Factors affecting Vietnamese tourists’ decision to travel overseas

in general, the decision-making theories mentioned in the development of consumer behavior studies can be divided into two groups: classical theories: approaching decision-making behavior in 3 common ways: 1) Risk reduction is that an individual may be faced with having to predict the risks and consequences of make a purchase when making decisions about the products and services they will buy (Taylor, 1974). 2) Problem solving is an approach based on the assumption that any individual has a need or desire to consume issues before making a consumer decision. 3) Information processing: Bettman et al. (1998) based on the assumption that customers are always seeking and processing the information they gain. Post-modern and modern theories: Nowadays, with the explosion of information technology, especially the popularity of the internet and social networks, it will bring consumers more complete and diverse information from diverse sources. Consumers have more choices (brands, substitutes) when making decisions. The studies to update on consumption decisions are generally divided into five research-approach groups, including: Economics; psychology; behavioral sciences; cognitive science; and humanities. This thesis inherits the theory of planned behavior (TPB), proposed by Ajzen (1991), combined with other models which explain behavior of tourists: Um and Crompton (1990); Woodside and MacDonald (1994); Decrop (2006b). 2.3. The theory of consumer behavior in tourism 2.3.1 Models of travel decision-making The theoretical model of consumption decisions in tourism was developed by many scholars, but it can be divided into three main types: microeconomic model; Cognitive models (including structured cognitive models and cognitive processes); and paradigm paradigm (Decrop, 2006b). The decision-making model is based on microeconomic factors that emerged in the 1970s. While trying to explain the relationship between demand and cost, 6 this model does not address the specific issues of tourism products (Decrop, 2006b). Different from the economic model, decision-making models based on customers' perceptions are mainly focusing on the relationship of economic benefits; needs; and expectations of customers. In these models, consumer awareness and information processing play a core role in the purchasing decision process. It can be divided into two types of cognitive models: 1) A structured decision-making model, based on the relationship between the inputs (marketing variables, personal resources variables ...) and factors. output; 2) Decision-making process model approaches the problem by explaining how decision-making behavior takes place in the cognitive process. 2.4.2 Overview of factors affecting travel decision-making In this study, factors influencing travel decisions are divided into major groups, including: Attitude: studies of general consumer attitudes (Fishbein and Ajzen, 1975; Shimp, 1981) showed that attitude and beliefs are the expression of human emotions about a product, whether it can be good or bad, but they are all based on cognitive results. Although the research on attitudinal factors mentioned in previous research is quite comprehensive and clear. Travel motivation: motivation is mentioned in the previous studies as Dann (1977); Crompton (1979); Ross (1994); Murphy (2013). Also, it is constituted by the following aspects: 1) Escape from daily life; 2) Discover something new; 3) Rest; 4) Seek experience; 5) Seeking relaxion; 6) Asserting one's own values (Dann, 1977); and 7) Go to true place (MacCannell, 1973). Destination Images: reflecting the perception of tourists when thinking of destination. The destination image is made up of basic factors such as natural resources, cultural and historical values, customs ... that directly or indirectly influence consumers' decision-making in tourism (Hyde, 2004; McCracken, 2005; Swarbrooke and Horner, 2007). Customer approach activities: is one of the most important factors influencing travel destination selection behavior (Woodside and Lysonski, 1989). 7 Kotler (2017) asserts that the price and quality of services included in the tour are the two most important factors in the group of marketing factors that influence the decision-making of tourists. Reference groups: reference groups can influence the purpose of the trip (eg: visiting friends, visiting relatives, etc.) but can also influence from the perspective of the group's participation before, during and after making a decision. In recent years, along with the explosion of global information networks and especially the popularity of social networks, there has been the emergence of electronic word of mouth (eWOM) based on media and the internet. Personal control variables: including factors which influence decision- making such as personal characteristics including age, family status, education and occupation, personality and lifestyle ... 2.4.2 Overview of studies on outbound travel In Asia, studies focusing mainly on countries with large markets of outbound tourists such as China, Japan, Korea, Taiwan (Nozawa, 1992; Chen, 2001; Chang, 2009; Sparks and Pan, 2009; Lee et al., 2012). Although there has been a closer approach to the research topics of travel decision-making, it is evident that these studies come from the context of the study of outbound tourism market countries. have been around for a long time. An overview of the studies shows that there have not been many studies in the typical Vietnamese context. 2.5. Research gaps From the literature review of previous studies, there is some emerged theoretical gaps as follows: Firstly, the differences of cultural environment affect tourists' consumption behavior. In each society with different cultures, the influence from personal and environmental factors on tourist consumption decisions is different. Although there are many works in the world offering theories to explain decision-making. However, the research models are mostly based on the research context which is in Western countries. So far, there are very few studies on the topic affecting the factors affecting foreign travel decisions in Southeast Asia, and especially very rare in Vietnam. 8 Secondly, the intensity of influence from factors on tourist travel decisions has changed over time, requiring constant updates. The rapid change of the internet and social networks gives rise to new factors (eWOM) that have a significant impact on tourist’s decision-making. Third: Previous studies have mentioned that these research models focus only on theoretical perspectives on one (or a few) individual factors affecting tourist decisions. in practical, it’s required a model that reflects the relationship between individual psychological factors; environmental factors; the overall set of impact on the decision to travel abroad. Fourthly, the current growth trend of the outbound tourism market in Vietnam poses practical challenges that require theoretical supplementation to clarify the issues raised. 2.6. Research model and Hypotheses 2.6.1 Research model To point out the factors influencing Vietnamese outbound tourist’s decision- making and the relationship between these factors, this study was inherited Ajzen's theory of planned behavior (1991). and theories explaining factors affecting tourist behavior including Um and Crompton (1990); Woodside and MacDonald (1994); Decrop (2006b). 9 Figure 2.8 Proposed research model The basic factors inherited from the theory of planned behavior include: 1) Attitude towards travel abroad (Attitude toward the Behavior); 2) Travel motivation; 3) Destination image: inherited from Um and Crompton (1990); 4) Customer approach activities: inherited from Woodside and MacDonald (1994); 5) Reference group: Inheriting from Decrop (2006b); Woodside and MacDonald, (1994) 2.6.2 Research hypotheses Hypothesis 1 (H1): Destination image has a positive impact on the attitude of tourist Hypothesis 2 (H2): Destination image has a positive impact on the travel motivation of tourist Hypothesis 3 (H3): Customer approach activities have a positive impact on customer attitudes to travel abroad. Hypothesis 4 (H4): Customer approach activities have a positive impact on travel motivation Control variables Destination Images Travel Decision Customer Approach Activities Reference Group Attitude Travel Motivation H2 H3 H4 H5 H6 H7 H8 H1 10 Hypothesis 5 (H5): Reference group has a positive impact on customer attitudes of tourist Hypothesis 6 (H6): Reference group has a positive effect on travel motivation Hypothesis 7 (H7): Attitudes towards travel have a positive impact on decision-making to travel abroad. Hypothesis 8 (H8): Travel motivation has a positive impact on decision- making to travel abroad 11 CHAPTER 3: METHODOLOGY 3.1. General research framework In this study, the research process consists of 5 steps: Step 1: Identify research issues; Step 2: Overview of research works, theoretical basis and identify research gaps; Step 3: Qualitative research; Step 4: Quantitative research; Step 5: Finalize the report Data sample was randomly selected from Vietnamese tourists who travelled abroad at the waiting area in 3 major international airports: Noi Bai, Tan Son Nhat and Da Nang. Total of survey sent was 2000 and number of received was 850, the response rate was 42.5%. After eliminating the invalid votes 754 were left. The time for collecting votes is from March 10, 2019 to July 10, 2019. Data were processed and analyzed on SPSS 22 and AMOS software. 3.2. Variables and scales Likert 5-level scale from 1 - "Strongly disagree"; 2. Disagree; 3. Normal (neutral); 4. Agree and 5 - "Strongly agree" for the observed variables. Scale of Image destination, which was used in the thesis, is inherited from studies of Byon and Zhang (2010); Hsu et al. (2017); Obenour et al. (2005). Scale of Customer approach activities are inherited from the studies of Middleton et al (2009); Kotler (2017); Luo and Zhong (2015); Gruen (2005); Crick (2003). Scale of Reference group inherited Gitelson and Kerstetter (1995); Murphy et al. (2007); Xiang and Gretzel (2010). The measure scale of attitude variable was inherits from Lankford and Howard (1994); Phillips and Jang (2008); Sparks and Pan (2009). The concept of travel motivation variable is a Second-order factor, made up of first-class factors, including: Discovering new things, gathering knowledge and experiences (DIS); Share experiences with others (SHA); Seek pleasure (FIN); and Self-concept (SEL), as details: Measure scales of variable discover new things, gathering knowledge and experiences based on the scales of Fodness (1994); Chetthamrongchai (2017) and adjusted, based on qualitative research results. Scale of variable share experiences and with others, built on the scale inheritance in Marzuki et al. (2017). Scale of 12 find pleasure based on a scale in Fodness (1994). Variable scale Search and assert self-worth based on previous studies by Munar and Jacobsen (2014); Gnoth (1997); Crompton (1979). The Travel decision variable is measured using a scale inherited from Decrop (2006b); Wong and Kwong (2004); Chen et al. (2019). 3.3. Qualitative research methods The qualitative research was initially conducted on two subjects who are experts with in-depth knowledge in the field of customer behavior and consumer behavior of tourists (07 experts); and randomly selected from outbound tourists at airports (10 guests) 3.4 Quantitative research methods 3.4.1 Descriptive observed variables Descriptive analysis initially reflects an overview of actual situation from the data collected in this study. The study conducted a statistical descriptive analysis of the frequency of observed variables in the model through criteria including: maximum value, minimum value, average value and standard deviation. 3.4.2 Evaluation of measurement scales by Cronbach’s Alpha The reliability of the scale is evaluated through Cronbach’s Alpha coefficient and total correlation coefficient. This thesis will base on the scales that ensure Cronbach’s Alpha coefficient ≥ 0.6 and have the total correlation coefficient ≥ 0.3. 3.4.3 Testing the scale by exploratory factor analysis (EFA) Applying EFA method with PAF factor extraction with non-perpendicular rotation of Promax. The following criteria: KMO coefficient: 0.5 ≤ KMO ≤ 1; Bartlett test <0.05; Factor load factor ≥ 0.5; Extraction variance ≥ 50%. 3.4.4 Confirmation Factor Analysis (CFA) Evaluation criteria inherited from Hair et al. (2009); Baumgartner and Homburg (1996) including Chi-square / df 0.9; GFI> 0.8; TLI> 0.9 and RMSEA <0.05 3.4.5 Structural Equation Model analysis (SEM) 13 The indexes considered in SEM analysis include regression estimate and P- value (reflecting statistical significance). Bootstrap test was performed with the number of repeating samples N = 800 times. 3.4.6 Analysis of variance This study uses analysis of variance to evaluate the difference in average of travel decisions and average of travel motivation among groups in personal control variables. Analysis techniques include: 1) Testing averaging 2 groups (independent sample T-test); 2) Test by variance analysis, using ANOVA, Welch analysis. 14 CHAPTER 4. RESULTS OF RESEARCH 4.1 Context of outbound tourism market of Vietnam In recent years, Vietnam has experienced a fast change in international integration in many areas such as politics, socio-culture, security and defense, science and technology, education, etc. . Also, the outbound tourism market has experienced a strong growth based on the increasing international integration of Vietnam in several years. The supply of outbound tourism markets has also grown fastly. The context of the outbound tourism market together with the explosion of information technology has changed the consumption habits of Vietnamese tourists. Figure 4.1 Growth chart of the number of Vietnamese tourists going abroad in main markets 4.2 Qualitative research results Results from qualitative analysis help adjust the measurement scales such as Destination images; Travel motivation; and Personal control variable. It help confirms the appropriateness of the proposed model and adjusts the items of variables consistent with the research context of Vietnam. - 200,000 400,000 600,000 800,000 1,000,000 1,200,000 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 Thailand Malaysia Singapore Taiwan South Korea Japan United States Australia 15 In-depth interviews with experts also showed a high consensus about the calibration of the destination image variable scale. The adjustment of the scale of easy to access to the destination, now is convenient or contraits for visa/immigration of Vietnamese tourists. Adjust the indicator on the scale of Byon and Zhang (2010) to fit to Vietnamese tourists. In addition, in-depth interviews with experts show that there is a high consensus that in order to ensure the generalization and relevance of the research model proposed in the thesis. It requires to reduce the travel motivation factors into 4 key factors include: Discovering new things, gathering knowledge and experiences; share experiences with others; find out pleasure; and Self-concept. On the other hand, 100% of experts' opinions agree on content and indicators in the Reference Group scale. Although there are new concepts (eWOM) in the reference group's content, the opinions also agree on the suitability of the proposed research model of this element. Thus, the results of qualitative research can see that the selection of the most basic and most influential factors in travel decisions is word of mouth (WOM) and electronic word of mouth (eWOM). Perfectly consistent with this study. It is also consistent with the inheritance of previous studies by Gitelson and Kerstetter (1995); Murphy et al. (2007); Xiang and Gretzel (2010). The qualitative research results gave 100% feedback to support the addition of the “accompanying people” to reflect more clearly the level of influence on travel motivation and travel decisions of Vietnamese tourist. At the same time, this can be the basis for showing more clearly the differences between customer groups by comparing control variables with the average of other variables in the research model. Therefore, the question "Who accompanies you on this trip" is added in the additional information section of the questionnaire. 4.3 Quantitative research results 4.3.1. Results of data collection and sample survey The total of valid votes is 754 and it is described by criteria such as: age; sex; academic level; family status; monthly income; employment groups; foreign 16 language knowledge; living area ... the data is valid and meets the requirements for further analysis. 4.3.2. Results of Reliability Analysis by Cronbach’s Alpha Variable of Destination image: this variable scale is measured by 5 observed variables: DES1, DES2, DES3, DES4 and DES5, Cronbach's Alpha coefficient is 0.938 (> 0.6) and total variable correlation coefficient of observed variables in the scale is greater than 0.3. Observed variables are reasonable, all observed variables are accepted; Variable of Customer approach activities: after remove unreasonable variables, there are remaining variables: ADS1, ADS2, ADS3, ADS4 and ADS5. Cronbach’s Alpha coefficient is 0.915 and all is accepted; Variable of Reference group: scale of this variable is measured by observed variables SOC1, SOC2, SOC3, SOC4 and SOC5. The Cronbach’s Alpha coefficient is 0.917 and the correlation coefficients are> 0.3. All observed variables used for measurement are accepted; Variables of Discover new things, gathering knowledge and experience after being adjusted, including observation variables DIS2, DIS3, DIS4, DIS5, achieving internal consistency and being accepted in this study; Variables of Share experiences with others, set on observed variables retained by SHA2, SHA3 and SHA4; Variables of Find out pleaseare: judged based on FIN1, FIN2, FIN3 and FIN4 with Cronbach’s Alpha being 0.922 and correlation coefficients of variables> 0.3; Variables of Search and self-concept including variables SEL1, SEL2, SEL3 and SEL4 with Cronbach's Alpha of 0.933 and all observed variables in the scale with the correlation coefficient of the total variables are satisfactory (> 0,30); Variable of Travel decision: after removing unreasonable observed variable, Cronbach's Alpha reaches 0.941 (> 0.6) and the total correlation coefficient> 0.30, is accepted and retained in the research model in the next steps. 4.3.3 Results of exploratory factor analysis (EFA) 17 The combined exploratory factor analysis simultaneously all observed variables retained in the previous step which was conducted. The table of the analysis results shows that the KMO coefficient is 0.937 (> 0.6) and the coefficient of variance Sig. is 0.000 (<0.05), is statistically significant. The analysis results shown also that there are 6 factors with Eigenvalues value in the last factor reaching 1,671 (> 1) and the cumulative percentage reaches 65,886% (> 50%), with distinct values. The analysis results in the pattern matrix show that no observed variables have a factor loading factor <0.5 and none of the observed variables have a high factor load factor on many factors. From 42 observed variables converge on 6 representative factors. 4.3.4. Results of testing the conformity of the research model Results of CFA show that Chi-square/df (cmin/df) reaches 3,217 meeting the set criteria. CFI was reached 0.933, it reflects the compliance of the proposed model with the data set and compares with the fit of the other model with the data set used in this study. The results show that CFI degree ≥ 90% is satisfactory. TLI index of 0.927 (> 0.9) shows the initial suitability of the model that meets the requirements. Although the RMSEA in this study is 0.054, which approximates to a confidence level of over 95%, it is therefore accepted in this study. The CFA analysis results also show that the GFI is currently 0.862 satisfactory and accepted for application in this study. Table of Bootstrap analysis results at the 95% confidence level estimated from the original model and from the average of 800 other estimates gave the same or reliable model. 4.3.5 Results of Structural Equation Model (SEM) analysis The results SEM analysis with the conformity assessment indicators such as Chi-Square/df, RMSEA, CFI, GFI ... above show that the proposed model was accepted and achieved. The criteria of the suitability with the data collection used in this thesis Based on the SEM analysis result, the bolded part reflects regression weight between Reference Group (SOC) and Attitude (ATT) factor; and Reference Group (SOC) with Engine (MOT) has the highest coefficients. 18 Table 4.26 Estimated results of Structural Equation Model analysis Relationship Estimated coefficient S.E. C.R. P ATT <--- DES 0,134 0,047 2,833 0,005 ATT <--- SOC 0,188 0,040 4,644 *** ATT <--- ADS 0,185 0,056 3,300 *** MOT <--- ADS 0,163 0,045 3,599 *** MOT <--- SOC 0,336 0,034 9,732 *** MOT <--- DES 0,081 0,038 2,127 0,033 DEC <--- MOT 0,274 0,038 7,183 *** DEC <--- ATT 0,274 0,038 7,262 *** The results helps confirm that the regression estimated of first-order factors including FIN, SEL, SHA and DIS with second-order factor (MOT) is relatively large. All P values <0.05 showed that there was statistical significance. Therefore, it can be confirmed that the scale for the second-order tourism engine is accepted in the research model of this thesis. 4.3.6 Test results of research hypotheses The results table of estimating SEM analysis shows that the smallest regression in this model is 0.081 and the largest is 0.334. There is no value of P> 0.05, which means that the hypothesis is accepted at 95% of the data. All hypotheses including H1, H2, H3, H4, H5, H6, H7, H8 are accepted. 4.3.7 Results of variance analysis The results of evaluating the average difference of quantitative variables among control groups show that there are specific differences as follows: On average the decision to travel abroad in the 18-34 age group is greater than the 35-44 age group; 45-60 years old; and the group is over 60 years old; 19 The average of travel motivation of family with young children is bigger than family with full-grown member; higher education group whose average decision to travel abroad is greater than that of ordinary education group; The freelance individuals group is lower than that of state employment group; The group of customers earning 10-20 million/month higher than the group of 21-30 million/month; The group of tourists traveling in Asia had an average decision to travel abroad higher than the rest; The group of tourists participating in the tour for more than 9 days is lower than the group of 4 days, 5 days and 6 days. 20 CHAPTER 5. DISCUSSION 5.1 Summary of research results The proposed theoretical model is established based on inheriting the research model of Ajzen (1991); Um and Crompton (1990); Woodside and MacDonald (1994); Decrop (2006b). The study identified new factors (eWOM). The results of the CFA test help confirm the suitability of the model compared to the data set used in this study. Analysis of SEM in this thesis confirms the degree of correlation between the factors. The results evaluate the average difference of quantitative variables among control groups 5.2 Discuss the research results 5.2.1 The impact of Destination image factor to decision-making The attraction of destinations is created from tourism resources such as indigenous cultural, natural landscapes, the safety of destinations ... indirect effects on decisions of outbound tourist through psychological factors in each individual such as attitude and travel motive. Hypotheses H1, H2, H7, H8 are tested and accepted showing that Destination image is still one of the key factors affecting the decision to travel abroad. 5.2.2 The impact of customer approach activities to decision-making The findings of this study show that the hypotheses reflect the impact on travel decisions o