Understanding online purchase intention: the mediating role of attitude towards advertising

Abstract Recognizing the growth of Facebook and advertising on this platform in an emerging market, we conducted a study on the impact of Facebook advertising characteristics on the market’s purchase intention. This study attempts to identify and test key factors that have a direct impact on customer online purchase intention for advertised products, and examines the mediating role of attitude towards advertising. In addition, for factors that have been shown in previous studies to have a direct impact on consumer purchase intention, such as entertainment, credibility, and personalization, the direct effects are also included in the model for further consideration. This study follows a quantitative approach through interviews with 477 participants. The questionnaires were distributed to the respondents by interviewers at milk tea and coffee shops in Ho Chi Minh City, Vietnam’s largest city. Data were analyzed through structural equation modelling. Empirical results have shown an indirect effect of informativeness, privacy concerns, and irritation on purchase intention through attitudes towards advertising, however, the factors interact, personalization, entertainment, and credibility do not show an impact on attitudes towards advertising. Finally, factors such as entertainment, personalization, and credibility of Facebook ads are also found to have a significant positive effect on purchase intention. This study was successful in developing an empirical model for understanding online purchase intention with the mediating role of attitude towards advertising in emerging markets, specifically Vietnam. Therefore, the study provides essential insights for firms and marketers investigating issues related to online purchase intentions, and attitudes towards advertising.


PUBLIC INTEREST STATEMENT
Recognizing the growth of Facebook and advertising on this platform in an emerging market. This study was conducted to build a user model for Facebook ads in Vietnam which is an emerging market. It aimed to identify and examine the main Facebook advertising features that could predict the customer's purchase intention for the products that are promoted using social media advertising. The study provides essential insights for firms and marketers investigating issues related to online purchase intentions, and attitudes towards advertising.

Introduction
The growth in social network use is evident not only in developed countries but also, in fact even more so, in developing countries (Wiese et al., 2020). The number of social media network users who use it at least once a month reached approximately one-third of the world's population in 2017 (E-Marketer, 2017), Facebook alone had 2.23 billion monthly users as of 2018 (Statistica, 2018). Thus, Wiese et al. (2020) deduced that Facebook is the world's foremost social network service (SNS) and a tool that has become an integrated part of consumers' lives. So, social networks have spread to all aspects of human life, changing the way we interact, which makes organizations start thinking about how to use social networks to attract customers as well as build profitable marketing relationships with those customers, and in a wide range of social media marketing activities, advertising is one of the activities that receive the most attention from both researchers' and practitioners' perspectives (Alalwan, 2018). Berthon et al. (1996) were one of the first researchers who investigated the way advertising might operate on the Internet. Recently, the Internet has been regarded as the most powerful advertising medium (Mahmoud, 2014) because they are likely to feel less irritated, more entertained, and well-informed about the advertised product. Social media advertising represents the superiority of company-customer communication in comparison with traditional mass media advertising or online advertising (used for Web 1.0 applications), companies can communicate more information and interact (two-way) with their customers thanks to the interactive nature of web 2.0 (Alalwan, 2018). Zhu and Kanjanamekanant (2021) proposed that users perceive personalized ad content (which is one of the notable characteristics of social media ads) as more appealing and more connected to their interests and have higher purchase intentions for the product recommended in personalized advertisements. Firms reportedly spend up to 50% of their total marketing budgets on personalized advertising. The targeting capabilities of social networks like Facebook help reach favorable consumers' attitudes towards the ads because of possible efforts on reshaping public beliefs positively about it (Mahmoud, 2014).
The main purpose of social media (such as Facebook) marketing activities is to shape the consumer's decision-making process (Alalwan, 2018), in which, purchase intentions are key predictors of purchase behavior (Kotler & Keller, 2016). Consumers' hesitation to purchase which arises from various problems in online shopping becomes an obstacle to e-commerce growth, and therefore, a better understanding of purchase intention is essential to understanding consumer behavior to improve the company's business (Dachyar & Banjarnahor, 2017). Studies on purchase intention are also claimed by Peña-García et al. (2020) as of the utmost importance for the success of any online retailer.
As the growing trend in businesses worldwide is to globalize, firms established in developed economies would benefit from entering emerging markets to expand their operations and maintain global competitiveness (Paul, 2020). However, most of the hitherto research mostly focuses on social media attitudes and practices in developed countries; it neglects other emerging countries, which might be considerably different in determining factors (Chetioui et al., 2021), while the difference between emerging markets and other markets is a key cultural difference among customers from different backgrounds (Banerjee & Chai, 2019). Therefore, further research needs to be conducted on these emerging markets to fill the gap in understanding these differences.
Vietnam ranks 4th among the top 11 Southeast Asian countries, measured by purchasing power parity-adjusted nominal gross domestic product in 2020 (General Statistics Office of Vietnam, 2021). Therefore, Vietnam can represent an emerging market in Southeast Asia. According to an article published on Hr1tech (2021), statistics show that by June 2021, the total number of Facebook users in Vietnam was nearly 76 million, accounting for more than 70 percent of the national population. This is an increase of 31 million users from 2019 and leads the list of popular social networks in Vietnam. Statistics also show that, on average, Facebook has one new Vietnamese user every three seconds. During the COVID-19 pandemic, Facebook users' activities mainly focused on posting entertaining videos/photos (with young users) and online shopping (with users aged 24 and over). Moreover, Vietnam is considered a popular place to use Messenger for online promotion, selling, and taking care of customers, much more so than the other countries in the region (Hr1tech, 2021). Although the rapid growth of Facebook as well as the great potential of advertising on this platform, but research on this topic in Vietnam is still quite limited.
As the growth of emerging markets offers great opportunities to develop or discover new perspectives and practices in marketing, studies for these markets are important in business (Paul, 2020). Recognizing the growth of Facebook as well as advertising on this platform in Vietnam, an emerging market, we have conducted a study on the impact of Facebook advertising characteristics ("the world's foremost social network service", Wiese et al., 2020) to purchase intention for this market. This study attempts to identify and examine the main Facebook advertising features that could predict the customer's purchase intention for the products that are promoted using social media advertising. The impact of these features on purchase intention is examined mainly through consumers' attitudes towards Facebook advertising. Besides, for the factors that have been proven in previous studies to have a direct impact on consumer purchase intention such as Entertainment, Credibility, and Personalization (eg. Alalwan, 2018;Kurtz et al., 2021;Lee et al., 2017), the direct effects are also included in the model for further consideration.

Theory of planned behavior (TPB)
The theory of planned behavior was elaborated for the purpose of improving the predictive power of the theory of reasoned action, which aims to address the limitations of the theory of reasoned action model in dealing with behaviors over which the individual had incomplete volitional control (Ajzen, 1991). Like the theory of reasoned action, the theory of planned behavior has the overall objective to predict and explain human behavior in specific situations (Fortes & Rita, 2016).
The theory of planned behavior assumes that individuals' attitudes predict their intention to perform a particular behavior. In other words, when an individual has a stronger intention to perform a certain behavior, the likelihood of the person performing that particular behavior accordingly expands (Chetioui et al., 2021). The theory of planned behavior has been used to explain and predict human behavior in various contexts, including the adoption of interactive communication technologies . This is also one of the most used consumer behavior models in the literature (Fortes & Rita, 2016) and has been widely used in the context of e-commerce behavior (Lim & Dubinsky, 2005). media communication (Azam, 2015). Audience activity is central to U&G theory research and communication motives are key aspects of audience activity (Menon & Meghana, 2021).
The theory considers not only the pleasure people search for in media but also the attitudes of the audience towards the medium and its content (Roy, 2009). On the other hand, Among the Social networking sites, Facebook has emerged as the most popular and largely used medium for digital content creation and distribution (Menon & Meghana, 2021). This theory contributes to the understanding of mass communication processes (Roy, 2009). Therefore, the theoretical framework on which the present study is principally evaluated is based on the U&G theory.

Purchase intention
Purchase intention indicates the likelihood that consumers will plan or be willing to purchase a certain product or service in the future (Wu et al., 2011). In the advertising context, purchase intention refers to the process by which consumers make a plan to purchase a product/service due to advertisements delivered through advertising media (Lee et al., 2017). Online purchase intention in the study of Peña-García et al. (2020) is understood as the degree to which a consumer is willing to buy a product through an online store. Martins et al. (2019) have demonstrated that an increase in purchase intention implies an increase in the chance of purchasing, which means if consumers have a positive purchase intention, then a positive engagement will promote that purchase.

Advertising on Facebook
Social media can be described as a group of internet-based applications that build on the ideological and technological foundations of Web 2.0 and enable the creation and exchange of User-generated Content (Kaplan & Haenlein, 2010). Taylor et al. (2011) characterized social networks as platforms that share commonalities such as providing users with the ability to create profiles within a limited system, to maintain lists of "friends" with whom they share connections, and to view and browse their own lists of connections and of others. The rapid development of social media has changed the way consumers choose, share and verify information, thereby, reducing the influence of traditional advertising channels (Duffet, 2015). Through the wide reach of social media, users' reviews or comments have a much greater word-of-mouth impact than ever before (Yilmaz & Enginkaya, 2015). Social media advertising is becoming an increasingly popular tool for marketers (Logan et al., 2012).
Facebook gets special attention from marketers because of its number of users and its popularity among young people. In addition, Facebook gives advertisers many advertising options, such as driving traffic to particular websites or promoting a particular Facebook page, event or app. All of Facebook's options offer the opportunity to add a "like" link, allowing consumers to instantly give feedback (Logan et al., 2012).

Attitude towards advertising
Attitudes toward advertising are defined as a predisposition to respond favorably or unfavorably to a particular advertising stimulus during a particular exposure occasion (Lutz, 1985). An attitude toward an ad is formed when the consumer evaluates the visual and/or verbal substance of the claim and its content (Sander et al., 2021). Consumers' attitude toward advertising is one of the influential indicators of advertising effectiveness because consumers' cognitive ability toward advertising is reflected in their thoughts and feelings, which subsequently influence their attitude toward advertising (Nagar, 2015). In the internet environment, Mahmoud (2012) illustrated attitudes towards Web advertising as a general disposition to like or dislike advertising messages delivered online.
Advertising creates both positive and negative emotions, which in turn, affects the overall attitude of users (Ducoffe, 1996). When sentimental responses are evoked, consumers can decide what they will buy based on the attitude of the advertisement alone, without having to fully process all the brand information (Zhu & Kanjanamekanant, 2021). Customer acceptance is especially important for both social media advertisers and social media providers (Taylor et al., 2011). The affect transfer hypothesis (ATH) proposed by Lutz (1985) suggests a direct one-way causal flow from consumers' attitudes towards advertisements to their attitudes towards brands. Moreover, this attitude affects consumers' purchase decision-making process as well, and different perceptions towards the ad lead to different levels of purchase intention (Zhu & Kanjanamekanant, 2021). Mitchell and Olson (1981), some of the first authors to emphasize the importance of the concept of consumer attitudes towards advertising in marketing literature, proposed that consumers' attitudes toward advertising influence their attitudes toward brands by changing the way consumers perceive the advertised brand. Attitude towards advertising also is an important measure of advertising effectiveness (Wang & Sun, 2010). It is announced that how a person evaluates an advertising message depends on their attitude towards the advertisement (Arora & Agarwal, 2019). Attitudes towards advertising, thus, have always been an interest of marketers because attitude is seen as a predictor of consumer behavior (Gaber et al., 2019).
To use online advertising effectively, marketers need to understand how users perceive the web as an advertising source because media perception influences attitudes towards advertising (Ducoffe, 1996). Along with the rapid development of powerful advertising media on the internet, the research on attitudes towards advertising in the online space is also expanding (Wang & Sun, 2010). The knowledge of attitudes towards advertising helps marketers to work selectively with attitudes that shape or influence attitudes towards a particular advertising message, which in turn affect a brand's attitude towards the advertising (Arora & Agarwal, 2019).

Personalization
On social media, advertisers are more capable of tailoring and customizing the types of messages, the content posted based on customer interests (Alalwan, 2018). These days, network advertisers tend to use consumer-oriented advertising in order to optimize the effectiveness of the ads and enhance customers' access to useful information (Lee et al., 2017).
Personalization refers to the delivery of advertising messages based on user demographics, interests, and context at the time of delivery to make advertising less irritating to customers. It also can be seen as a company's ability to identify and treat customers through messages, advertising banners, special offers, or other personal transactions (Gaber et al., 2019). In the personalized advertising context, Zhu and Chang (2016) defined relevance (the degree of personalization) as the degree to which consumers perceive a personalized advertisement to be selfrelated or in some way instrumental in achieving their personal goals and values.
Personalisation, which has also been recognised as a primary antecedent factor for advertisement value (Kurtz et al., 2021), is essential for understanding individual circumstances (Lee et al., 2017). Pavlou and Stewart (2000) suggested that personalized advertising has the potential to be more effective for marketers. The personalization of an advertisement is claimed to be a top prerequisite for customers to accept the messages and receive the information in the advertisement (Xu, 2006). Jung (2017) found that perceived ad relevance has an influence on advertising effectiveness such as increasing attention and decreasing ad avoidance. Gaber et al. (2019) declared that customers prefer to receive content that is more relevant to them and their interests. Thus, we propose:

H1: Personalization positively influences customers' attitudes towards advertisements on Facebook
Ahn and Bailenson (2011) have experimentally confirmed that the favorable brand attitude and purchase intention triggered by self-endorsing is mediated by the self-referencing effect. Research by Tam and Ho (2006) showed that, when exposed to messages with relevant content, customers tend to spend less effort searching for more information as well as spend less time making decisions. Personalized advertising has the potential to ensure that consumers receive only relevant information, which also means advertisers have the ability to send messages that are most likely to generate purchases or other desired responses (Pavlou & Stewart, 2000). Xia and Bechwati (2008) declared that when receiving useful promotional information, users tend to give it more weight, which in turn influences purchase intention.
Personalization was observed to have a positive impact on purchase intention in previous studies (Alalwan, 2018;Xia & Bechwati, 2008). Thus, our hypothesis states: H2: Personalization positively influences purchase intention of products presented in Facebook advertising.

Entertainment
In the context of advertising, it is illustrated that the value of entertainment lies in the advertisement capacity to fulfill audience needs for escaping and distracting from real-life difficulty, aesthetic enjoyment, or emotional (Ducoffe, 1996). Thanks to the change of the main advertising channel from newspaper to online advertising, entertainment is increasingly influential (Lee et al., 2017).
Advertisers seek benefits from entertainment to increase advertising effectiveness, and they particularly believe that entertaining ads will increase positive attitudes towards the brand (Logan et al., 2012). Entertainment is significantly and positively related to attitude toward a commercial website (Gao & Koufaris, 2006). Entertainment is one of the two content-driven features of advertising that have a critical impact on consumers' attitudes toward advertising (Taylor et al., 2011). The entertainment of advertising can also enhance the experience of advertising for consumers (Ducoffe, 1996). To have a positive customer attitude, hedonic perspectives have to be carefully addressed in social media advertising activities (Alalwan et al., 2017). When the ads on social media networks are creative and attractive, customers seem more attracted to them (Alalwan, 2018). Moreover, the entertainment of advertising messages becomes very useful in attracting the attention of customers (Gaber et al., 2019).
Empirical pieces of evidence have shown the significant effects of entertainment on attitude towards Web advertising (Ducoffe, 1996;Mahmoud, 2014), Sponsored links advertising (F.H. Lin & Hung, 2009), social media advertising in general (M. Shareef et al., 2017;Taylor et al., 2011), and Instagram advertising (Gaber et al., 2019). Hence, the next hypothesis is as follows: Venkatesh et al. (2012) declared that hedonic motivation (entertainment) is an important determinant of customers' behavioral intention. Entertainment was also claimed to play an important role in predicting customer reaction (Alalwan, 2018). In the research of Barata and Coelho (2021), hedonic motivation (entertainment) has been empirically proven as one of the variables which explain behavioral intention to buy a premium account of music streaming services. This factor was also a key predictor of purchase intention in the research of Alalwan (2018). Therefore, we propose: H4: Entertainment positively influences purchase intention of products presented in Facebook advertising.

Interactivity
Online environments have been shown to be more interactive and have a greater ability, in comparison with traditional media, to stimulate one-on-one interaction (Wright & Campbell, 2008). Interactivity has become an important feature in social media platforms and has received significant interest from many scholars regarding the related field (Alalwan, 2018). In the context of advertising and media communications, interactivity's purpose is to supply users the extent to exchange information in real-time on an on-demand basis where the timing, content, and sequence of the communication are under the control of the end-user (Sreejesh et al., 2020).
Interactivity is the ability for customers to gather relevant information and make purchasing decisions through social interaction and mechanical interaction (Jiang et al., 2010). According to Yim et al. (2017), technological outcome and user perception are the existing complete viewpoint that provides a comprehensive definition of interactivity. While the technological outcome emphasizes the more easily interact and be involved in the content, the user perception focuses on people characteristics that induce a sense of interactivity. Sreejesh et al. (2020) describe two different perspectives of interactivity, the user-to-user interactivity, which focused on the interpersonal communication perspective, and the user-to-system perspective, which focused on interactivity as the media characteristics. Huang (2012) defined interactive communication as a communication that provides users with active control, the ability to select information and direct interaction, and reciprocal communication, the ability to communicate between two or more entities.
There are many benefits of interactivity allowing users to have synchronous and reciprocal interactions, engagement, sense of control, and choices of media. Social media, with this compelling media interactivity, allows consumers to fully participate in the communication strategy, directing them to commit their full cognitive resources in information processing (Sreejesh et al., 2020).
When online advertising is developed with a higher interactivity level to match the inherent interactivity of the online environment, the advertisement becomes more suitable for internet users' experience. As a result, the level of interactivity is expected to have an impact on the user's ability to interact in the online environment and, accordingly, the user's attitude towards different aspects of the online environment (Wright & Campbell, 2008). Besides, the roles of interactivity such as critical influence on the customer's intention to use Mobile commerce; indirect impact on users' engagement over the social commerce website; or shaping customers' online buying behavior have been supported in a lot of studies (Alalwan, 2018). In addition, high levels of interactivity on Facebook are positively related to favorable attitudes towards the brand (Duffet, 2015).
According to the aforementioned discussion, it could be argued that the level of interactivity existing in Facebook advertising may have an impact on attitude towards this social media network ads. Thus, it is hypothesized that:

H5: Interactivity positively influences customers' attitudes towards advertisements on Facebook
Social media network, as a Web 2.0 platform, has a high level of interactivity, which allows users to have more capacity to interact and make their own contribution. This ability, in turn, increases the intrinsic and psychological benefits related to using and following social media advertising (Alalwan, 2018). Besides, consumers are likely to access the ad contents that have the information they need as well as an enjoyment itself thanks to the high-level interactivity of social media (Yang et al., 2013). Interacting with virtual products through image interactivity technology makes customers tend to be enjoyable and have a positive attitude towards the technology (Li et al., 2001). Enjoyment is shown as one of the interactivity results of users towards the website (Cyr et al., 2009). Jiang et al. (2010) revealed that interactivity features such as colors, sound, animation can enhance the entertaining value of online shopping. Therefore, we hypothesize: H6: Interactivity positively influences the entertainment on Facebook 2.2.7. Irritation Greyser (1973) presented in their research that the cutter of the advertisement or even the content of the advertisement itself can be irritating to customers. Irritation is considered the main challenge of advertising (Tsang et al., 2004). Advertising in high customer-targeted-possibility environments like social media can be more irritating than other channels (Taylor et al., 2011). On social media, perceiving that the ad is interfering with their goal-directing tasks will make customers find the ad irritating (Gaber et al., 2019).
Irritation, in terms of advertising, can be described as the feelings of customers when advertising tactics and techniques make them annoyed, offended, and overly manipulated (Ducoffe, 1996). Irritation caused by advertisements tactics may lead to the devaluation of advertising effectiveness (Ducoffe, 1996). The negative attitudes towards advertising are supposed to be related to the perception that advertising is a source of irritation (Mahmoud, 2014). According to M. Shareef et al. (2017), if anything about the ad creates irritation, consumers will likely feel annoyed and will eventually not be persuaded by the ad. When customers perceive social media advertising as being irritating, it can generate negative attitudes (Taylor et al., 2011).

Informativeness
Researches that consider information as a motivation to use the Internet have received much attention from many scholars (Mahmoud, 2014). Logan et al. (2012) stated that one of the main motivations of users when using social networks is to exchange information, and therefore, it is not surprising that the information-oriented capabilities of social networks make advertising information in this field receptive to users.
Informativeness can be seen as the process in which consumers realize the content in an advertisement and feel that it is informative regarding the product/service being advertised (Lee et al., 2017). Alalwan (2018) stated informativeness as the extent to which a company can provide sufficient information for the customers so that they could make better purchasing decisions.
M. Shareef et al. (2017) claimed that informativeness is crucial in persuading customers, moreover, the information provided by the ads is a predictor for customers to derive greater value from advertising. The advertising's capability to present a more precise description of products may drive consumer perceptions of its value (Taylor et al., 2011).
Informativeness is the other content-driven feature of advertising that has a critical impact on consumers' attitudes toward advertising that were mentioned in the research of Taylor et al. (2011). Informativeness, in the context of U&G theory, can be considered as the need of the consumers, which when is fulfilled may turn customers' perception into purchase intention (Rajesh et al., 2019). The perceived precision, timeliness, and usefulness of the information presented in the ads will have an impact on customers' attitudes towards advertising (Gaber et al., 2019).
Empirical shreds of evidence have shown the significant effects of Informativeness on attitude towards online advertising (Wang & Sun, 2010), Web advertising (Mahmoud, 2014), social media advertising in general (Taylor et al., 2011), and Instagram advertising (Gaber et al., 2019). Therefore, we hypothesize:

Privacy concerns
The fact that advertisers use data about previous browsing habits or even user-shared content to build targeted ads has raised concerns about user privacy (Taylor et al., 2011). When searching for information through online sources or when exposing private information to advertisers, consumers tend to express privacy concerns (C. Lin & Kim, 2016) Originally, information privacy was identified as people's ability to control the conditions under which their personal information was collected and used (Fortes & Rita, 2016). Privacy concerns, in terms of social media advertising, refer to users' desire to control how companies collect and use information about them through their online activities (Taylor et al., 2011).
Privacy concerns are "a precursor to prevention-related defective, defensive (control over receiving ads), and disruptive behaviors" (Wiese et al., 2020). Concerned that other people can infer their personality through ad likes, users tend not to click the "like" button (Lee & Hong, 2016). When users associate privacy concerns with viewing social media ads, they may have a negative bias towards accepting these ads (Taylor et al., 2011). Ence, the next hypothesis is as follows:

Credibility
Credibility has been an important research issue for Internet media and E-commerce studies (F.H. Lin & Hung, 2009). In the case of online purchasing, when the shops, the owners, and the products' quality are mostly unknown, the degree of required credibility becomes much greater (Heijden et al., 2003). Comments displayed on social media posts make advertising on this medium considered a reliable source of information (Arora & Agarwal, 2019).
Advertising credibility can be described as the degree of truthfulness and trustworthiness that consumers perceive about the claims made in the advertisement (F.H. Lin & Hung, 2009). Credibility refers to a customer's belief in the authenticity, integrity, and reliability of an advertisement (Rajesh et al., 2019).
As one of the main sources of advertising value in the online environment, advertising credibility has been shown to have a positive effect on consumer attitudes (Gaber et al., 2019). The credibility of the advertisements can positively influence consumers' attitudes and behavior (Arora & Agarwal, 2019). In the context of social media literature, credibility has been shown to positively affect consumers' perceived value of social media advertising, which in turn positively affects purchase intention (Gaber et al., 2019).

H10: Credibility positively influences customers' attitudes towards advertisements on Facebook
Finding that in the marketing industry, consumers have a moderate degree of privacy concern and a low level of trust in information, Milne and Boza (1999) have examined the relationship between the two factors. The results showed a negative relationship between credibility level and privacy concerns. Liu et al. (2005) declared that privacy has a strong impact on consumers' trust in an electronic supplier. Credibility is reduced when users perceive heightened awareness of information collection on the web (Olivero & Lunt, 2004). There are studies which show that privacy concern has a negative impact on credibility (Fortes & Rita, 2016;Liu et al., 2005). Therefore, we hypothesize:

H11: Privacy concerns negatively influence the credibility of Facebook advertising
In the online environment, customers cannot touch or look directly at the actual product, nor can they interact face to face with the seller, therefore, customers often feel "a lot of uncertainty and a huge risk in their online purchasing decisions" (Fortes & Rita, 2016). Gefen and Heart (2006) empirically proved that integrity (one of the three components of trust belief) affects consumers' purchase intention, and the ability (another component of trust belief) affects intentions to inquire about the product without actual purchase. When there is trust, the user's sense of social complexity and vulnerability tends to decrease, thereby, subjectively eliminating potential risks that may occur when making a purchase in e-commerce. As such, trust (credibility) can help buyers reduce their risk perceptions when dealing with online vendors, thereby encouraging them to engage in "trust-related behaviors", such as sharing information or making purchases (Lu et al., 2016). Thus, it is hypothesized that:

H12: Credibility positively influences purchase intention of products presented in Facebook advertising.
Attitude is described in the theory of planned behavior as the main driver of behavioral intentions (Ajzen, 1991). In the process of customers' purchase decision-making, attitude also plays a crucial role (Chen et al., 2017). Once emotions are aroused, customers tend to make buying decisions based on their attitudes toward advertising only, without having to deal with all the information about the brand (Zhu & Kanjanamekanant, 2021). Attitudes towards advertising have been shown to have an influence on behavioral factors such as behavior towards the brand, behavior towards the message (Wang & Sun, 2010;Wiese et al., 2020). The empirical results of previous studies also indicated that attitude towards ads has a positive impact on purchase intention (Kurtz et al., 2021;Lee et al., 2017;Zhu & Kanjanamekanant, 2021). Thus, our hypothesis states:

Measurement
The theoretical constructs used in this study were measured using validated items from previous research (see, Table 2). The questionnaire designed for this study was originally developed in English. As this study was conducted in Vietnam and as Vietnamese is the main language there, the current questionnaire was translated into Vietnamese by two independent marketing researchers, who spoke English and Vietnamese fluently and were familiar with the constructs. Next, the translators derived a combined translation of their independent translations of the measures, resolving translation discrepancies by discussing them. Eight constructs were used in the study. A five-point Likert scale anchor from strongly agree to strongly disagree was used to measure the main questionnaire items. The questionnaire was presented to academics and practitioners in the field of marketing to evaluate their perceptions of the study's topic and assess its measurability, as well as its context. The questionnaire was also pretested by collecting responses from Facebook users who had been exposed to Facebook advertisements to check its wording, sequencing, and completeness. Based on the feedback of respondents, the sequencing of the questionnaire was modified, ambiguous questions were deleted, and some wording was changed. Thus, the feedback improved the clarity, relevance, and consistency of the questionnaire.

Sampling and data collection
The hypotheses were tested using a convenience sample of Facebook users, who had been exposed to Facebook advertisements. According to Vietnam Briefing, Ho Chi Minh City is home to more than 9 million people, making it the largest city in Vietnam, with a strong economy supported by the most dynamic human resources in the country, the city holds the leading role in driving the economic and social development of the whole country (An, 2020). Therefore, we assumed that the research sample collected in Ho Chi Minh City is capable of representative of Vietnam and decided to conduct research data collection in Ho Chi Minh City. The questionnaires were distributed to the respondents by interviewers at milk tea and coffee shops in Ho Chi Minh City with instructions on how to complete them. Data was collected offline instead of online because we wanted to make sure respondents understood the question well and were given proper instructions before answering. The instructions emphasized that the study focused only on personal opinions to minimize possible response bias. There were no correct or incorrect answers. Respondents were informed that their participation would entitle them to a small gift. Cover letters were attached to the questionnaires to explain the aim and purpose of the research. In addition, the respondents were guaranteed the confidentiality and anonymity of their responses. To avoid potential bias stemming from a "sameness" in the consumers, trained interviewers were instructed not to interview more than five consumers from the same places. Respondents were asked to complete the self-administered questionnaire onsite within approximately 15 minutes. A total of 580 questionnaires were distributed between September 2021 and March 2022. After

Results testing scale
The measurement model was assessed via the reliability and validity of the studied constructs; the details are presented in Tables 2 and 3. The recommended thresholds for key reliability and validity indexes, such as Cronbach's alpha, composite reliability (CR), average variance explained (AVE) values, and factor loadings were adopted following Hair et al. (2010).
Construct reliability was measured using composite reliability. The value ranged from 0.778 to 0.887, which was higher than the recommended criteria of 0.6 (Hair et al., 2010). We measured the internal consistency of the items of each construct using Cronbach's α; the value ranged from 0.778 to 0.887, which was higher than 0.6, which met the criterion of 0.5 (Hair et al., 2010). Convergent validity was measured using factor loading, and average variance was extracted. The standardized factor loading of all items ranged from 0.669 to 0.862, which is higher than the recommended criteria of 0.5 (Hair et al., 2010). Further, to evaluate discriminant validity, the value of the average variance extracted ranged from 0.505 to 0.725, which was higher than the criterion of 0.5 (Hair et al., 2010). Moreover, to confirm discriminant validity, the square root of a construct's AVE should be higher than its bivariate correlation with other constructs in the model (Hair et al., 2010).

Structural model results
Owing to the complexity of the model and the need to test the relationships between the constructs simultaneously, we used structural equation modeling by applying the maximum likelihood method. Figure 1 shows the results of the testing model with Chi 2 = 1127.807, df = 482, Cmin/df = 2.340, Tucker-Lewis fit index (TLI) = 0.910 (> 0.9), comparative fit index (CFI) = 0.918 (> 0.9), and root mean square error of approximation (RMSEA) = 0.053 (< 0.07; Hair et al., 2010). Therefore, the hypothesized model offered an acceptable fit to the data.
The results also support H 6 , indicating that interactivity has a significant effect on entertainment (β = 0.102, p = 0.049). Finally, these findings support the notion that privacy concerns have direct and negative effects on credibility (β = −0.564, p = 0.000). Therefore, H 11 is supported.

Discussion
This study was conducted to build a user model for Facebook ads in Vietnam which is an emerging market. It aimed to identify and examine the main Facebook advertising features that could predict the customer's purchase intention for the products that are promoted using social media advertising. Overall, the study confirmed most of the proposed hypotheses and the overall validation of the research model. The findings show that customers' perceptions of informativeness are important factors that motivate them to attribute positive attitudes towards Facebook advertisements. By contrast, privacy concerns and irritation lead consumers to attribute negative attitudes toward advertising. Attitudes towards Facebook advertisements, then, in turn, have a significant positive effect on purchase intention. Moreover, factors such as entertainment, personalization, and credibility of Facebook ads are also found to have a significant positive effect on purchase intention. These findings are in line with previous research that examined the influence of advertising facilitators to Purchase Intention through Attitudes toward advertisements (Kurtz et al., 2021;Lu & Chen, 2021;Martins et al., 2019). Our findings also indicate that interactivity has a significantly positive effect on entertainment, and, privacy concerns were also shown to have a significantly negative impact on credibility.
Nevertheless, the influence of interactivity and personalization on attitudes towards advertising is not proven by the experimental results of this study. This means that a customer's perception of Note: *Significant at 10% level; **Significant at 5% level; ***Significant at 1% level; n.s: n non significant an ad's interactivity or personalization does not affect their attitude towards the ad. This result is inconsistent with the experimental results of the study by Wright and Campbell (2008), and Kurtz et al. (2021) but consistent with the results given by Gaber et al. (2019) and Arora and Agarwal (2019). The effects of entertainment and credibility on attitude towards advertising were also not empirically proven in this study. This means that entertainment and credibility are not crucial aspects in determining consumers' attitudes towards Facebook advertisements. These results are inconsistent with most studies on attitudes toward advertisement (eg. Aslam et al., 2021;Gaber et al., 2019;Taylor et al., 2011). These inconsistencies may stem from the research context in Vietnam, a Southeast Asia emerging country with a history and culture that is different from previously studied countries. In a different approach, these inconsistencies can also stem from the characteristics of the social networking platform Facebook itself. Further studies are needed to be done to identify the reason for these inconsistencies.

Theoretical contributions
The most important theoretical contribution is to the current understanding of the impact of Facebook advertising on Purchase Intention. Along with the development of technology, different features of advertising are also discovered, which can be mentioned as entertainment, informativeness, irritation, and credibility (Ducoffe, 1996;F.H. Lin & Hung, 2009;Mahmoud, 2014;Wang & Sun, 2010) appearing in online advertisements on the web 1.0. Later, the appearance of social media platforms allows for interactive and personalized advertising (Alalwan, 2018;Gaber et al., 2019;Kurtz et al., 2021;Lee et al., 2017), however, the increased display of tailored sponsored ads has made consumers concerned about privacy risks (Fortes & Rita, 2016;C. Lin & Kim, 2016;Taylor et al., 2011;Wiese et al., 2020). The above characteristics are considered by us to be the key influencers of customer attitudes towards Facebook advertisements. By capturing key influencers of customer attitudes towards Facebook advertisements in the current study model, this study was able to build an empirical model for Facebook ads in emerging markets, in this case, Vietnam. Aspects of Facebook advertising including credibility, entertainment, informativeness, interactivity, personalization, privacy concerns, irritation, and attitude towards ads have been once again confirmed as important factors in research models on the influence of advertising on consumers' purchase intention in emerging markets. In addition, the mediating role of Attitudes towards advertising in the relationship between advertising characteristics and the intention to purchase the advertised product is also established and examined through this study.
The study also contributes to the current understanding of the aspects of Facebook advertising by examining the relationships between key constructs in the model. Empirical results have proven the cause-and-effect relationships between interactivity and entertainment, as well as between privacy concerns and credibility.

Implications
As shown in Figure 1, attitude toward Facebook advertising has a significant impact on Purchase intention. This means that when customers have a positive attitude toward ads, specifically Facebook advertising, they have stronger buying intentions, in other words, they are more likely to buy the advertised product. Therefore, marketers in this emerging market need to pay attention to generating customers' positive attitudes toward advertisements. Moreover, according to the research results, informativeness has a positive effect on attitudes toward advertisements, at the same time, privacy concerns and irritation in descending order have a negative effect on attitudes toward advertisements. This means, that when an advertisement becomes more informativeness, less irritating, or does not raise audiences' concerns about privacy, viewers tend to have a more positive attitude towards that advertisement, and hence, are more likely to buy the advertised product. Therefore, practitioners and marketers should pay attention to increasing informativeness, reducing privacy concerns, and irritation of advertising. The impact of these factors should be combined with features suggested to concentrate in emerging markets such as adaptations to dynamic environments, knowledge transfers, relationships within a market, and the importance of political relationships (Paul, 2020). In that way, customers are more likely will show a positive attitude towards advertising and thereby have stronger purchase intentions.
Credibility was the second strongest factor that have a direct effect on consumers' purchase intention. These results are consistent with those of other studies that have tested the role of credibility (eg. Dachyar & Banjarnahor, 2017;Lu & Chen, 2021). This implies that when customers feel that an ad is trustworthy, the intention to buy the advertised item is stronger when they find the advertisement unreliable. Building social relationships with trust is also concluded as a critical strategy for international marketing success in emerging markets (Paul, 2020). Subsequently, advertisers should be aware of the credibility of the messages they provide and carefully consider the content of the information in the ads to increase customers' trust in marketers. They should provide customers with precise information that the customers desire, not information that they may interpret as spam. This finding also has implications for Facebook users. If they want to increase brands' expenditure on advertising on their platforms, they must ensure and demonstrate their credibility. In addition, privacy concerns were found to have a significantly negative impact on credibility. This means that when users perceive that their privacy has been violated, they are more likely to perceive the ads they see as untrustworthy, and subsequently, they have less possibility to buy the advertised item. This finding has more implications for Facebook executives than marketers. First, this social networking platform needs to make commitments, assuring users that their privacy will not be violated when they use it, and then ensuring strict compliance with what has been said. These are claimed to be favorable conditions for the development of the platform in general, and the development of Facebook advertising features in particular.
Entertainment has been empirically supported as a key factor in leading consumers to have purchase intention. In the relevant literature, the role of entertainment has been supported, such as studies by Venkatesh et al. (2012), Barata and Coelho (2021), and Lee et al. (2017). This means that if an individual feels that an advertisement is entertaining, they tend to form a stronger intention to buy that advertised item. Therefore, advertisers should make their ads more entertaining to increase the likelihood that Facebook users will react favorably. Business organizations should also take advantage of the entertainment aspect of advertising on social networking sites and make advertising more attractive. In addition, interactivity showed a strong positive effect on entertainment. This implies that if customers perceive a high level of engagement related to a Facebook ad, most of them will find such an ad more interesting, and accordingly, this synergistic effect will motivate their purchase intention. Subsequently, marketers should motivate their customers to engage more with advertisements posted on Facebook by providing feedback, comments, and information (Jiang et al., 2010). Companies can also ask marketing teams to monitor and respond to customer comments, questions, and feedback on their page.
The findings also show that personalization is an important factor in motivating consumers' purchase intention. Different studies have supported the role of personalization, such as Lee et al. (2017), Kurtz et al. (2021), and Alalwan (2018). This implies that if customers feel that an ad on Facebook is relevant to them, they tend to show a greater possibility of the item that is being advertised. Understanding and operating according to the target segment's needs is a key to increasing the performance of the multinational enterprises in any country, especially those within the emerging economies. Relationships within a market-from both an institutional (Business to Business) and consumer (Business to Consumer) perspective-were also suggested as an important feature that needs to be concentrated in an emerging market (Paul, 2020). Accordingly, marketers should find ways to deliver relevant advertising messages to the appropriate audience, making sure that their target audiences feel related. Owing to the company's "self-service" advertising capability, which allows smaller advertisers to create advertising inexpensively and target users based on demographic and psychographic variables, marketers can now easily create personalized ads. If properly used, such advertisements can be effective (Taylor et al., 2011). However, these tools must be used carefully to avoid privacy concerns because they may have a significant negative influence on credibility, which have a direct effect on consumers' purchase intention.
To sum things up, in emerging markets, all the aspects of advertising covered in the study have been experimentally proven to have either direct or indirect impacts on purchase intention. Therefore, when implementing advertising campaigns on Facebook, marketers in emerging markets need to pay attention to these factors so that their campaigns can increase customers' purchase intention, which then increases the chance of purchase (Martins et al., 2019). The strong direct impact of credibility on purchase intention in the research results is also something that practitioners need to pay attention to. Accordingly, when establishing advertising plans in emerging markets, marketers particularly need to pay attention to the credibility of their advertising messages. The research results also have implications for Facebook developers in emerging markets. First of all, these developers should pay more attention to the algorithms in the advertising tools they offer users. These tools need to have the ability to support users to create ads that are entertaining, informative, highly reliable, or create little irritation for customers. Besides, developers also need to make good use of their database to create advertising recommendations that are suitable for users, so that these ads can increase purchase intention. However, it is important to keep in mind that the gathering of customer information should be done to a reasonable extent and avoid creating privacy concerns for them, otherwise, both credibility and attitude towards advertising (two variables with the greatest impact on purchase intention considered in) the study) will be negatively affected, consequently backfiring and reducing the purchase intention of consumers.

Limitations
Despite the contributions of the current study, it has several limitations. First, the sample was constrained to Ho Chi Minh City, which may have limited the generalisability of the findings. Second, this study considered only a limited number of predictors of attitudes towards Facebook advertising and Purchase intention. Future studies can consider more constructs to build a more comprehensive advertising model. The next limitation is that our research was conducted for products and services advertised on Facebook in general and did not focus on a specific category of products or services. Future studies can be oriented to building a research model of attitudes towards advertising for a specific industry or field, not just attitudes towards Facebook ads in general. Finally, as mentioned before, the empirical result of this study has some inconsistencies with previous studies in this field, and therefore, further studies are needed to be done to identify the reason for these inconsistencies.