Online Purchase Intention Under the Integration of Theory of Planned Behavior and Technology Acceptance Model

This study focuses on predicting the online purchase intention of consumers purchasing from e-commerce platforms using an integrated model of the theory of planned behavior (TPB) and extended technology acceptance model (TAM). The results showed that TAM factors, perceived value, and perceived enjoyment had significant effects on attitude toward online purchase, while the TPB inputs, perceived enjoyment, and e-commerce exchange image played strong roles in influencing online purchase intention. The results also showed that perceived enjoyment helped to raise the predictive power of attitude dramatically, while the e-commerce exchange image contributed a high portion to explain online purchase intention variation. The results indicated that the e-commerce trading platforms need to focus on improving their images and on increasing customers’ experience to improve their business performance. Plain Language Summary Prediction of online purchase intention on e-commerce platforms This study focuses on predicting the online purchase intention of consumers purchasing from e-commerce platforms using an integrated model of the theory of planned behavior (TPB) and technology acceptance model (TAM). The study proposed a model with 13 hypotheses from the literature review. The hypotheses were tested using data collected from 326 e-commerce consumers in Vietnam. The PLS-SEM approach was used for analysis. The results showed that TAM factors, perceived value and perceived enjoyment had significant effects on attitude towards online purchase, while attitude, perceived enjoyment and e-commerce exchange image played strong roles influencing online purchase intention. The results also showed that the perceived enjoyment helped to raise the predictive power of attitude dramatically, while e-commerce exchange image contributed a high portion to explain online purchase intention variation. Other specific contextual factors may be incorporated into the integrated model to better explain online purchase intention. The results indicated that the e-commerce trading platforms need to focus on improving their image and on increasing customers’ experience in order to improve their business performance. This is among the first studies that investigates the integration of the TPB and TAM for e-commerce behavioral intention research incorporating e-commerce exchange image as an important factor that affects online purchase intention. The new integrated model has verified its value to considerably enhance the capability to predict the variations of attitude towards online purchase and online purchase intention.


Introduction
Business-to-customers (B2C) e-commerce has had a breakthrough in recent years due to the expansion of information technology infrastructure under Industrial Revolution 4.0.The quick growth of internet and smartphone (and tablet) users pushed the development of B2C e-commerce and mobile commerce (Chiu et al., 2018).The e-commerce companies foster their business through e-commerce exchanges (or e-commerce trading floors) which are facilitated by websites and mobile phone applications (Pillai et al., 2022).Because of its flexibility, using mobile phones seems to be a preferable way of practicing B2C e-commerce.This leads to the fast development of mobile commerce or m-commerce.
This research investigates e-commerce B2C trading in an emerging economy (Le-Anh & Nguyen-To, 2020) which is Vietnam.According to C c_Th ng_m i_ i n_t _và_Kinh_t _s (2022), the estimated scale of the e-commerce B2C market of Vietnam in 2022 is about 16.4 billion USD raising from 13.7 billion USD in 2021 and e-commerce trading is becoming an important trading mode which accounts for about 7.8% total consumer product retail volume of the country.The data also reveals that the percentage trading volume through e-commerce exchanges in the Vietnam online B2C market is 96.1%.This means that B2C online trading using e-commerce exchanges is the dominating mode in the Vietnamese online B2C market.Therefore, this research mainly focuses on online B2C trading through e-commerce exchanges.
Purchasing through an e-commerce exchange has many advantages such as reduction of prices, saving time, and eliminating several physical difficulties involved in traditional shopping (Tsao et al., 2016).With an ecommerce exchange, customers can surf the platform, compare the characteristics and prices of products from different sellers (manufacturers, retailers, etc.), and select the suitable ones that they want.The variety of products selling on e-commerce exchanges is growing quickly with several suppliers participating in e-commerce exchanges.This increases the ability of customers to find the right products and attracts more customers to use e-commerce exchanges which further raises the growth of the ecommerce market in the future (Liao et al., 2021).
Along with the growth of the e-commerce market, the number of studies on e-commerce related issues has been increasing quickly, in which online purchase intention of customers on e-commerce platforms received much interest from researchers (Akram et al., 2021;Chiu et al., 2018;Pen˜a-Garcı´a et al., 2020).
Studies in the literature revealed that utilitarian and hedonic motivations (Akram et al., 2021) are important factors influencing online purchase intention, in which hedonic motivations have a higher impact than utilitarian motivations.Opposite to motivations, risks caused negative impacts on purchase intention on ecommerce platforms directly or through mediators (Aghekyan-Simonian et al., 2012;Liao et al., 2021;Pillai et al., 2022).To study online purchase intention, several theoretical models were frequently used including the theory of planned behavior (TPB) and its extensions (Leong et al., 2023), the technology acceptance model (TAM) (Chen & Teng, 2013), or the stimulus-organismresponse (SOR) model (Liu et al., 2016).Besides these three frequently used theories, other theories were used as well (Abumalloh, 2018;Ayalew & Zewdie, 2022).
In practice, authors integrated the TPB and TAM to better explain online purchase intention (German Ruiz-Herrera et al., 2023;Pen˜a-Garcı´a et al., 2020;Rehman et al., 2019;Singh, 2015;Song & Jo, 2023).Song and Jo (2023) and Rehman et al. (2019) combined the TPB and TAM in a way that input factors of the TPB and TAM directly influenced purchase intention in the model.However, the TAM inputs showed unstable impacts on purchase intention.Pen˜a-Garcı´a et al. (2020), on the other hand, showed the constant impact of the perceived usefulness on attitude in the TPB model.According to the literature, the impact of enjoyment which is an important factor of the extended TAM model (Davis et al., 1992), has not been examined in the integrated TAM-TPB framework which is an interest of this research.
In the online business environment, the image of products (Aghekyan-Simonian et al., 2012), retail stores (Aghekyan-Simonian et al., 2012;Rizal et al., 2022), and companies (Kuo & Tang, 2013;Srivastava & Sharma, 2013) certainly may have impacts on the customers' purchase intention.The studies in the literature indicated that the online store image and the corporate image showed different (significant and insignificant) impacts across cases.Moreover, the impact of e-commerce platform image (for retail stores trading) on online purchase intention has not been examined so far.
Therefore, this research contributes to online purchase behavior by evaluating the impacts of perceived enjoyment and the e-commerce platform image in an integrated TPB and TAM framework.The perceived value (Fishbein & Ajzen, 2009), which can be seen as an important factor of the TPB model that may cause an impact on the e-commerce platform image was included in the proposed framework as well.The main aims are to find out the contributions of the extended TPB-TAM with the incorporation of perceived enjoyment and ecommerce platform image in the explanation of the online purchase intention on e-commerce platforms.
The rest of the paper is organized as follows: section 2 presents the paper's conceptual framework, section 3 describes the research methodology of the paper, section 4 provides results and discussion, and the last section presents conclusions of the study.

Conceptual Framework
This research focused on customer purchase intention in e-commerce exchanges.E-commerce exchanges or ecommerce trading floors are the main e-commerce trading medium in Vietnam.An e-commerce exchange, in our research, can be viewed as an e-commerce website (and the corresponding app on Android or Apple iOS) and back-office facilities serving as a mediator for the trading of goods.
To study the customer purchase intention on ecommerce exchanges, the theory of planned behavior can be considered one of the most appropriate theoretical frameworks, in which attitude toward online purchase, subjective norms, and perceived behavioral control are precedence of online purchase intention.This research, therefore, used the TPB as the core of the model.The TPB model was extended by adding perceived value as a precedence of attitude (Aertsens et al., 2009).To better explain the online purchase intention, the extended TAM model was integrated with the TPB (Figure 1).The original TAM framework used the perceived ease of use and perceived usefulness to predict the use acceptance of computers (Davis & Davis, 1989).In this research, the TAM factors were added into the TPB, in which perceived ease of use and perceived usefulness caused direct influences on attitude toward online purchase, while perceived ease of use was an antecedent of perceived usefulness and perceived usefulness might influence online purchase intention.The perceived enjoyment in this case was an extended factor of TAM (Chang et al., 2015) and was integrated into the proposed framework.E-commerce exchange image was also added as another precedence factor of online purchase intention.

Online Purchase Intention
Ajzen (1991) defined intention as ''indicators of how hard people are willing to try, of how much of an effort they are planning to exert.''Behavioral intention is at the center of the TPB theory and is the main predictor of purchase behavior (Ajzen, 1991).Online purchase intention here means the intention of a customer to buy goods/ products from e-commerce trading floors.The central role of the purchase intention had been verified in different research areas including energy saving (Gao et al., 2017), recycling (Park & Ha, 2014), organic food (OF) consumption (Paul et al., 2016).

E-Commerce Exchange Image
The e-commerce exchange image has not been studied in the literature on online purchase behavior so far.The concept of e-commerce exchange image is close to the corporate image (Swoboda et al., 2016) and online store image (Aghekyan-Simonian et al., 2012).Kuo and Tang (2013) viewed corporate image as ''the overall cognitive and affective perception in relation to the corporate reputation, products and services, core concepts of management, and quality of communication'' and verified that the corporate image was significantly affected by the quality of the accessible environment.
Other authors reflected corporate image as a multidimensional concept which included reputation and credibility dimensions (Srivastava & Sharma, 2013).It was also assumed that the corporate image is established and developed in the consumers' minds through communication and experience.A well-established corporate image in the consumers' minds certainly influences a customer's intentions and behavior.Srivastava and Sharma (2013) believed that corporate image creates a halo effect on customers' satisfaction judgments.They evaluated that the corporate image was a valuable asset for companies and a favorable image can boost sales.Their research also revealed that a credible corporate image resulted in high repurchase intention and less switching behavior.A similar effect was observed in the research of (Hu et al., 2009) where the corporate image is an important predecessor of behavioral intention.In another case, the corporate image did not show a significant impact on the intention to choose high-speed rail services for elderly persons in Taiwan (Kuo & Tang, 2013).The possible reason was that the elderly persons did not care much about company image but satisfaction instead.
In the online purchase environment, the impact of online store (or e-store) image on purchase intention has been examined by several authors (Aghekyan-Simonian et al., 2012;Jiang et al., 2023;Rizal et al., 2022).Aghekyan-Simonian et al. (2012) showed that product brands had a significant effect on purchase intention from e-store, while e-store image had not.They argued that the e-store image did not have a strong role as a physical store image because of the private nature of online shopping, therefore it causes a negligible impact on purchase intention.However, e-store had an indirect impact on online purchase intention in their research.Jiang et al. (2023), in another research, defined e-store image as the result of consumers' perception of the functional and psychological attributes of stores and verified the impact of e-store image on consumers' intentions.
An e-commerce exchange is a trading floor for e-stores that is managed by a company, therefore an e-commerce exchange image shares some similarities with a corporate image and an e-store image and is likely to impact online purchase intention (Aghekyan-Simonian et al., 2012;Ma et al., 2022).Moreover, in practice, C c_Th ng_m i_ i n_t _và_Kinh_t _s (2022) showed that customers' decisions to buy products on an ecommerce trading floor (e.g., Shoppee) were affected strongly by the image of the trading floor.The above arguments lead to the following hypothesis: H1. E-commerce exchange image positively affects online purchase intention

Perceived Value
Perceived value has been acknowledged as an important construct in service marketing (Boksberger & Melsen, 2011).Authors defined perceived value from different angles as from utilitarian (Parasuraman & Grewal, 2000) or behavioral (Sheth et al., 1991) perspectives.Aertsens et al. (2009) stated that perceived value was connected to an individual's cognitive system and identified a set of 10 values that can incorporate all specific values.Aertsens et al. (2009) also showed that there existed a link between perceived value and attitude as indicated by the expectancy-value model of attitudes.Thi Nguyen and Dang (2022), in another research, verified the positive impact of perceived value on attitude under the behavioral reasoning theory framework while studying organic food purchase behaviors.Hence, we propose the following hypothesis:

H2a. Perceived value positively influences attitude toward online purchase
Studies in the literature pointed out that there was a relationship between perceived value and image (related to a company).Hu et al. (2009) suggested that a company would have a strong image in case customers believed that they were getting high value when they bought from it.They also verified the positive impact of perceived value on the corporate image in their study.These arguments lead to the following hypothesis: H2b. Perceived value positively influences e-commerce exchange image

TPB Input Factors
Attitude.Attitude is an important psychological concept that has been widely used by many authors (Ajzen, 1991;Fishbein & Ajzen, 2009;Gao et al., 2017;Paul et al., 2016).Originally, Ajzen (1991) defined attitude toward a behavior as ''the degree to which a person has a favorable or unfavorable evaluation or appraisal of the behavior in question.''Both TPB and TAM identified attitude as an important predictor of behavioral intention (Chang et al., 2015;Fishbein & Ajzen, 2009).Under volitional conditions, the TRA theory stated that attitude and subjective norms are the sole predictors of behavioral intention (Madden et al., 1992), while the TPB added perceived behavioral control as another predictor in the case of non-volitional control.Nonetheless, in both theories, attitude always plays a significant role in influencing behavioral intention.Pen˜a-Garcı´a et al. (2020), in a research on online shopping behavior, identified attitude along with other factors including perceived behavioral control having a significant influence on online purchase intention.Lin (2007) compared consumer intention to shop online using three competing theories and found that attitude played a significant role in influencing behavioral intention in all three experimental models.These arguments lead to the following hypothesis: H3. Attitude toward online purchase positively affects online purchase intention Subjective Norms.According to Ajzen (1991), subjective norms referred to ''the degree to which a person has a favorable or unfavorable evaluation or appraisal of the behavior in question'' and it was seen as an important predictor of behavioral intention.Subjective norm has shown its significant influence in several proenvironmental related research (Wang et al., 2016), however, its impact seemed inconsistent from case to case.
As subjective norms are an important factor impacting individual behavioral intention and may have a significant effect on online purchase intention, we propose the following hypothesis:

H4. Subjective norms positively affect online purchase intention
Perceived Behavioral Control.Perceived Behavioral Control (PBC) can be seen as the way that a person perceives a behavior as easy or difficult to perform (Ajzen, 1991;Gao et al., 2017).Ajzen (2002) decomposed PBC into two components which are self-efficacy and controllability, in which self-efficacy was likely to influence intention, while controllability affected actual behavior.PBC consistently authenticated its impacts on behavioral intention in different fields of research (Gao et al., 2017;Nguyen et al., 2019;Paul et al., 2016;Wang et al., 2016;Yadav & Pathak, 2017).Therefore, this research proposes the following hypothesis:

H5. Perceived behavioral control positively affects online purchase intention
Extended TAM Model TAM Factors.The adoption of online shopping can be considered as a type of new technology implementation in retail.Hence, it is important to understand the acceptance of this new way of retailing practice.The Technology Acceptance Model (TAM) has been used popularly to assess the level of acceptance of new technology implementation in practice (Davis & Davis, 1989;Kim & Forsythe, 2010;Kim et al., 2009).
In the original model, Davis and Davis (1989) defined perceived ease of use (PEU) as ''the degree to which a person believes that using a particular system would be free of effort'' and perceived usefulness (PU) as ''the degree to which a person believes that using a particular system would enhance his or her job performance.''They attested that the PEU and PU were the two main factors that decided the user acceptance of a new technology.Davis and Davis (1989) showed that PU caused a much stronger impact on user acceptance than PEU and they also suggested that PEU should be an antecedent of PU.
In the online business environment, the ease of use characteristic makes online purchases easy to practice, hence increasing the frequency of online purchases or raising the usefulness of purchases online.The EOU and PU of online purchases attract customers to buy online, therefore causing positive effects on customers' attitudes and purchase intentions.
In extended versions of TAM (Chang et al., 2015;Chen et al., 2018;Lee et al., 2006), perceived enjoyment was added to the original TAM and this factor played a role in influencing online perceived intention.Chang et al. (2015) also indicated that PEU had a direct impact on perceived enjoyment when studying the behavioral intentions of consumers to use m-commerce.These arguments lead to the following hypotheses: H6a. Perceived usefulness is positively associated with attitude toward online purchase H6b.Perceived usefulness is positively associated with online purchase intention H7a.Perceived ease of use is positively associated with attitude toward online purchase H7b.Perceived ease of use is positively associated with perceived usefulness H7c.Perceived ease of use is positively associated with perceived enjoyment Perceived Enjoyment.Perceived enjoyment can be considered as a joint factor of both the TAM and TPB models.In the context of TAM, perceived enjoyment refers to ''the extent to which the activity of interacting with the e-commerce website is perceived to be enjoyable in its own right aside from the utilitarian value of the site'' (Davis et al., 1992).In the context of the TPB, perceived enjoyment can be considered an affective state that represents an aspect of experience related to a specific behavior (Fishbein & Ajzen, 2009).Practically, perceived ease of use in the online purchase environment raises the favorable experiences of customers and then encourages persons to continue their respective actions and, therefore, are likely leading to a positive attitude toward a specific behavior and increase the possibility or intention to practice the corresponding behavior.Aertsens et al. (2011) suggested there existed a link between experience and attitude in their study.Hence, perceived enjoyment (a type of experience in this research) can have an impact on attitude.Therefore, we propose the following hypothesis: H8a. Perceived enjoyment positively influences attitude toward online purchase In their experiment with affective states (anticipated regret), Fishbein and Ajzen (2009) showed that by adding anticipated regret, the predicting power of intention was increased significantly.This suggested a positive correlation between affective states which reflect an aspect of intrinsic motivation with behavioral intention.In the context of e-commerce research, Qiu and Li (2008) suggested that perceived enjoyment was close to intrinsic motivation and can exert influences on the adoption intentions of e-commerce.These arguments lead to the following hypothesis: H8b. Perceived enjoyment positively influences online purchase intention

Constructs and Measurement
In this research, we proposed an integrated framework of the TPB and extended TAM.Constructs and indicators have been used and adapted from previous studies.We prefer to select indicators from recently related studies in the research area that are best suited to the research situation.Indicators (questions) were adapted from different studies as follows: attitude toward online shopping, subjective norms, PBC (Pen˜a-Garcı´a et al., 2020), perceived value (Parasuraman et al., 2005), perceived enjoyment (Chen et al., 2018), E-commerce exchange image (Swoboda et al., 2016), perceived usefulness, perceived ease of use (Lee et al., 2006), online purchase intention (Gan & Wang, 2017).In this research, several constructs' indicators were adapted from the previous study which were appropriate for the current research.Experts in the field were asked to give their opinions on the appropriateness and validity of the proposed constructs and indicators.Adjustments were applied when it is necessary.Each item in this research was measured on a 7-point Likert scale ranging from 1 (''strongly disagree'') to 7 (''strongly agree'').
The questionnaire, in this research, had been designed with two sections: (a) questions to gather general information on a respondent, and (b) questions to collect information on respondents' perception of factors in the proposed research model.The content of the questionnaire had been discussed with experts in the field and a pilot study of 60 respondents was used for questionnaire adjustment.The pilot study was used to guarantee that the proposed constructs and indicators were appropriate for this study.The online questionnaire (google form) was sent out to potential respondents.This survey collected data from consumers who had already purchased online at least once.

Sample and Data Collection
In our study, online Google forms were sent out to consumers who had already practiced online purchases in Vietnam using the volunteer (self-selection) sampling method (Saunders et al., 2019).As described by Saunders et al. (2019), the volunteer sampling method is a type of non-probability sampling technique that provides better samples than the popularly convenient sampling method.In total, we collected 326 usable filled questionnaires that can be used for data analysis.The survey was conducted during the first half the year of 2021.The sample size in this research has been verified by several statistical criteria: (a) the number of observations should be at least five times the number of indicators (the proposed model has 30 indicators hence we need at least 150 observations), (b) to guarantee the 80% statistical power and detect minimum R 2 of 0.1 with (max) six arrows pointing at online purchase intention construct and 95% significance level, the sample size should be at least 130 (Hair et al., 2017).In this case, we collected 326 observations that were appropriate for analysis.

Data Analysis
Table 1 shows basic information about online shopping consumers with detailed characteristics on gender, occupation, age, monthly income, number of annual online purchases, and the chosen e-commerce exchange.Table 1 indicates that the number of female respondents is about double in comparison with the number of male respondents since females usually use more time for online shopping than males.Similar to other studies in online purchasing research (Pen˜a-Garcı´a et al., 2020), consumer groups aged under 40 are the dominant groups of customers.This research focused on consumers who purchased from e-commerce trading floors such as Lazada, Shoppee, Tiki, rather than ones that directly bought from retailers.These exchanges serve as trading floors for sellers (retailers and producers) and buyers and provide logistic services for sellers.In practice, the sellers may carry out logistic activities themselves.
This research proposed an integrated model that combines the TPB and extended TAM to study factors influencing online purchasing behaviors.Several indicators in the research model were adapted from previous studies, therefore, we used exploratory factor analysis (EFA) to verify the appropriateness of indicators and constructs.Lately, SmartPLS has been used to analyze the structural model.
In our analysis, indicators, that had EFA loading higher than 0.5 and had PLS-SEM outer loading greater than 0.7 (reflective model was used in this case), had been kept for later statistical analysis (Table 2).We checked the reliability of all constructs and indicators in the model to ensure that all of them met statistical requirements according to Hair et al. (2017).As shown by Hair et al. (2017), the main requirements for data reliability are outer-loading ø 0.7; Cronbach Alpha and Composite Reliability ø 0.7; Average Variance Extracted (AVE) ø 0.5.Indicators that did not satisfy statistical requirements had been eliminated (Table 2).The data analysis was conducted using the PLS-SEM reflective model in SmartPLS 4 (Hair et al., 2017).
According to Hair et al. (2017), f 2 (effect size) values in our experiment indicated that all impacts in the model are significant (f 2 .0.2-threshold level of small impact).Even though the model fit has not been popularly used for PLS-SEM, the model SRMR value of 0.048 (\ 0.08) was good enough in this case (Hair et al., 2017).The Q 2 predicted values in Table 4 mean the PLS-SEM model offers a better predictive performance.
Figure 2 illustrates the bootstrapping results of the integrated model using SmartPLS4.The results showed that most impacts in the model were significant, except for the two connections PU Ò Online purchase intention, PBC Ò Online purchase intention.This meant that the hypotheses H1, H2a, H2b, H3, H4, H6a, H6b, H7a, H7b, H7c, H8a, and H8b were confirmed and two hypotheses H5, H6b were rejected.
As can be seen in Figure 2, the R 2 values of attitude toward online purchase and online purchase intention are pretty high (68.7% and 62% respectively).The R 2 value of 0.62 corresponding to online purchase intention means the input factors can explain a 62% variation in online purchase intention.
Table 5 shows the total indirect effects and Table 6 presents the total impacts of input factors in the model.The results illustrated that most factors caused significantly positive impacts and that only one factor caused directly negligible influences on purchase intention which was perceived usefulness.Yet, perceived usefulness caused an indirect significant influence on purchase intention through attitude toward online shopping (Table 5).Table 6 also indicated that e-commerce exchange image and perceived value caused the highest impacts on online purchase intention.

Discussion
This research introduced the integrated model which combined the TPB, the extended TAM, and two other factors (perceived value and e-commerce exchange image) to explain the variation of attitude toward online purchase and consumer online purchase intention.The survey data focused on the customers who have bought products from popular e-commerce trading floors in Vietnam such as Lazada, Shoppee, Tiki, etc.
Considering the TPB part of the integrated model in this research, subjective norms caused significant impacts and PBC caused an insignificant effect on online purchase intention.The significant impact of SN is agreeable   with the results from several authors (Song & Jo, 2023;Wang et al., 2016).This meant that individual online purchase intention was affected by the important surrounding persons.
The insignificant effect of PBC in this research was similar to the result of Aboelmaged ( 2021) but was different from the results of (Paul et al., 2016;Pen˜a-Garcı´a et al., 2020).This negligible impact can be explained by the lack of trust or the inconsistency of product quality provided by online product suppliers.The lack of trust certainly reduced the self-efficacy of customers which was an important component of PBC.Among TPB inputs, attitude toward online purchase indicated a strong influence (b = .236)on online purchase intention    as in other research (Aboelmaged, 2021;Gao et al., 2017;Paul et al., 2016;Pen˜a-Garcı´a et al., 2020;Wang et al., 2016).This factor seemed to be the most reliable construct of the TPB framework.
Taking the TAM input components of the integrated model into account, it was found that the effects of factors in the model were similar to the previous studies (Lee et al., 2006;Li et al., 2017), except for perceived usefulness which caused negligible impact on purchase intention.PU showed their consistent influences on attitude toward online purchase (German Ruiz-Herrera et al., 2023;Pen˜a-Garcı´a et al., 2020).PEU also indicated a significant impact on PU in our research that is similar to (Chen & Teng, 2013).In this research, PEU showed a higher influence on attitude than PU which is different from the study of (Pen˜a-Garcı´a et al., 2020).This can be explained by the fact that online shopping cannot replace traditional shopping currently and consumers in Vietnam may not have a high level of trust (C c_Th ng_m i_ i n_t _và_Kinh_t _s , 2022) in online purchase.This fact may also explain the directly negligible effect of PU on online purchase intention.
While considering the indirect effect of PU, the results indicated that PU caused an indirect impact on purchase intention through attitude as a mediator (b = .047,pvalue = .017).
The integrated model in our research also considered supplemental factors of the TPB and TAM.Perceived enjoyment, which is an extended factor of TAM, can be considered a supplement factor of the TPB as well (Fishbein & Ajzen, 2009).In the extended TAM, Davis et al. (1992) introduced perceived enjoyment as ''the extent to which the activity of using the computer is perceived to be enjoyable in its own right, apart from any performance consequences that may be anticipated'', which is intrinsic motivation, and they attested the positive impact of perceived enjoyment on behavioral intention.
From the TPB perspective, perceived enjoyment is considered as an affective state and may influence behavioral intention.The results of our research verified the significant impact of perceived enjoyment on attitude and online purchase intention.The perceived enjoyment caused the strongest effect on attitude (b = .417)and a moderate influence on online purchase intention (b = .194).These results are agreeable with the results of several previous studies (Chang et al., 2015;Davis et al., 1992;Li et al., 2017;Wu & Li, 2017).This implied that perceived enjoyment is an important factor in online behavioral intention research.
The results also indicated the significant impact of PEU on perceived enjoyment and subsequently on purchase intention.The influence of PEU on perceived enjoyment was identified in the research of Chang et al. (2015), in which PEU showed a strong impact on perceived enjoyment.In a report in 2022, C c_Th ng_m i_ i n_t _và_Kinh_t _s (2022)  also identified that PEU is among important factors (with 34% rating) that influenced consumer purchase decisions in the B2C e-commerce market in Vietnam in 2022.
In this research, we introduced perceived value as another supplement factor that had already been discussed in the BRT framework (Westaby, 2005).As discussed in the BRT theory, perceived value is an important factor that may impact attitude toward online purchase.The results of this research have already verified this impact.In our experiment, perceived value caused a direct impact on attitude (b = .216).
In an online business environment, the e-commerce exchange image represents the image of the whole ecommerce platform for various e-stores.This image has two dimensions of a corporate image that are reputation and credibility (Srivastava & Sharma, 2013).An ecommerce exchange's reputation is accumulated from the customer's assessment of the e-commerce platform's operations and the e-commerce exchange's credibility accounts for the customer's evaluation of different aspects such as trustworthiness, dependability, and concern for customers.
Practically, an e-commerce platform that can bring in more value to customers will create a higher credibility and increase its reputation, hence raising the e-commerce exchange image.This relationship is proven by the significant influence of perceived value on e-commerce exchange image (b = .614)in this research.This result is agreeable to the result of Hu et al. (2009), in which the perceived value had a strong effect on corporate image in their study.
In addition, an e-commerce platform with a high reputation and credibility will be likely to raise the customer's intention to buy products on this ecommerce platform.The results of this research support this argument by confirming the significant influence of e-commerce exchange image on online purchase intention (b = .446),which is agreeable with the results of (Hu et al., 2009) for corporate image.According to the white book on e-commerce in Vietnam (C c_Th ng_m i_ i n_t _và_Kinh_t _s , 2022), consu- mers identified a charisma of an e-commerce exchange as the most important factor (with 74% rating) affecting their online purchase decisions.This confirmed the strong role of e-commerce exchange image in practice.
The results of this research also showed that the introduction of perceived enjoyment and e-commerce exchange image improved the prediction power of the online purchase intention dramatically.The R 2 value of the online purchase intention increased from 48.3% to 62%.It means that the predictive power of the online purchase intention increased by 13.7%.Particularly, the e-commerce exchange image increased the R 2 value of online purchase intention by 13.1% from 48.9% to 62%.While the R 2 value of attitude increased from 25.8% to 68.7% by incorporating the extended TAM into the TPB framework.It means that the predictive power of attitude is raised by 42.9% by integrating the extended TAM into the TPB model.
Regarding the levels of impact on online purchase intention, the e-commerce exchange image showed the highest impact (b = .446),and the second strongest impact was from perceived value (b = .325).Despite showing a very consistent effect on behavioral intention across areas (Gao et al., 2017;Paul et al., 2016), the level of attitude influencing purchase intention (b = .236)was even lower than the impact of perceived enjoyment (b = .292)in our experiment.This suggested that both hedonic and utilitarian values played significant roles in influencing purchase intention which was similar to the results of Gan and Wang (2017).

Implications
Theoretical Implication.This research has several theoretical implications.Firstly, the integration of the TPB and the extended TAM theories incorporating exchange image and perceived value certainly raises the predicting power of the attitude and online purchasing intention.As shown before, the capability of explaining the variation in attitude and online purchase intention increased by 42.9% and 13.7% respectively.The TAM components and perceived value helped to explain the main variation of the attitude toward online purchase, while the extended TAM-TPB joint component (perceived enjoyment) and e-commerce exchange image played important roles in explaining the online purchase intention variation.This illustrated the importance of integration of the two behavioral theories (TAM and TPB) to better predict online purchasing intention.
Secondly, this research verified the strong role of perceived enjoyment in influencing attitude toward online purchase and online purchase intention through ecommerce platforms.This role of perceived enjoyment has been identified for the group with a majority under 30 years old that used m-commerce (Chang et al., 2015).The results of this research verified the significant impacts of perceived enjoyment on both attitude and online purchase intention for different groups of customers who had already practiced purchasing products through e-commerce platforms.
Thirdly, this study examined and proved the key role of e-commerce exchange image affecting online purchase intention through e-commerce platforms.E-commerce exchange image was derived from corporate image (Hu et al., 2009) and showed to have a significantly strong influence on purchase intention and even caused the strongest impact in our case.The result also indicated that the perceived value, which is an important input of the TPB model (Fishbein & Ajzen, 2009) had a strong influence on the e-commerce exchange image in this study.
Finally, we found that attitude played a central role in the integration model which was significantly impacted by various factors of the TAM and TPB framework.This proved that attitude had an undeniable role in human behavioral theories.Besides having a direct impact on purchase intention, this variable also played a mediating role for different other variables in the behavioral research.
Practical Implication.From a practical point of view, this research contributed to practice in several ways.Firstly, this study verified that the online business area is strongly related to technological applications, so the TAM model certainly plays an important role in explaining the consumer's behavior and this theory is very important to this field of practice.
Secondly, the results indicated that perceived enjoyment had a strong effect on online purchase intention.This means that the behavioral intention had been, practically, influenced strongly by the customers' experience with online purchases, and to some extent by other subjective factors.Therefore, to increase consumer's online purchase intention, e-commerce trading floors should improve the customer's experience by providing good products and services and guaranteeing that customers can trust the e-commerce exchange.
Thirdly, the strong role of e-commerce exchange image means that customers are likely to purchase from an e-commerce exchange that they believe to have a favorable image.This suggests that e-commerce exchanges need to pay attention to improving their image under the lens of customers.Moreover, the image was strongly influenced by perceived value, hence, the job of bringing good value to consumers is crucial for ecommerce platforms.
The significant roles of purchase enjoyment and ecommerce exchange image suggest that the e-commerce trading platforms should focus on lifting the customer's experience purchasing on e-commerce platforms.An important task to enhance a customer's experience is to improve the e-commerce platform's credibility.As discussed before, this requires e-commerce platforms to provide trustworthiness, dependability, and concern for customers.To lift customer experience to a new height and make customers excited with the experience of purchasing, e-commerce platforms should induce an ease to use platform which sells a wide range of good value products.In this case, an e-commerce platform should attract a large number of product providers (or e-stores) to sell their products on the e-commerce platform and provide quality services by monitoring e-stores strictly.The ecommerce platform needs to provide efficient logistics operations and effective reverse logistics functions for customers as well.
Finally, the results in Table 5 suggested that the hedonic value (exchange image, perceived enjoyment, and attitude) seemed to have a stronger influence on purchase intention than the utilitarian value (perceived value and perceived ease of use).Therefore, an important task of ecommerce exchanges is to bring more hedonic value to consumers to ameliorate their business.

Limitations and Future Research
Despite having useful contributions, this research still has several limitations that can shed light on the new research directions.In this research, we did not take the trust factor into account, and as discussed previously, low trust may reduce the level of self-efficacy and may have a moderating impact on the PBC Ò purchase intention relationship.As described by the (C c_Th ng_m i_ i n_t _và_Kinh_t _s , 2022), trust is among the most important factors prevent customers from purchase online.The main reason is that, in the case of purchasing online, customers cannot touch or have direct experiences with the purchasing products, therefore, they have to trust sellers for the products' quality and accompanying services.Practically, both the attitude and the self-efficacy (an important element of PBC) of customers are influenced by trust.Low trust is likely to lessen the strength of the attitude Ò purchase intention relationship.In addition, low trust reduces the self-efficacy of customers which leads to weakening the PBC Ò purchase intention relationship.
As described in the behavioral reasoning theory (BRT), behavioral intention can be affected by reasons (for and against).According to Westaby (2005), reasons are motivational (justification and defense) mechanisms that an individual uses to justify his/ her intention or behavior.In the online shopping environment, reasons have not been applied to study consumers' purchase intentions yet, so it will be interesting to implement the BRT framework in this research area and see how reasons interact with other factors in the framework.
For future research directions, we are interested in the implementation of a BRT model that integrates reasons (for and against) to have a deeper understanding of online purchase intention in practice.Future research may also take into account the moderating impact of trust on attitude and purchase intention.

Conclusions
This research proposed a framework that integrated the TPB and extended TAM, incorporating the e-commerce exchange image and perceived value to study consumer's online purchase intention from e-commerce exchanges in an emerging economy (e.g., Vietnam).The research results benefit both theory and practice.
The results showed that integrating the TPB and extended TAM can dramatically improve the predicting power of attitude toward online purchase and online purchase intention.Specifically, the TAM factors helped to improve the predicting power of attitude, and the perceived enjoyment, which is a share factor of both TAM and TPB, contributed a large portion to explaining online purchase intention.
The research results also identified the significant contribution of e-commerce exchange images to online purchase intention.This confirmed the strong role of e-commerce exchange images in practice.Perceived value also verified its strong role in influencing attitude toward online purchase and e-commerce exchange image in our case.

Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Figure 2 .
Figure 2. Results of hypothesis testing and influenced coefficients.

Table 2 .
Constructs and Indicators of the Model.

Table 4 .
Variance Inflation Factors (VIF) and Effect Size f 2 of Constructs in the Model.
Italicized values indicate very small effect.

Table 5 .
Total Indirect Impacts of Factors in the Model.

Table 6 .
Total Impacts of Factors in the Model.