Proposal model
Based on the foundations of former research, the model in the study of Gummerus, et al
(2012) is mostly suitable to our objectives as stated above We re-applied this model as figured out below in the Vietnam airline industry context
Hypotheses summary
H1a Community Engagement behavior has positive impact Social benefit
H1b: Community Engagement behavior has positive impact Entertainment benefit
H1c: Community Engagement behavior has positive impact Economic benefit
H2a: Transaction Engagement behavior has positive impact Social benefit
H2b: Transaction Engagement behavior has positive impact Entertainment benefit
H2c: Transaction Engagement behavior has positive impact Economic benefit
H3a Social benefit has positive impact on Customer Brand Page Commitment
H3b Social benefit has positive impact on Customer Word of Mouth
H3c Social benefit has positive impact on Customer Brand Page Commitment
H4a Entertainment benefit has positive impact on Brand Page Commitment
H4b Entertainment benefit has positive impact on Customer Word of Mouth
H4c Entertainment benefit has positive impact on Satisfaction
H5a Economic benefit has positive impact on Brand Page Commitment
H5b Economic benefit has positive impact on Customer Word of Mouth
H5c Economic benefit has positive impact on Satisfaction
This chapter outlines the proposed research model based on a comprehensive literature review, focusing on customer engagement behaviors on a brand’s Facebook page, categorized into community engagement and transaction engagement Both constructs positively influence perceived customer benefits, including social, entertainment, and economic advantages associated with joining a brand's Facebook page Additionally, the model examines how these benefits affect customer commitment to the brand page, word-of-mouth communication, and overall satisfaction These factors were chosen due to their established relationships demonstrated in prior research.
Hence, there are fifteen hypotheses proposed for this research The next chapter will discuss about methodology that used to analyze the data and test hypotheses of the research model
This chapter outlines the research design and methodology employed to investigate the relationships among key variables, including community engagement behavior, transaction engagement behavior, social benefits, entertainment benefits, economic benefits, brand page commitment, customer word of mouth, and customer satisfaction It begins with a detailed description of the research design process and the development of the questionnaire, followed by an introduction to the analysis methods Additionally, the chapter discusses the sampling method and the types of analysis that will be utilized to evaluate the study's findings.
The mixed methodology includes quantitative method and qualitative method Denzin
The mixed method approach, as defined by 1978, combines various research methods to explore scientific phenomena Recently, it has gained popularity for conducting research (Creswell, 2003), as it integrates qualitative and quantitative methodologies to gain a deeper understanding of concepts and gather accurate data This research utilizes theories from different authors to address research problems and hypotheses through qualitative methods Subsequently, quantitative interviews are conducted to gather responses from a diverse range of participants, varying in age, education, and employment This process includes testing the alignment between theoretical frameworks and the real-world data collected.
Sample
According to Hair, Anderson, Tatham, and Black (1998), the minimum sample size for accurate analysis should be at least five times the number of variables, with a minimum threshold of 100 participants to ensure reliable results.
This research involves 33 variables, necessitating a minimum of 165 responses However, to effectively conduct Structural Equation Modelling (SEM), Garver and Mentser (1999) recommend a sample size of at least 300 for reliable results.
300 samples, this research was intended to have about 350 samples delivered to respondents.
Measurement scale
The research model was meticulously crafted to ensure the relevance of the study's scope, with all variables measured using multiple items developed by previous researchers and informed by qualitative findings Community engagement behavior was assessed through three items based on the work of Cvijikj and Michahelles (2013), Gummerus et al (2012), and Pham and Tran (2014) Transaction engagement behavior was evaluated using two items from Gummerus et al (2012) and Pham and Tran (2014), modified for the context of airline fan pages Additionally, social benefits were measured using seven items, and entertainment benefits were captured with three items, as suggested by Dholakia et al.
(2004) and Gummerus et al., (2012) were used; The economic profit was measured by four items from research of Gwinner et al (1998); Yen and Gwinner, (2003), Gummerus et al.,
In 2012, the concept of satisfaction was grounded in Oliver's (1997) framework, while customer loyalty was assessed using metrics from Ouwersloot and Odekerken-Schruder (2008) and Punniyamoorthy and Raj (2007) To minimize response bias, the questionnaire omitted specific concept names, presenting all questions in a continuous table format.
Table 3.1 Source of measurement scale
Indicators Reseachers Vietnamese Modified Items /
Visit facebook Group of brands
1 Tôi truy cập (mở) facebook của hãng hàng không mỗi khi tôi vào mạng xã hội CEB1
2 Tôi đọc các thông tin (chương trình khuyến mại, giá vé, chuyến bay, kinh nghiệm du lịch v.v ) được đăng trên facebook Hãng Hàng Không CEB2
3 Tôi bấm like (thích) các bài đăng (status - trạng thái, hình ảnh hoặc video) trên facebook của hãng hàng không CEB3
Indicators Reseachers Vietnamese Modified Items /
4 Tôi comment (bình luận) về các nội dung (Trạng thái, hình ảnh, video) được đăng trên facebook của hãng hàng không CEB4
5 Tôi hỏi/ thắc mắc về chương trình giá vé/ chính sách/khuyến mãi của hãng hàng không trên facebook của hãng CEB5
Use links provided by the company/brand on
Tôi truy cập vào các liên kết trên Facebook của hãng hàng không để tìm hiểu thông tin sản phẩm, chính sách, khuyến mãi, giá vé, du lịch và các thông tin liên quan khác.
7 Tôi thực hiện mua vé máy bay ngay trên facebook của hãng hàng không TEB1
Participate in company’s/brand’s raffles on Facebook
8 Tôi tham gia vào các cuộc thi được tổ chức trên facebook của hãng hàng không TEB2
9 9 Tôi chơi các trò chơi trên facebook của hãng hàng không TEB3
Facebook group member to get information (e.g new products) Dholakia,
10 Tôi được cung cấp thông tin (về sản phẩm, sản phẩm mới, giá vé, chính sách, khuyến mãi, thông tin du lịch…) từ facebook hãng hàng không SB1
Facebook group member to provide other group members with information
Tôi chia sẻ thông tin hữu ích về sản phẩm, sản phẩm mới, giá vé, chính sách, khuyến mãi và thông tin du lịch cho các thành viên tham gia nhóm Facebook của hãng hàng không.
Facebook group member to share my ideas with other group members
Tôi chia sẻ ý kiến cá nhân của mình về chương trình, dịch vụ, sản phẩm, giá vé và kinh nghiệm du lịch với các thành viên khác trên trang Facebook của hãng hàng không.
Indicators Reseachers Vietnamese Modified Items /
Facebook group member to get to know other community members
13 Tôi kết bạn, giao lưu với nhiều thành viên khác cùng tham gia facebook của Hãng Hàng Không
Facebook group member to get help from other community members
Tôi đã nhận được sự hỗ trợ từ các thành viên trong nhóm Facebook của hãng hàng không, bao gồm trợ giúp mua vé, hướng dẫn thủ tục mua vé, hướng dẫn lên máy bay và thông tin du lịch hữu ích.
Facebook group member to help from other community members
Tôi hỗ trợ các thành viên tham gia nhóm Facebook của hãng hàng không bằng cách giúp họ mua vé, hướng dẫn thủ tục mua vé, hướng dẫn quy trình lên máy bay và cung cấp thông tin du lịch hữu ích.
Facebook group member to feel needed by [brand] or other community members
16 Các thành viên khác cùng tham gia trên facebook của Hãng Hàng Không cần tôi
Facebook group member to get entertained
17 Tham gia vào facebook của hãng hàng không giúp tôi giải trí
Facebook group member to relax
18 Tham gia vào facebook của hãng hàng không giúp thư giãn
Facebook group member to pass time when I am bored
19 Tham gia vào facebook của hãng hàng không giúp tôi giết thời gian khi tôi buồn chán
Facebook group member to try to get bonuses
20 Tôi nhận được các phần thưởng/ điểm thưởng từ các cuộc thi/ trò chơi tổ chức trên facebook của Hãng Hàng Không
Facebook group member to participate in
21 Tôi mua được vé giá rẻ khi tham gia vào facebook của hãng hàng không
Indicators Reseachers Vietnamese Modified Items /
Facebook group member to get better service
22 Tôi nhận được dịch vụ tốt hơn từ hãng hàng không khi tham gia vào facebook của hãng
Facebook group member to get fast responses
Tôi đã nhận được phản hồi nhanh chóng từ các hãng hàng không khi hỏi về chính sách, giá vé, khuyến mãi, thông tin du lịch và khiếu nại về dịch vụ.
I feel emotionally attached to [Brand]
24 Tôi cảm thấy có sự gắn kết về mặt tình cảm với facebook của Hãng Hàng Không
[Brand] has a great deal of personal meaning for me
25 Facebook của Hãng Hàng Không có ý nghĩa đối với cá nhân tôi
I feel a strong sense of identification with [Brand]
I feel as a part of the [Brand] Facebook Community
26 Tôi cảm thấy mình là một phần của facebook Hãng Hàng Không
I recommend the brand to other people
27 Tôi nhắc đến hãng hàng không khi nói chuyện/ trao đổi với người khác
28 I introduce the brand to other people
28 Tôi chủ động giới thiệu hãng hàng không cho người khác
I say positive things about the brand to other people
29 Tôi nói những điều tích cực về hãng hàng không với người khác
I am satisfied with my decision to become a member/fan of the [brand] Facebook
30 Tôi hài lòng khi là thành viên của facebook hãng hàng không
Indicators Reseachers Vietnamese Modified Items /
I think that I did the right thing when I decided to become a [brand] Facebook group member/fan
31 Tôi quyết định đúng khi tham gia vào facebook của hãng hàng không
I am satisfied with my decision to become a [brand] customer
32 Tôi hài lòng với quyết định trở thành khách hàng của hãng hàng không
33 Nói chung tôi hài lòng về hãng hàng không SF4
This study explores the intersection of online gaming and the tourism industry, a topic previously unexamined in the airline sector It was conducted in two phases: a qualitative study followed by a quantitative study The initial draft questionnaire included eight measurement scales: community engagement behavior, transaction engagement behavior, social benefit, entertainment benefit, economic benefit, brand page commitment, customer word of mouth, and customer satisfaction The questionnaire was translated from English to Vietnamese, and in-depth interviews were performed to refine the items In the quantitative phase, the adjusted questionnaires were utilized to gather data and evaluate the measurement and structural models, as illustrated in Figure 3.1.
Qualitative phase
In the qualitative phase, the Vietnamese version of the survey questionnaire underwent pre-testing through in-depth interviews conducted over two weeks with eight participants, comprising three experts in airline marketing and five customers who follow the airline's Fanpage This process aimed to ensure that the participants clearly understood the scales used in the survey.
Eliminating items had the low corrected item – total correlation Check Cronbach Alpha
Eliminating items had the Eigen-value less than 0.5 and item that distributes in 2 or more components/factors with difference less than 0.3. Combining variables into group of variables
Eliminating items had low CFA coefficients Check the suitability of model
Calculating the total reliability coefficient and cumulative variance Check the uni-dimension of the scales, convergent and discriminant validity
Check the suitability of model Check hypothesis in Viet Nam All the comments from the interviewees were gathered with the aim to modify the measurement scale
Based on the feedback of respondents, the survey questionnaire was slightly modified to make it clearer and more understandable (see Appendix 1).
Main survey - Quantitative research
In association with the theoretical literature, other researches and the results of qualitative phase, the final questionnaire was given out to implicating the main survey (see Appendix 2 – Final questionnaire)
The main survey was conducted in the large scale with the attendant of three airlines fan page members:
- Vietnam Airlines: 251,981 “liked” https://www.facebook.com/VietnamAirlinesCorp?fref=ts
- Jetstar Pacific Airlines: 376,879 “liked” https://www.facebook.com/JetstarVN?fref=ts
- VietJetAir.com: 652,452 “liked” https://www.facebook.com/vietjetvietnam?fref=ts
To assess the degree of respondent’s answer, Liker scale 5 was used in the answer sheet
1 (strongly disagree), 2 (disagree), 3 (neutral), 4 (agree), and 5 (strongly agree)
The primary objective of the survey was to target respondents who engage with airline Facebook pages A Google Form was utilized to create the questionnaire, which was distributed via the fan pages of Vietnam Airlines (VNA), Jetstar Pacific Airlines (JP), and VietJet Air (VJ), as well as through personal Facebook walls and emails Over a three-week period, from May 15 to June 4, 2016, a total of 353 questionnaires were collected After excluding 7 invalid responses due to identical answers or blank items, 346 questionnaires were deemed valid for analysis, meeting the minimum sample size requirements satisfactorily.
Table 3.2 Sources of data collection
Data entry was conducted using SPSS software, analyzing responses from 346 participants Reliability was assessed through Cronbach’s Alpha coefficient, followed by Exploratory Factor Analysis (EFA) to identify underlying structures Confirmatory Factor Analysis (CFA) was then employed to refine and validate the measurement model Finally, Structural Equation Modeling (SEM) analysis was performed using AMOS 2.0 to test the proposed hypotheses.
This chapter outlines the selection of sample size, construction of measurement scales, and research methods used for data processing An online questionnaire was created and distributed via an airline fan page on Facebook and through email The study was conducted in two phases: the first phase involved qualitative research through in-depth interviews, which were used to refine the measurement scale Following this, the questionnaire was slightly modified before the official survey The main survey included a sample size of 246 valid questionnaires, which were analyzed using regression methods The subsequent chapter will present the results of the data analysis from the main survey.
This chapter presents the survey results and interpretations, starting with a demographic breakdown of respondents by gender and age It then discusses scale validation through Cronbach’s alpha to assess the reliability and validity of the constructs Following this, exploratory and confirmatory factor analyses are conducted to evaluate discriminant and convergent validity Finally, the research model is analyzed using structural equation modeling, and the outcomes of the hypotheses are interpreted.
Table 4.1 summarizes the demographic analysis results, revealing that 41% of respondents favored the Vietnam Airlines fan page, while 46% preferred VietJet Air and 13% chose Jetstar Pacific Additionally, the gender distribution of respondents showed that 34.7% were female and 65.3% were male.
The study revealed that a significant majority, 80.4%, of participants were young individuals aged 18 to 30 Among the respondents, 37.3% fell within the age bracket of 18 to 22 years, while 18.5% represented those aged over 40 This demographic distribution highlights the prevalence of youth engagement in the research.
The majority of airline fan page participants are young adults, with 24.6% aged 26 to 30 and 80% of Facebook users in Vietnam falling within the 18-34 age range (We Are Social, 2015) Additionally, 13.9% of respondents are between 31 and 40 years old, while only 5.5% are over 40 This trend highlights the prevalence of younger customers engaging with airline fan pages.
A significant majority of respondents, 82.7%, hold a college or bachelor's degree, while 12.7% have obtained a postgraduate degree Income levels among customers vary, with 28.9% earning below 7,000,000 VND, 32.9% earning between 7,000,000 VND and 10,000,000 VND, and 20.2% earning between 10,000,000 VND and 15,000,000 VND Only 17.9% of customers have an income exceeding 15,000,000 VND Overall, the respondents display a diverse range of demographics, including gender, age, education level, and income.
Table 4.1 Description of sample Demographic profile Category Frequency Percentage (%) Airlines Fan’s page
College degree or bachelor's degree 286 82.7
A vital aspect of data analysis involves ensuring the reliability of all items within a scale, which accurately measure the corresponding research concept Consequently, conducting a reliability test for each construct in the measurement scale is essential.
To verify the reliability of the research instrument, Cronbach's Alpha was utilized to assess the internal consistency of the scales This test facilitated the elimination of unstandardized scale items, enhancing the overall validity of the measurement tool.
To ensure internal reliability, Cronbach’s Alpha should be at least 0.6 (Nunnally & Burnstein, as cited in Nguyen & Barrett, 2006) Additionally, the Corrected Item-Total Correlation is crucial; items with a correlation above 0.3 are likely well correlated with other scale items and contribute positively to the overall rating Conversely, items with a negative or low correlation (below 0.3) should be re-evaluated for wording and conceptual clarity (Leech, Barrett, & Morgan, 2005).
A modification or deletion for such items in this case was essential The results of reliability test for each construct in the model were summarized in the table below
Scale Mean if Item Deleted
Scale Mean if Item Deleted
Cronbach's Alpha if Item Deleted Community engagement behavior – CEB
Scale Mean if Item Deleted
Scale Mean if Item Deleted
Cronbach's Alpha if Item Deleted Social benefits – SB
All eight measurement scales demonstrated reliability in assessing the research concepts, with Cronbach’s Alpha values exceeding 0.6 However, the Corrected Item-Total Correlation for items CEB5, SB6, and SB7 fell below the acceptable threshold of 0.3, leading to their elimination Consequently, the concepts of community engagement behavior and social benefits were re-evaluated, with the results presented in the table below.
Tables 4.3 Retest Cronbach’s alpha results
Scale Mean if Item Deleted
Scale Mean if Item Deleted
Cronbach's Alpha if Item Deleted Community engagement behavior – CEB
After refining the scale, the reliability of community engagement behavior improved significantly, achieving a Cronbach’s Alpha of 0.842, up from the original 0.791 Additionally, the social benefits scale also demonstrated enhanced reliability with an alpha value of 0.843 All items met the standard for Corrected Item-Total Correlation, remaining below the threshold of 0.3 Consequently, all items in the scales were deemed satisfactory and retained for Exploratory Factor Analysis (EFA).
Factor analysis is a statistical method that reveals the latent structure of a set of variables, effectively condensing a larger number of attributes into a more manageable model Key criteria for conducting factor analysis include the Kaiser-Meyer-Olkin (KMO) measure of sampling adequacy, which should range from 0.6 to 1, and Bartlett’s test of Sphericity, which must be significant (p < 0.05) Additionally, the Total Variance Explained should exceed 50%, and each factor's Eigen-value must be greater than 1 Items with loading factors below 0.5 or those that distribute across multiple components with a difference of less than 0.3 are typically excluded from the analysis.
In the first rotated round, 30 items/variables were grouped into 7 components (see Table 4.4.) with KMO is 828 and Sig 000 Total Variance Explained is 56.068 %
The initial round of Exploratory Factor Analysis (EFA) revealed that items TEB1 and TEB3 had loading factors below 0.5, leading to their elimination Following this, EFA was conducted in four rounds, resulting in the removal of items TEB1, TEB3, and SB2 The final outcomes of the analysis are presented in Table 4.5.
Confirmatory factor analysis (CFA)
The results of the measurement model check are presented through confirmatory factor analysis (CFA), a statistical method that verifies the hypothesized relationships between observed variables and their underlying latent constructs (Steenkamp & van Trijp, as cited in Nguyen & Nguyen, 2011) Additionally, CFA allows for the assessment of the validity of research data.
CFA specifically, relies on several statistical tests to determine the adequacy of model fit to the data As shown:
The chi-square test measures the discrepancy between expected and observed covariance matrices, with a value near zero suggesting minimal difference For the results to be considered statistically significant, the probability level should exceed 0.05 when the chi-square value is close to zero.
The Comparative Fit Index (CFI) measures model fit by adjusting the discrepancy function for sample size, with values ranging from 0 to 1; a higher CFI indicates a better fit A CFI value of 0.90 or above is considered acceptable for model fit (Hu & Bentler, as cited in Suhr, 2012).
The Root Mean Square Error of Approximation (RMSEA) measures the residuals in a model, with values ranging from 0 to 1 A lower RMSEA value signifies a better fit for the model, and an RMSEA of 0.06 or less is considered acceptable (Hu & Bentler, as cited in Suhr, 2012).
The measurement model with all seven constructs was assessed using Confirmatory Factor Analysis (CFA) AMOS Software was used for CFA Diagram of latent variables was draws as Figure 4.2 below
The linear structural analysis results indicated that the saturated model had a Chi-square value of 490.803 with 303 degrees of freedom, resulting in a normalized Chi-square of 1.620, which is below the threshold of 2 Additionally, the RMSEA was 0.042 (less than 0.05), the TLI was 0.948 (greater than 0.9), and the CFI was 0.908 (also greater than 0.9) Furthermore, all factor loadings exceeded 0.5, confirming that the model aligns well with the market data.
All standardized regression weights for the scales exceeded 0.5, indicating statistical significance with p-values approaching zero, thus confirming convergent validity (Gerbing & Anderson, 1988) Additionally, convergent validity was evaluated through Composite Reliability (CR) and Average Variance Extracted (AVE) (Udugama & Ikram, 2012) For optimal reliability and internal consistency, CR should ideally be 0.7 or higher (Fornell & Bookstein, 1982 as cited in Udugama).
To ensure adequate convergence and distinct validity in scale measurement, the Average Variance Extracted (AVE) should be 0.5 or higher, as established by Fornell & Larker (1981) and referenced in Udugama & Ikram (2012) The scales in this study meet the required criteria for both Composite Reliability (CR) and AVE, indicating that the Confirmatory Factor Analysis (CFA) results are acceptable.
Table 4.7 Composite Reliability and Average Variance Extracted
Customer Word of Mouth WOM 0.817 0.599
In this phase, the model underwent testing for discriminant validity and the correlations among the variables were evaluated The results from the Confirmatory Factor Analysis (CFA) revealed that the relationships between the constructs in the research model were distinct, as they differed from 1.00 Furthermore, discriminant validity was assessed by ensuring that the Average Variance Extracted (AVE) for each construct exceeded the correlation between the constructs, following the guidelines established by Chin (1998) and Fornell & Larcker.
1981) These values are showed in Table 4.3 All the estimated correlation relationships were lower than 1.0 with of the p-values were all 0.000 and significant
Table 4.8 Discriminant validity test Correlations: (Group number 1 -
Default model) n46 r2 SE CR p_value
In conclusion, the assessment of the measurement scale, which included confirmatory factor analysis and discriminant validity analysis, indicates that the measurement model effectively meets the criteria for goodness-of-fit with market data, demonstrating reliability, convergent validity, and discriminant validity.
Structural equation model
In Chapter Three, the structural equation modeling (SEM) analysis method is utilized to evaluate the research model Consistent with the measurement model verification, the Maximum Likelihood method is employed to estimate the regression weights of the research model This step aims to determine the validity of the hypotheses presented in the literature review.
In Chapter Two, the official research model was introduced, incorporating seven key constructs: Customer Engagement Behavior (EBE), Social Benefit (SB), Entertainment Benefit (EB), Economic Benefit (ECO), and Customer Commitment (CM).
(6) Customer Word of Mouth – WOM, (7) Cusromer Satisfaction - SF
The SEM results shows that the theoretical model received an acceptable fit to the data Table 4.8 presented the fitness of measurement of measurement model
Table 4.9 Fitness of conceptualized model
Key indices Acceptable level Current value Result Conceptualized model
Chi-square/df (Cmin/df)
