1. Introduction

Purchase decisions are very important in business, especially with consumer goods. Korean styles and products appearing in popular Korean dramas, or the image and actions of famous idols greatly affected the decision to buy Korean consumer goods of generation Z. Currently, there have been consensuses regarding the ending birth-year of this group but most demographers and researchers used. Gen Z as those who had been born after 1993 (Statistics Canada, 2011), and after 2001 (Philippine Retailers Association, 2016). In this study, Generation Z in Vietnam is referred to a group of people born between 1990 to 2005. In 2016, Vietnam had 16.621 million people in generation Z, accounting for 17 percent of the Vietnam population. They spend a lot of time on social media and are affected by Hallyu and products whose country of origin is Korea. The decision to buy Korean consumer goods of Vietnam people is influenced strongly by Hallyu and social media (Nguyen, 2018). However, this problem has not been fully studied and applied. In 2018, Korea exported USD 48.6 billion worth of goods to Vietnam. Despite the increased number of Korean consumer goods in Vietnam, little research has investigated the reasons leading to the purchase decision of Korean products. To develop a useful marketing strategy, it is important to examine the relationship between different factors and the buying decision. The results provide Korean firms and marketers with valuable and important insights to develop appropriate marketing strategies.

2. Literature Review

2.1. Social Media Usage (SM)

Social media has become a way of life for youngsters and personal activities are made public (Edge, 2017). According to Ba and Pavlou (2002), more positive comments, feedback, and higher ratings lead to a higher level of trust in a vendor. Summarised the previous results research showed the quality and quantity of information produced by customer reviews influence intention to buy through increased trust. The research result of Pan and Chiou (2011) showed that social media significantly affects the perceived trust of consumers. The interactions on social media generated influences trust (Weisberg et al., 2011). Consumers reviews on social media are perceived to be useful and affect attitude and their intention to buy through the impression created about a product or service(Purnawirawan et al., 2012). Monfortiand Marichal (2014) stated that social media had a positive effect on trust. Base on that, this study established hypothesis follows:

H1: Social media has a positive influence on trust.

Some studies were carried out on students’ usage of social media, its impact on their social norms, education, and academic performance (Al-Sharq, Hashimand Kutbi, 2015). Social media is used by millions of people, with teens and young adults making up the largest user demographic. According to McLauhlinand Vital (2011), in social space, social norms provide structure and guidelines on what is healthy and unhealthy behavior. The social norm is established among different social groups and that norms become represented in societal rules and laws. According to Feiler (2014), the users often feel social media anxiety and the need for popularity, which makes social media impact on the social norm. Hence, the research postulates this hypothesis:

H2: Social media has a positive influence on social norms.

Perceived usefulness is defined as the degree to which a person believes that using a particular technology would enhance his or her job performance (Davis et al., 1989). According to Gefen (2002), previous studies had found a positive connection between information technology acceptance and perceived usefulness. Usefulness for sharing problems is one of the predictors to social media adoption(Quan-Hasse & Young, 2010). Quality of information is one of the usefulness factors that influence the adoption of social media (DeLone & McLean, 2003). From the results of previous research, the hypothesis:

H3: Social media has a positive influence on perceived usefulness.

The unique features, platform enforcement together with the nature of social media have created an environment in which healthy social norms are not ubiquitously followed. Currently, in Vietnam, most young people are using social media and they are influenced by the hot trend on these channels. According to Laroche, Habibi, and Richard (2013), social media help individuals commentary on products, respondents to new products and demonstrate a relationship to the community. Consumers may comment or engage in discussions about a product on social media. Firms must create a marketing plan based on Effective measuring of consumer engagement in product brand-related social media content (Schivinski, Christodoulides, & Dabrowski, 2016). Besides, the research conducted by Nguyen (2018) showed that social media also influenced Korean products purchase decision of generation C in Vietnam. Base on that, this study established hypotheses as follows:

H4: Social media has a positive influence on product quality perceived.

2.2. Hallyu (HA)

Hallyu is a Korean cultural production trend that comprises all Korea-related fields. Consumer products related to Korean culture such as tourism, contents, cosmetic products could be greatly influenced (Yu & Kim, 2012). Lee (2007) found the positive and significant relationship between the Korean Wave and attitude toward Korean products. According to Kim Bok-rae (2015), Hallyu 1.0 emerged with K-drama exports in the late 1990s, then came Hallyu 2.0 with K-music, Hallyu 3.0 with K-culture, and Hallyu 4.0 with K-style. Overall, Vietnam had got over 64 million internet users (48,1% population) as of December 31, 2017 (Internetworldstats, 2018) and 58 million Facebook users as of April 2018. Generation Z in Vietnam are influenced by Hallyu and social media and they usually search for information, prospective reviews from other people through social networking sites. They have been blown away by the new wind of Hallyu which inspires them to become younger, more dynamic and active. Hallyu 4.0 had a positive impact on Korean products purchase decision of generation C in Vietnam (Nguyen, 2018). Therefore, there were hypotheses:

H5: Hallyu has a positive influence on attitude toward behavior;

H6: Hallyu has a positive influence on buying intention.

2.3. Country of Origin Effect (CO)

The country of origin effect was referred to as the made in the image and the nationality. The CO would still be an important driver for building brand image, and it would affect purchase intentions even indirectly (Diamantopoulos et al.,involvement all had a significantly positive effect on consumer purchase decision (Lin & Chen, 2006). The CO image had a direct effect on product involvement, which in turn influenced the perceived product’s quality and purchase intention (Afnan, & Md. Hossain, 2015). Ahmed and Astous (2007) stated that respondents who were highly involved in automobiles were able to make better distinctions between the country of origin than fewer involved ones. Vietnamese Generation Z expresses their interest in Korean goods. Thus, this has contributed to the rapid increase of Korean exports to Vietnam. Vietnam is Korea's third-largest export market after China and the United States in 2017. In 2017, the imported goods from Korea to Vietnam reached 46.7 billion, 45.3% more than that of 2016. This study, hence, set a hypothesis:

H7: Country of origin has a positive influence on Product involvement.

2.4. Product Involvement (PI)

Product involvement was described as the general level of interest in the object or the centrality of the object to the person’s ego structure. Results of previous researches debated that PI might positively or negatively impact the correlation with CO. To be specific, PI was said to be affected by country of origin effect (Chattalas et al., 2008; Cilingir & Basfirinci, 2014). Maheswaran (1994) stated that CO information would be more important to those who were purchasing lower involvement products. On the other hand, Petty et al. (1983) showed that consumers tended to utilize a central approach to high involvement conditions and chose a peripheral approach in low involvement conditions. According to Celsi and Olson (1988), higher involvement with a product would lead to searching, utilizing and paying more attention to the cues of products before evaluation and purchase. This study then came up with a hypothesis:

H8: Product involvement has a positive influence on Perceived quality.

2.5. Perceived Quality (PQ)

Perceived quality was defined as consumer findings of a product’s performance and how this product was like in comparison with their expectation. Chaudhuri and Holbrook (2001) stated that the focus of brand superiority was based on findings that there was a strong positive relationship between brand superiority and brand purchase. According to Grunert (2005), perceived quality became the estimation relying on the whole sets of basic as well as the outer dimension of the service or product. Intention buying was directly influenced by product perceived quality. Customers usually have some perceptions in their minds about the quality, styles of a product or service, and price before entering the real purchasing stage. With the above arguments, this study looked forward to exposing a direct influence between perceived product quality and purchase decision of Vietnam’s generation Z through the mediating constructs of buying intention. Hence, there was a hypothesis:

H9: Perceived quality has a positive influence on the purchase decision.

2.6. Social Norms (SN)

Social norms provided structure and guidelines on what behaviors were appropriate and healthy (MacLaughlin and Vitak, 2011). The theory reasoned action (TRA) showed that if people evaluated the suggested attitude as positive, and if they thought that their significant others wanted them to perform the behavior (subjective norms), this resulted in a higher intention (motivation) and they were more likely to undertake that behavior. Many studies had shown a high correlation of subjective norms and attitudes to behavioral intention, and subsequently to behavior (Sheppard; Hartwick; Warshaw, 1988). Social factors played an important role in influencing consumer purchase decisions. From some results research above, it can be summarized that social norms, in general, are an individual’s set of perceptions of whether they should or should not do the behavior, taking social pressure and opinions of important persons into consideration. In addition, the TPB and TRA can explain the individual's social behavior by considering social norm as an important construct. Consumer purchasing decisions were influenced by their social norms. Therefore, the specific hypothesis to be tested was:

H10: Social norms have a positive influence on Buying intention.

2.7. Perceived Usefulness (PU)

Perceived usefulness was the degree to which a person believed that using a particular system would enhance his or her job performance (Davis et al., 1989). There were different aspects that defined perceived usefulness in an e-commerce environment. For instance, the quality of a website in terms of system, information quality and service had a significant impact on perceived usefulness, which in turn encouraged consumers to buy. With the rise of social media and online communities, individuals could easily share and access the information (Wu et al., 2010). Social media were where consumers made social connections and participate in cyberspace (Mueller & Fueller et al., 2011). Social media placed a positive impact on perceived usefulness. Perceived usefulness affected users’ intentions to use e-commerce (Gefen, 2002). The more consumers shopped in SNSs that they perceived to be useful, the more²⁰¹¹⁾. The CO image, product knowledge, and product they intended to buy in SNSs (Jiyoung, 2009). According to Nguyen (2015), perceived usefulness had a positive impact on the intention to buy a product. Hence, the study formulated the following hypothesis:

H11: Perceived usefulness has a positive influence on Attitude.

2.8. Trust (TR)

Trust derived from reciprocity, which can be learned only in cooperation with others. In the relationship between business and consumers, trust in an electronic supplier was very important when assessing the effectiveness and risks of a transaction (McCole & Williams, 2010). Trust and perceptions of safety on social networking sites played an important role, impacting attitudes and purchasing intention (Jiyoung, 2009). Members of the network community exchanged information and experience with each other, therefore trust and willingness to buy were motivated. The trust in online transaction affected the online purchasing intention (McCole & Williams, 2010) and increased purchasing intention (Shin, 2010). The trust was a deciding factor in the consideration stage (Gefen, 2002). Yoon (2002) stated that both online and offline trust had had a great impact on the consumer’s purchase decision. Trust strongly affected specific purchasing intentions and also had a strong influence on three key consumer intentions (Nguyen, 2015). Trust could affect the intention to buy (Mauricio & Paul, 2003). Therefore, this study set out the hypothesis:

H12: Trust has a positive influence on Buying intention.

2.9. Attitude Toward Behavior(AT)

Attitude defined the reviews, thoughts, relatively consistent tendencies of humans against an object or an idea. It puts humans into the framework to think about things they like or don't like. The theory of reasoned action (TRA) and the theory of planned behavior (TPB) were developed focusing on trust and attitudes as determinants of consumer buying behavior. Both TRA and TPB proposed that a person’s overall attitude toward an object was derived from his or her beliefs and feelings about various attributes of that object. The relative contributions of attitudes and subjective norms would not necessarily be equal in predicting behavior. Actual behavior was again derived largely from behavioral intention but was mediated to some degrees by perceived behavioral control (Ajzen, 1991). Consumers often buy when they trust and have a positive attitude to a seller. Attitude has a long-term trend based on the values and beliefs of people and is difficult to change. Attitudes affect buying intentions, thus posing the hypothesis:

H13: Attitude positively influences buying intention.

2.10. Buying Intention (BI)

Purchase intention was defined as the probability that customers in a certain purchasing situation chose a certain brand of a product category (Crosno et al., 2009). Intention mentioned a willingness to make a purchase decision. The theoretical action model (TRA) and the technical acceptability model (TAM) showed that there was a high correlation between intention and actual use. Tirtiroglu and Elbeck (2008) suggested that purchasing intention should be described as customer willingness to buy a product, anticipating purchasing intention as the initial step to anticipate actual customer acquisition behavior. The positive relationship between intention and purchase decision was described by (Mauricio & Paul, 2003 & Ajzen 1991). The study conducted by Sommer (2011) showed that behavioral intention referred to individual motivation and conscious plan to expand the effort and perform a behavior. The results of studies by Nguyen (2015 and 2018) suggested that intention toward buying positively influenced purchasing decisions, which we used to test the following hypothesis:

H14: Buying intention has a positive influence on the Purchase decision.

2.11. Purchase Decision (PD)

The purchase decision is the final stage of buying behavior. It is defined as the act of buying or decision-making without having to worry about anything, just waiting for the receipt and payment. Nowadays, thanks to the information and communication technology, many people find information about products, prices, brands to predetermine and make a purchase at the right time. Research results have shown that the majority of respondents agree to purchase if the purchase is convenient or useful. According to Akhter (2010), purchasing intentions had a significantly positive effect on the actual purchasing decision.

3. Conceptual Model

In order to achieve the study’s aims, the conceptual model was developed based on the theory of reasoned action – TRA, theory of planned behavior – TPB, technology acceptance model, the trust model (Mayer, Davis, & Schoorman, 1995) and previous research (Table 1), this conceptual model is modified by applying 11 constructs to find out the impact of independent constructs: Hallyu, social media, country of origin through intermediate constructs (social norms, trust, perceived usefulness, product involvement, perceived quality, attitude toward behavior, buying intention) on purchase decision (Figure 1).

Table 1: Summary of Theoretical Background and Preliminary Research

Figure 1: Conceptual Model

4. Methodology

A mixed-method approach was utilized. Focus group of 9 experts has carried out a semi-structured questionnaire for exploring constructs and testing content validity by CVR (Laewshe, 1975). In this study, the measurement items in the questionnaire were inherited and adopted from previous research, so it needs to be explored and reconsidered to increase the validity. An in-depth interview with 27 participants has carried out a semi-structured questionnaire for exploring measurement items and content validity by CVR. A pilot study to test reliability measurement items was delivered by cross-sectional questionnaires with a sample of 115 respondents. The official study applied cross-sectional with 439 respondents used testing the reliability, the validity of the measurement items, testing hypotheses and the final model by EFA, CFA, and SEM.

For the analysis and testing, SPSS 20 and AMOS 20 software were used. This research population is approximately 16.621 million people in generation Z, accounting for 17 percent of the total Vietnam population. Consistent with the purpose of this study, target respondents to be sampled were carrying out all the population distribution in the north, in the central and in the south in Vietnam. Therefore this study applied non-probability with convenience sampling. Data was collected using both face to face and a self-administered questionnaire distributed electronically. The questionnaire applied a 5-point Likert scale (1= strongly disagree; 5= strongly agree). The Bootstrap was used for estimating the model parameters for retesting reliability of the estimates.

5. Results

5.1. Preliminary Research

The original conceptual model with 11 constructs and 55 measurement items was developed from some theoretical background and previous research then explored and retested with the focus group and in-depth interview. The result showed that 11 constructs and 55 measurement items have gained content validity to be further analyzed.

5.2. Pilot Study

The result pilot study showed that Cronbach's alpha of 11 constructs were more than 0.6, which was acceptable reliability. There were only 2 measurement items (CO5 and PI5) had corrected item-total correlation < 0.3. Therefore, 51 remaining measurement items were accepted, ready to continue in the official study.

5.3. Official Study Descriptive Analysis Result

There were 460 respondents, in which 21 invalid respondents were eliminated and 439 respondents were retained for analysis (respond rate 95.43%). There were 197 boys (44,87%) and 242 girls (55,13%). The result showed that 100% of respondents belonged to Generation Z (born between 1990 and 2005) and purchased Korean consumer goods through two screening questionnaires. Almost 78,79% of the respondents’ education background was the college/university level and above. They purchased fast-moving consumer good ranked first with 64,92%, followed by fashion goods with 23,92% and other items with 6,38%.

Reliability Test

A higher Cronbach’s alpha coefficient indicated greater reliability and its acceptable lower limit was 0.6. In this study, the Cronbach alpha coefficient of constructs was from 0.647 to 0.829. The measurement items: PI3; PI4; SN5; IB5; and PD3 had corrected item-total correlation < 0.3 so they were removed from the study and 46 measurement items remained to go through EFA.

Exploratory Factor Analysis (EFA)

In this study, an extraction method principal axis factoring with Promax rotation was used. The result: Kaiser-Meyer- Olkin measure of sampling adequacy (KMO) = 0.782, Sig. =0.000. Pattern Matrix had 11 factors (Table 2) and the SN3; TR3; PU5; PQ1; AT3 and PD2 were not accepted because the factor loadings were < 0.5. Factor loadings of the indicators for each construct were statistically significant and sufficiently high to demonstrate that the indicators and their underlying constructs were accepted. The conceptual model with 11 factors with 42 measurement items was qualified for confirmatory factor analysis.

Table 2: Pattern Matrix

Kaiser-Meyer-Olkin Measure of Sampling Adequacy = .782;

Sig. = .000

Extraction Method: Principal Axis Factoring. Rotation Method: Promax with Kaiser Normalization.

Confirmatory Factor Analysis (CFA)

The results showed an acceptable fit of data: Chi-squared =1265.377; df=759; Chi-square/df=1.667; P=0.000; CFI=0.920; TLI=0.910; IFI=0.922; GFI=0.881; AGFI=0.862. RMSEA=0.039. Although GFI and AGFI < 0.9, but other fit mode indicators were satisfied so it is acceptable consistency. All factor loadings were highly 0.5 significant on the expected constructs. Together, these demonstrated an adequate convergent validity of the measures. Composite reliable and variance extracted were calculated on the basis of standardized weights estimated in the CFA model. Before the hypothesis test, the constructs were tested for two psychometric properties reliability and validity.

Reliability: According to Bagozzi and Kimmel (1995), a factor displayed its reliability if its composite reliability was greater than 0.6. The Cronbach’s alpha coefficient retested of all constructs in this study were from 0.615 to 0.844 (Table 3). Therefore, 11 constructs have achieved reliability.

Figure 1: The Confirmatory Factor Analysis

Table 3: Construct Test

Validity: The research tests validity through both content validity and constructs validity. In content validity, the study has checked both by focus group and in-depth interview. The result showed that all measurement items and contructs have achieved content validity. This study has checked construct validity both by Composite reliability and Average variance extracted. Although variance extracted few constructs (SM, HA, PU, AT and PD) weighted nearly 0.5, the composite reliability of all constructs was 0.615 to 0.852. Therefore, it is considered all constructs have achieved validity (Table 3).

Convergent validity: Convergent validity can be ensured by assessing discriminant and convergent validity. Measurement items achieved convergence validity if standardized regression weighted > 0.5 and regression weighted significant p < 0.05. The results of this study showed that all standardized regression weights of 34 measurement items were from 0.542 to 0.933, larger than 0.5, therefore they were convergent validity. This model of measurement items was consistent and there was no correlation between the measurement error should it achieved unidimensional. The correlation coefficient between constructs with others had to be < 0.9 or p-value had to be< 0.05, the new gained discriminant validity (Table 3).

Structural Equation Modeling (SEM)

SEM allowed the combination to be the underlying concept with our measurement and could consider the measure independently or in combination with theoretical models at once. In addition, SEM is a comprehensive statistical approach for testing hypotheses about relations between observed and latent construct.

It combines features of factor analysis and multiple regressions for studying both the measurement and the structural properties of theoretical models. SEM methodology accounted for independent construct errors and model multiple relationships simultaneously, which resulted in more powerful tests of mean differences.

The first SEM order, parameter HA to BI have p-value =0.413 (above 0.05) therefore, it was deleted from the model. Results of SEM second-order obtained for model showed excellent fit (Chi-square/df = 1.717; P=0.00; CFI=0.910; IFI=0.911; TLI=0.903; GFI=0.870; AGFI=0.856; RMSEA=0.040). Although GFI and AGFI < 0.9, other fit mode indicators were acceptable consistency. Therefore, we could proceed to examine the path coefficients of the structural model. The results show that all the parameters in the SEM model for the P-value were less than 0.05 so the relationships were significant. All of the constructs were positively correlated, except intention (Table 4).

Table 4: Regression Weights

(1) The indicators in line BI <--- HA results in first SEM order

Bootstrap Estimate Model

This study used methods Maximum Likelihood (ML) to estimate the parameters in the models of SEM. Usually in the quantitative research, so as to test the reliability of the estimates, researchers have to divide it into two samples. The first half is used for estimating the model parameters and the other half is for retesting. The other way is to repeat the study using a different sample and the Bootstrap method was suitable. This study used the bootstrap method with the number of repeated sample N=500 to test the model in practice. The CR (bias/SE-bias) absolute value was< 2.0; therefore, the estimates could be trusted and fit.

Model Fitting and Hypotheses Testing

Yu, H. C. Kim, and H. Kim. (2012). Investigating Cosmetic Purchase of International Tourists in South Korea. International Journal of Management Cases, 14(1), 398-410.

The conceptual model and fourteen hypotheses were tested using regression weights in SEM. The results, shown in (Figure 3) and (Table 4), indicated the value of the fit model and that the overall constructs were supported. 13 hypotheses were significant with significance at less than the p < 0.05 level, except for H6. Overall, all the path coefficient-related hypotheses were supported from 0.107 to 0.548. It is shown that social media positively influences significant to trust, social norms, perceived usefulness, and product quality perceived. This is consistent with previous research (Monforti & Marichal, 2014; Al-Sharq, Hashim, & Kutbi, 2015; McLauhlin & Vital, 2011; Gefen, 2002; Quan- Hasse & Young, 2010; DeLone & McLean, 2003; Schivinski, Christodoulides, & Dabrowski, 2016; Nguyen, 2018).

Figure 3: Structural Equation Modeling (Second order)

The research also provides some implications for Korea firms. The result of research has shown the key role of Hallyu, Country of origin had strong impacts on purchase decision through Attitude and Product involvement. This is consistent with previous research (Yu & Kim, 2012; Lee, 2007; Nguyen, 2018; Diamantopoulos et al., 2011; Lin & Chen, 2006; Afnan, 2015; Ahmed & Astous, 2007). The results highlight the relevance of TRA theory, TPB theory, TAM theory and consistent with previous research on the behavior buying, the study shows that the Social norms, Trust and Attitude positively influences towards buying intention. The result of his research showed that intermediate constructs (Perceived usefulness, Trust, Attitude, Product quality, Buying intention) were affected by Hallyu, Social media, Social norms and positively influence to Purchase decisions. This result like other studies from Chaudhuri and Holbrook (2001); Grunert (2005); Tirtiroglu and Elbeck (2008); Mauricio and Paul (2003); Sommer (2011); Nguyen (2015, 2018).

Multiple-Group Analysis

The multiple-group analysis was applied to investigate the effects of the Hallyu, social media, country of origin and mediating constructs on Korean goods purchase decision of generation Z in Vietnam. Data showed that the demographic factor was divided into Boy and Girl group. AMOS assumed that the baseline model (our original model) was true. The model (equal loading model) that specified a group-invariant factor pattern, was supported by the sample data. ML estimation method was used in the analysis of the multi-group model. The test was used to compare between 2 invariances and variance across groups. The results of constructs and invariant multivariate testing were divided into two groups of boy and girl, the chi-squared difference value of two models (non-variance) is 212.677 with df=113 degrees of freedom (chidist (Δ chi-squared, Δdf)=0.000 below than 0.05 so that invariance multiple models were used. The resulting test structural weights witnessed the same results between the boy group and girl group, which were: SN did not influence BI and SM did not influence PQ. The difference was, for the boy group, TR did not influence BI; for the girl group, PU did not influence AT and PI did not influence BI.

6. Conclusions

Nowadays, culture is very important to do business in the global market. This study was developed based on previous theoretical and empirical studies and the theories of Hallyu, social media, country-of-origin and how they influenced the purchase decision of Generation Z in Vietnam. To summarize, it could be concluded that Hallyu, social media and country of origin impact purchase decision through mediating constructs such as social norms, trust, product involvement, perceived quality, perceived usefulness, attitude and buying intention. The constructs of individual including Hallyu, social media, and country of origin are believed to positively moderate trust, social norms, perceived usefulness, product involvement, perceived quality, and attitude. The hypotheses generated from the model were tested. Thirteen hypotheses were accepted and one was rejected. There is a difference between the male group and the female group. For the male group, TR did not influence BI, while PU affected AT, whereas for the female group, TR influenced BI, while PU did not influence AT and PI did not influence BI.

The results revealed that the purchase decision for Korean goods of Vietnam’s generation Z was influenced by Hallyu, social media and country of origin impact purchase decision through mediating constructs such as social norms, trust, product involvement, perceived quality, perceived usefulness, attitude and buying intention. These results confirmed factors identified by previous research studies. Therefore, Korean firms exporting to Vietnam in particular and those who base in Vietnam, in general, should take into account the essential roles of the above factors, in order to understand and respond promptly to the demands and expectations of generation Z in Vietnam. The results of this study can be a valuable source of information for marketing managers to implement strategies and plans, to not only attract more generation Z but also enhance their desires and encourage them to keep buying Korean goods. One of the important implications of this paper is that Korean firms should take advantage of and increase the influence of Hallyu, country of origin, social media to better reach and motivate the buying decision of Vietnam’s generation Z.

7. Limitations and Future Research

The limitation is the use of cross-sectional design so it may be influenced by the Hallyu at the time of research. These limitations require further research investigation in order to discover the underlying influence for time series. Another limitation of this research is that it did not yet seek to differentiate between different groups in demographics, geographic areas, etc. Nevertheless, it would be interesting topics for future research.