and augmented products. The study examines the positioning of four selected Middle Eastern airlines in the South African business and leisure travel environment. Since the study is related with four airlines in a specific region, it can be said that the coverage of the study is high according to studies measuring one airline. That study examines the relationship between service quality and age, number of flights, income group, purpose of travel and the carrier. Gourdin [15] classified airline service quality in terms of three items: price, safety, and timelines. Gilbert and Wong [16] used employees, facilities, customization, flight patterns, assurance, reliability, responsiveness as the dimensions of service quality. They detected significant differences among passengers of different ethnic groups/nationalities as well as among passengers who travel for different purposes, such as business, holiday and visiting friends/relatives. Pakdil and Aydın [17] identified employees, tangibles, responsiveness, reliability and assurance, flight patterns, availability, image, and empathy as dimensions of their study. In that study responsiveness and empathy dimensions are very close to each other in terms of meaning. They proposed that the passengers’ educational level is an important variable affecting the quality of service. Chang and Yeh [18] proposed on-board comfort, airline employees, reliability of service, convenience of service, handling of abnormal conditions as service quality dimensions. A fuzzy multicriteria analysis (MA) model was used to formulate the problem. The study also covers four airlines as in Surovitskikh and Lubbe’s study [14].
 
Although there are some changes in the quality dimensions used in evaluating similar airline services, SERVQUAL is a commonly used method. Airline service quality dimensions can be grouped under the heading of SERVQUAL. Furthermore, the relationships can be expanded according to the purpose of the study.
 
While some aspects of our study are similar to the studies in the literature, there is a crucial point differentiating our study from the literature. In this research, we employed weighted SERVQUAL method based on the Skytrax criteria. This approach is the first in the literature.
 
Methodology
 
Questionnaire design
 
Survey forms were designed according to Skytrax, the most accepted and prestigious official airline quality 5-star rating organization, system and criteria. 34-item questionnaire was designed to overlap with Skytrax’s 34 basic criteria. The survey forms were prepared in English and Turkish to increase communication, convenience, and the response rate. The forms consisted of two sections. The first focused on passengers’ demographic profiles like sex, flight frequency, and so on. The second consisted of 34 statements covering the service satisfaction of airline passengers and a likert scale. In our design, a 5-likert scale was used in accordance with Skytrax’s 5 stars. Perceptions were rated from 1 = strongly disagree to 5 = strongly agree.
 
All customers' expectations have been accepted as 5 so that the best was aimed in the improvement of the service quality. In other words, we wanted to show the distance to be taken to reach a service quality which is beyond the customer's expectation. Since the expectations have been accepted as 5, instead of measuring both expectations and perceptions, the questionnaire was designed to measure the perceptions only. In the last step, the questionnaire was examined by executives at the airline and their contributions were incorporated.
 
Survey distribution and evaluation
 
This study was conducted in all long-haul flights in a specific day, 75 flights in total. Questionnaires were applied during flight. To make the sample more representatives, the study was performed in the last hour of the flights. 9000 questionnaires were distributed in total and 2000 completions of the questionnaire were calculated. After eliminating the missing data, 1110 convenient questionnaires remained. Data obtained from surveys were evaluated using SPSS 15.0.
 
Research method
 
All customers' expectations have been accepted as 5. As a method, the weighted SERVQUAL method was applied to measure service quality in Turkish Airlines.
 
Implementation and Analysis
 
Reliability analysis
 
Reliability analysis was performed to prove whether all survey questions are closely related with each other, and perceived by all respondents in the same way or not. For reliability, Cronbach's Alpha Internal Consistency Analysis was applied. A Cronbach alpha coefficient varies between 0.00 and 1.00 and it is expected to have 0.70 or higher [19,20]. In this study α was found 0.964 which is very satisfactory. However, to make a healthy assessment, contribution of each question to this coefficient was also investigated. Results show that all cases were ≥ 0.7. Thus the internal consistency is very high for each question. In other words, each question is reliable.
 
However, each factor’s internal consistency must also be provided. Firstly, the overall reliability for each factor, then reliability for each case of removal of the each question from relevant factor, was calculated as shown in table 1. The questions in each factor were analyzed.
 
 
Table 1: Factors and reliabilities.
 
Since all α coefficients greater than 0.7, all the factors are reliable in themselves. Besides this, for each case of the removal of each question from the related factor is analyzed. According to the results,
 
If any question is removed from the first factor, reliability does not increase considerably. So no question is removed.
 
If any question is removed from the second factor, reliability does not increase considerably. Therefore, none of the questions is removed from the analysis.
 
If any question is removed from the third factor, reliability does not increase considerably. Therefore, none of the questions is removed from the analysis.
 
If the 14th question is removed, the fourth factor’s reliability increases to 0.921. Thus, the 14th question is removed and the reliability is recalculated. The fourth factor’s internal consistency (reliability) increased to 921.
 
If the 21th question is removed, the fourth factor’s reliability increases to 0.840. So the 21th question is removed and the reliability is recalculated. The fifth factor’s internal consistency (reliability) increased to 0.840.
 
Lastly, since the sixth factor has two questions, so calculations are continued with these questions.
 
Validity analysis
 
It is necessary for the validity of a questionnaire to be reliable, but not enough. Therefore, after reliability has been proven, factor analysis must be done to demonstrate the validity. Firstly two prerequisites of the factor analysis were performed.
 
Bartlett Test (Sphericity Test): Bartlett’s measure defines whether factor analysis is necessary or not. Bartlett measure was found 0.000 which is highly significant. That is, there are some relationships between the variables we hope to include in the analysis. Therefore, the factor analysis is appropriate.
 
Kaiser-Meyer-Olkin (KMO) Test: The KMO measures the sampling adequacy which should be greater than 0.5 for a satisfactory factor analysis. Furthermore, values between 0.5 and 0.7 are mediocre, values between 0.7 and 0.8 are good, values between 0.8 and 0.9 are great and values above 0.9 are superb. KMO measure was found 0,938 which fall into the range of being superb. Therefore, we are confident that the factor analysis is appropriate for these sample size.
 
Anti-Image Matrix: Via this matrix, it is determined whether each question is suitable for factor analysis or not. To perform a factor analysis, all elements on the diagonal of this matrix should be greater than 0.5. Since all the variables are independent from one another, varimax method was applied. Since Covaraince matrix is >0.50, all questions placed in the questionnaire are suitable for factor analysis in our study.
 
Rotated Factor Loading Matrix: In this table (Table 2), the questions with the highest factor loadings have been brought together. If a question has close factor loadings under more than one factor, that question was removed from the analysis. According to this, the 5th, 11th, 18th, 22nd and 23 questions were removed. In addition, if a question has less than a 0.45 factor load, it was removed due to the rule.According to this, the 10th question was removed from the analysis. In addition, according to the results of the reliability analysis, the 14th and 21th questions were removed.
 
Table 2: Rotated component matrix
 
After the reliability analysis was performed and related questions were removed from the study, the final factor analysis was performed and 5 factors were extracted from the data set via principal components analysis. But it was noticed that questions 32, 33 and 34 placed in the first factor are related with image and these questions could constitute a single factor which would be more specific in itself. So it was considered as a separate factor. In this way, 6 factors were identified in total. In Table 3, each factor, factor loadings and reliabilities for each factor are shown.
 
Table 3: New Reliabilities.
 
Demographic results
 
Questions in the first part of the questionnaires were analyzed. Frequencies related with the customers’ demographic data were calculated and shown in the Table 4.
 
Table 4: Frequency data summary.
 
According to the results, nationality distribution is broadly symmetrical. This is an important finding so that findings of this study are valid globally. Another result is that experience is the most important reason for selecting an airline and price is following that.Furthermore, customers generally make their own decisions in airline choices instead of using travel agencies.
 
Calculation of weighted SERVQUAL score
 
Steps to be followed in applying the method are as follows:
 
Step 1: Servqual score is calculated for each question in relevant factor for each customer in accordance with the following formula.
 
SERVQUAL score (SS) = Perception Score– Expectation Score
 
Step 2: For each question, weighted SERVQUAL score is calculated by multiplying the question’s factor loading. This procedure is applied to all questions.
 
Step 3: In the relevant factor, each question’s weighted SERVQUAL scores are collected. Thus, customer’s weighted servqual score is obtained for relevant factor.
 
Step 4: The same processes are repeated for all customers. Thus, for each customer, weighted SERVQUAL score in each dimension is obtained.
 
Step 5: All customers’ weighted SERVQUAL scores for the first dimension are collected and divided by the total number of customers. Thus, the final weighted SERVQUAL score of the first dimension is obtained.
 
Step 6: The same process is repeated for each dimension.
 
Step 7: Each factor’s SERVQUAL scores are collected and this result is divided by the total number of the dimensions and the final weighted SERVQUAL score is found.
 
By applying the steps described above, the weighted SERVQUAL scores for each question mean scores, standard deviations, and weighted SERVQUAL scores for each factor were calculated and are presented (Table 5).
 
Table 5: Servqual Scores.
 
Servqual factors for each factor
 
First, overall average customer assessments were found for questions separated according to the six dimensions. In addition, according to the ratings given by Skytrax in six dimensions, average scores were calculated and then the differences between customer satisfaction and Skytrax are calculated. The difference between the Skytrax score ever since the last issue in relation to each factor, were calculated to show the direction of change in the perception of customers (Table 6).
 
Table 6: Skytrax rating customer assessment and the SERVQUAL score.
 
The positive values show the improvement achieved since the day of evaluation. Since the customer perception and satisfaction is higher than last Skytrax evaluation score, these areas experienced a positive increase. According to these data, the largest increase was experienced in 'empathy' dimension.
 
Negative values represent the cases that customer perception is lower than Skytrax’s last rating. This situation shows that the importance shown to these properties (questions) is not sustainable. In-flight services and e-commerce services could be given as examples.
 
Accordingly, for a sustainable quality in in-flight services, a survey study on customer preferences will be effective. With presentations and meeting the customer expectations, both satisfaction and sustainability will be achieved.
 
Decline in customer perception of e-commerce is related with ensuring up to date content and the introduction of new technologies related to the web. Today, a very intense competition is experienced in the aviation sector and this competition is usually in electronic environment. To give an example newly established airline companies’ web sites allow changes on a ticket. However, when this study was conducted, Turkish Airlines’ web site did not allow changes on a ticket because of mainframe-based sub-structure. This created many problems for customers. Based on this, Turkish Airlines’ web site needs to be improved since a web site is the most crucial tool in today’s competitive business world. Turkish Airlines conducted a project in 2011, and now on its web site, ticket change (re-issue) has become possible.
 
The results suggested that according to the gap scores, empathy dimension has the highest customer satisfaction level; image and e-commerce dimensions followed the empathy. In these three dimensions, customer satisfaction is met in general.
 
Ground handling service has the lowest satisfaction level. This is followed by the dimensions, and respectively, employees and in-flight services. These three dimensions are quite far from Skytrax’s 5 star satisfaction level.
 
Evaluation of the dimension of image: This dimension has the highest satisfaction with SERVQUAL score -0.440. This is also an indicator of the distance to catch up Skytrax’s 5 stars satisfaction level. The perception scores of features in this dimension are approximately the same. In other words, Turkish Airlines has a brand image in minds of its customers, and this image can be felt from the physical appearance of the planes to the image of the employees in all areas. Park et al. [21] claim that passengers’ repurchase and recommendation intentions are influenced by the airline image; so, this result is important. Based on this, if the Corporate Communications Department carries out some sustainable innovations as image study, this may help to improve the customer perception more.
 
Evaluation of the dimension of employees: The second dimension is the dimension of employees that provides the highest satisfaction with -0,513. Many reasons lie behind this gap score. One of them is a continuous increase in the number of cabin crew during 2009 and 2010. At the end of the year 2010, the number of teams was increased by 33% compared to January. In December, that number increased even more. The continuous and intense increase in the number of employees reduced the sustainable quality. Training of a small number of employees and training of a large number of employees cannot be given the same result. Turkish Airlines has increased the training of cabin crew based on this idea. New studies are expected to reflect the results of these trainings.
 
Evaluation of the dimension of empathy: Empathy dimension has the third highest customer satisfaction with SERVQUAL score -0,658. This dimension consists of two properties: behaviors in the unexpected situations and behaviors in needs. The results show that customers' needs in these two properties are met in general. These two features have a very similar degree of individual satisfaction.
 
Evaluation of the dimension of e-commerce transactions: There is a serious sense of dissatisfaction about e-commerce services. -0,989 SERVQUAL score is so far from Skytrax’s 5 stars satisfaction level.
 
E-commerce dimension consists of two features: ease of use and content of the website. Web site's ease of use has relatively lower satisfaction according to the content. This may be the result of the customers’ comparison between Turkish Airlines’ web site and the other companies’. Turkish Airlines' website was established many years ago. There hasn’t been a radical design change since then. In this period, many new airline companies were established in the world, and these firms have recently set up web sites using the new technology and the analysis techniques to meet customer expectations. Turkish Airlines’ planned new, advanced, innovative and competitive web site should be put into service as soon as possible. As a suggestion, the web site should serve in different languages and this will be an important functionality. In addition, another attractive feature may be new payment channels.
 
Evaluation of the dimension of in-flight services: The second issue that has emerged as the most distant perfection. This dimension with -0,897 SERVQUAL score is the second most distant dimension from SKYTRAX’s 5 star satisfaction level.
 
In the category of Economy Class Meal Service, Turkish Airlines was awarded with the 'World's Best Airline' award in the World Airline Awards survey conducted by Skytrax. Therefore the questions were analyzed in detail and the results showed that features related with meals and flight entertainment programs had higher scores. Reason for the SERVQUAL score -0,897 is dissatisfaction with the phone service and the distribution of newspapers and magazines during the flight. (Table 7)
 
Table 7: In flight services SERVQUAL scores.
 
Turkish Airlines initiated a project in 2010, about telephone service during the flight. In order to achieve the satisfaction of Skytrax’ 5 star in in-flight services; these project activities should be accelerated and put into effect. It is expected that the results will be reflected in future studies.
 
Flight entertainment programs are good for overall satisfaction, but it still needs to be improved to achieve the satisfaction of Skytrax 5 stars. In the flight entertainment screens, it is possible to measure the rating. This is meaningful in terms of a feedback from customers. In addition, program loading plans made according to feedback can help to catch a 5 star satisfaction.
 
Evaluation of the dimension of ground handling services: This dimension with -0,701 SERVQUAL score is the third most distant dimension from SKYTRAX’s 5 star satisfaction level. The underlying reason is that the period of the questionnaires was the period when Turkish Airlines' ground handling services was transferred to another company. It is known as the transition period. If the same situation continues, urgent policy should be taken on ground handling services. When a current state analysis is conducted, results of this study will be a guide since the data of this study is the actual customer data. In this way, customer losses caused by dissatisfaction on the ground handling services could be prevented.
 
The most important feature affecting customer satisfaction
 
To know the feature affecting the customer satisfaction the most is important for the determination of the point which is critical to get starting customer satisfaction studies. By this way, the most influential features in satisfaction were identified. Multiple regression analysis was used as the method.
 
The values in R Square column show that the properties in the survey explains 76% of the variance overall satisfaction. Coefficient (Coefficients) table shows the regression coefficients used for regression equation and their significance levels. Column of the significance of ANOVA table shows the relationship between these variables at p <0.01 significant level statistically.
 
The most important dimension affecting customer satisfaction is accordingly delicious and variety of dishes (Question_6) and ground handling services (Question_19). Therefore, changes or innovations to improve customer satisfaction must be initiated from the food and transfer services. Here, there is a remarkable thing that SERVQUAL dimension scores had the highest dissatisfaction at ground handling services dimension, although this dimension has the highest impact on satisfaction. This is a very thought-provoking result and urgent measures should be taken.
 
Hypotheses and Tests
 
H1: People who prefer generally “other” airline companies has lower satisfaction about provision for the fare than people who prefer generally “Turkish Airlines” companies.
 
H2: There is a significant difference between men and women passengers about the adequacy of in-flight entertainment programs
 
H3: There is a significant difference between men and women about the adequacy and tastiness of food
 
H4: Service quality gaps vary by the purpose of the flight
 
H5: Foreign passengers’ satisfaction about variety of newspapers and magazines distributed during flights is lower than Turkish passengers’(Table 9).
 
According to the results, people who prefer generally “other” airline companies has lower satisfaction about provision for the fare than people who prefer generally “Turkish Airlines”. Another result is that there is a significant difference between men and women passengers about the adequacy of in-flight entertainment programs. Women passengers’ satisfaction level is lower than men. Based on this, when the entertainment programs are installed, women specific programs also should be included. Foreign passengers’ satisfaction about variety of newspapers and magazines distributed during flights is lower than Turkish passengers’. To overcome this dissatisfaction, magazines in different languages must be distributed. One important result that service quality gaps do not vary by the purpose of the flight.
 
Comparison with Existing Studies
 
In this section, the study that we have done is compared with the previous studies in terms of the scope and the results. (Table 10)
 
Table 8: Multiplier regression analysis.
 
Table 9: Test of Hypotheses.
 
Table 10: Comparison with existing studies.
 
This study differs in many ways from the past studies done on this topic. (Table 8) lists the differences between this study and the previous studies.
 
1. The sample size of this study is higher than all the previous studies. In addition, since all selected flights for this study are international flights, leg based wrong analysis were prevented.
 
2. Unlike previous studies, all the criteria of this study were established based on Skytrax criteria. Likert scale was designed with five intervals to be appropriate with the Skytrax evaluation system also. Unlike past studies, all the expectations were accepted as five in order to show the gap remaining to catch up Skytrax’s 5 star satisfaction level to the company's managers
 
3. In addition, even though some of the dimensions are similar to those presented in previous studies, e-commerce is a new dimension in our study and it is one of the most important concepts today.
 
Conclusion
 
Exploratory factor analysis was used to extract the items that provided a reliable measure of the constructs under investigation in this study and as a result, questions were grouped in six factors (dimensions). These factors are ground handling, employees, in-flight services, e-commerce, image and empathy. These six factors explained approximately 70% of the total variance.
 
The results suggested that according to the gap scores, image dimension has the highest customer satisfaction level; employees and empathy dimensions followed the image. In these three dimensions, it was observed that customer satisfaction was met in general. On the other hand, e-commerce dimension has the lowest satisfaction level. This is followed by the dimensions of in-flight services and ground handling and convenient respectively. These three dimensions are quite far from Skytrax’s 5 star satisfaction level. On the basis of the service quality gaps, it was discovered that none of passengers’ perceptions caught up Skytrax’s 5-star level in any of the dimensions. This finding confirms that airline management should be more committed to performance improvement and alert for improvement opportunities throughout the service delivery processes.
 
According to the results, the dimensions of in-flight services and e-commerce services were observed to have fallen behind the latest assessment of Skytrax. This is an indicator of deteriorating. Some suggestions were given about this. All comparisons and calculations were carried out based on the Skytrax’s economy category ratings. Thus, without distinction of class, it is aimed to develop a model which will provide the best in all categories. However, in economy class since there is no CIP lounge service, evaluations about CIP were carried out based on business class. In all categories, Skytrax’s long-haul scores were taken into consideration. The reason for this is because all the surveys were performed in more than 2, 5 hour flights.
 
The most important dimension affecting customer satisfaction is accordingly the delicious and a variety of dishes and ground handling services. This is a very critical finding since this result can be a fundamental guide for projects to improve the satisfaction. Therefore, changes or innovations to improve customer satisfaction must be initiated from the food and transfer services.
 
In this study five hypotheses were developed in total. While three of them are accepted, two of them were rejected. According to the results of hypothesis tests, people who prefer generally “other” airline companies has lower satisfaction about provision for the fare than people who prefer generally “Turkish Airlines” companies.
 
There is a significant difference between men and women passengers about adequacy of in-flight entertainment programs. Based on this result, it is recommended that the variety of the programs must be increased according to male and female preferences. When flight entertainment programs are selected, special programs for women must also be preferred. This will make a positive effect on enhancing the satisfaction of the women passengers. Decoration, home design, cooking programs can be given as examples. However, there is no difference by gender in food preference and satisfaction. Another result shows that customer satisfaction does not differ by the purpose of the flight.
 
Satisfaction of the foreign passengers is lower than national passengers on newspapers and magazines distributed during the flight. According to this result, Turkish Airlines must re-review the newspaper and magazine preferences. Related with this, the range of languages must be extended by bringing a new approach, instead of only English and Turkish language publications. For example some additional publications in languages like French and Japanese must be added so that the dissatisfaction in this matter can be cleared up. In particular, some publications in the language of the destination country will increase the satisfaction.
 
Since the data and the findings of this study were gathered from passengers half of whom were domestic and the other half international, it may provide valuable insight to all other airline companies worldwide. The ‘‘voice of the customer’’ should be taken into the design process using advanced techniques, such as experimental design, quality function deployment, and value engineering. After delivering the services, service providers should monitor how well the customers’ expectations have been met (Pakdil and Aydın, 2007). Based on this, in further research, a similar study can be carried out after improvements based on the findings of this study to show the progress made.
 
Acknowledgements
 
We should thank TUBITAK for supporting this study. Furthermore, we owe thanks to Dr. Zeliha AKCA for their valuable comments and advice, Mr. Tumer ALTAS for his proofreading and the anonymous referees for their valuable help.
 
 
References