Introduction
Background of the Study
The rise of the Internet and advancements in technology have profoundly transformed global lifestyles, leading to a significant surge in online shopping This evolution presents vast business opportunities by enabling commerce to thrive in cyberspace, connecting consumers and businesses without geographical constraints As a result, shoppers can conveniently purchase goods and services anytime, anywhere, free from the limitations of store hours, time zones, or traffic congestion.
The Internet has rapidly become a popular medium for shopping, with a significant increase in online purchases over the past decade According to ACNielsen (2008), 875 million people worldwide made online purchases in 2008, up from 627 million in 2006 South Korea leads in online shopping, with 99% of internet users engaging in e-commerce, followed closely by the United Kingdom, Germany, and Japan at 97%, while the United States ranks eighth at 94%.
According to a study by Indvik (2013), the e-commerce market is projected to expand significantly, increasing from $857 billion in 2011 to an estimated $1,860 billion by 2016 Additionally, online retail sales in the United States, which totaled $308 billion in 2011, are anticipated to rise to approximately $546 billion in 2016.
Online shopping is a relatively new phenomenon in Vietnam, marking a significant technological advancement in the retail sector According to the Vietnam E-commerce Report 2013 by the Vietnam E-Commerce and Information Technology Agency (VECITA), only 15% of enterprises in the wholesale and retail sectors have engaged with e-marketplaces, indicating a moderate level of participation in online shopping services.
According to VECITA (2014), the finance and estate sectors exhibited the highest participation rates in e-marketplaces, at 28% and 20% respectively Additionally, 85% of surveyed enterprises reported that their experiences in e-marketplaces yielded moderate to good efficiencies.
A 2013 VECITA survey revealed that 15% of respondents rated their e-commerce efficiencies as low The survey also indicated that 41% of businesses experienced revenue growth through e-commerce channels, while 13% reported a decline and 46% noted no significant change Notably, the effectiveness of e-commerce has remained relatively stable over the past few years (VECITA, 2014).
Source: Survey of Vecita in 2013
Figure 1.1: Trends of revenue from electronic means
In 2013, Vietnam's internet users comprised 36% of the population, with 57% engaging in online transactions, according to the Vietnam E-commerce and Information Technology Agency Each online buyer spent an average of $120, primarily on clothing, shoes, and cosmetics, followed by technology, airline tickets, food, and books The e-commerce market in Vietnam reached $2.2 billion that year, and projections indicated that by 2015, internet user penetration would rise to 40-45%.
By 2015, the e-commerce legal framework is expected to be finalized, alongside significant advancements in logistics and payment infrastructure Research indicates a rising trend in online shopping, with the number of Internet users making purchases projected to grow Additionally, each online consumer's spending is anticipated to increase by $30 in 2015 compared to 2013 However, despite these developments, the e-commerce market remains relatively small when compared to global leaders such as the U.S ($343 billion), Japan ($127 billion), the UK ($124 billion), and China ($110 billion) (PwC, 2013).
Appropriately, although sales of products from the Internet account for only a small percentage of total retail sales, millions of consumers shop and buy on the Internet
In Vietnam, many companies are hastily establishing an online presence, despite uncertainties regarding the impact of the internet on their businesses To boost online shopping in the country, it is crucial to prioritize understanding consumer behavior and the factors influencing online shopping Research shows that a significant majority of Vietnamese, particularly the youth, engage in non-shopping activities online, such as seeking information (87%), using social networks or forums (73%), and checking emails (71%), with only 20% participating in online shopping (VECITA, 2014).
Source: Survey of E-commerce and Information Technology Agency 2013
Figure 1.2: Purposes of accessing Internet
According to Levin, Levin, and Weller (2005), consumers utilize various information sources to verify product quality and increase satisfaction, influenced by the significance of different product attributes Researchers categorize online products based on whether their primary attributes are digital or non-digital (Biswas & Biswas, 2004; Lal & Sarvary, 1999) Digital products, which can be fully communicated online, generally carry less inherent risk compared to non-digital products that necessitate physical inspection Levin et al (2005) highlight that consumers value the tactile experience of apparel, leading them to favor traditional stores for such purchases Conversely, for digital products like books, consumers prioritize immediate access to information, making online shopping more appealing.
Numerous studies have explored the factors influencing consumers' online purchase intentions by utilizing traditional behavioral intention models and theories Key frameworks include the Theory of Reasoned Action (TRA), developed by Fishbein and Ajzen in 1975, the Technology Acceptance Model (TAM) introduced by Davis in 1989, and the Theory of Planned Behavior (TPB) proposed by Ajzen.
The Technology Acceptance Model (TAM) serves as the foundational framework for this research, as it is recognized as the most suitable model for effectively explaining behavioral intentions related to Information Technology (IT) and Information Systems (IS) This model is known for its validity and reliability in assessing user acceptance and usage of technology.
Problem Statement
The influence of personal attitudes on consumer decision-making and behavior is well-recognized, particularly in how service attitudes connect consumer characteristics to their needs (Ajzen, 1991) Online shopping behaviors are shaped by individual traits, including personality, demographics, and perceptions of online shopping benefits (Goldsmith & Flynn, 2004) Understanding the significance of various determinants influencing consumer choices is essential for grasping the reasons behind these behaviors Additionally, behavioral intention is closely linked to an individual's attitude towards the action, as awareness of the behavior's benefits, self-efficacy, and control over resources contribute to the development of the Theory of Reasoned Action (TRA) and the Theory of Planned Behavior.
The Theory of Planned Behavior (TPB) and the Technology Acceptance Model (TAM) offer valuable insights into consumer intentions regarding online shopping By integrating the principles of the Theory of Reasoned Action (TRA), these models help explain and predict how consumers adopt e-business practices Understanding these frameworks can enhance our comprehension of the factors influencing online shopping behavior.
Choi and Geistfeld (2004) highlight that online shopping is often viewed as riskier compared to traditional brick-and-mortar retail transactions Furthermore, given that consumer behavior varies across cultures, it remains uncertain if findings from Western consumers regarding online purchasing intentions are applicable to Eastern markets, such as Vietnam.
Dann and Dann (2001) highlight that the growth of online shopping is driven by the easy access to information, making consumers more informed and efficient Despite numerous studies on global online shopping trends, there remains a need for in-depth analysis of online shopping intentions in specific countries (Salisbury et al., 2001) In Vietnam, the internet is still emerging as a vital link between retailers and consumers, with customer retention posing significant challenges for e-retailers (Dai et al., 2014) To enhance understanding in this field, it is essential to seek precise answers to pertinent questions.
In 2013, Vietnam experienced significant growth in online transactions, particularly in the business-to-business (B2B) and business-to-consumer (B2C) sectors The emergence of platforms like muachung.vn, hotdeal.vn, and nhommua.vn at the end of 2010 transformed the e-commerce landscape, driving customer engagement and generating substantial economic benefits Despite this progress, the e-commerce market in Vietnam remains fragmented, with key players in the consumer-to-consumer (C2C) and B2C segments, such as vatgia.com, enbac.vn, and leading e-retailers like tiki.vn and lazada.vn Thegioididong.com and nguyenkim.com are recognized as the most popular online sales channels among Vietnamese consumers This study aims to provide insights that will help e-retail companies enhance their appeal and encourage more consumers to shop online in Vietnam.
The reliance on various information sources, such as search versus experience, in traditional shopping also extends to online shopping contexts Research indicates that consumers depend on different information sources based on the product type when making purchases in physical stores (Nelson, 1970) However, studies focusing on online shopping in Vietnam have largely overlooked how product categories affect consumer intentions (Dai et al., 2014) Furthermore, the categorization of products, such as search and experience, that is applicable in traditional retail may not seamlessly translate to online environments While certain products, like apparel, carry inherent risks when purchased online, existing literature has predominantly failed to demonstrate how online shoppers' risk perceptions differ by product category and how these perceptions impact their purchasing intentions, with the notable exception of Biswas and Biswas (2004).
Vietnamese consumers show significant potential for online shopping, yet there remains a gap in understanding their preferences For online retailers to thrive, they must deliver real value to these consumers Therefore, it is crucial for internet marketers to grasp customers’ expectations and shopping intentions Conducting thorough research can enable online retailers to better understand their clientele, meet their needs and desires, and ultimately create meaningful value for them.
Research Objective
This research aims to identify the factors influencing consumers' online purchase intentions in Ho Chi Minh City, Vietnam, with a focus on product category as a moderating variable The study specifically seeks to achieve key objectives related to understanding these determinants and their impact on consumer behavior in the context of online shopping.
- To determine factors influencing consumers’ online purchase intention in Ho Chi Minh City
- To examine the moderating effect of product category on the relationship between the factors and consumers’ online purchase intention.
Research Question
Based on the discussion above, and to accomplish the objective of this study, the proposed questions to be answered in this research are as follow:
- Does perceived usefulness of online shopping contribute towards consumer’s online purchase intention?
- Does perceived ease of use contribute towards consumer’s online purchase intention?
- Does perceived risk contribute towards consumer’s online purchase intention?
- Does brand orientation contribute towards consumer’s online purchase intention?
- Does prior online purchase experience contribute towards consumer’s online purchase intention?
- Are there differences above impacts among various product categories?
Above factors (perceived usefulness, perceived ease of use, perceived risk, brand orientation, and prior online purchase experience) are further discussed in the second chapter under conceptual framework.
Research contributions
This study explores an explanatory model to understand consumers' behavioral intentions in online purchasing, utilizing product category as a key control variable The Technology Acceptance Model (TAM) is selected for its proven effectiveness in elucidating the differences between behavioral intentions and actual purchasing behavior, particularly in technology-related products This model is deemed suitable for analyzing online shopping intentions and the moderating effects of product types, an area with limited existing literature Consequently, this research adds valuable insights to the understanding of consumer behavior in the online marketplace.
This study's findings extend beyond theoretical insights, offering significant practical contributions to business management Key applications include enhancing decision-making processes, improving operational efficiency, and fostering innovation within organizations.
- This study provides useful information for the business management to prioritize their resources for time, budget allocation, human resources, and investment
This study aims to enhance marketers' understanding of consumer intentions related to online shopping by examining the behavioral factors that influence purchasing decisions By identifying key variables affecting these intentions within specific product categories, marketing managers can develop targeted marketing strategies that meet the needs of online shoppers, ultimately increasing consumer satisfaction and optimizing the marketing mix.
- It provides useful input so that Web developers can design sites and pages of which content and layout is more compelling and effective to attract more business from consumers
Policymakers can leverage this information to enhance online business by improving infrastructure and regulations related to the Internet, fraud prevention, security, and privacy These improvements aim to facilitate and encourage greater consumer participation in online buying and selling.
Scope of the study
Due to the nature of this research, a number of delimitation of scope had to be set for the study as below:
This research focuses exclusively on residents of Ho Chi Minh City aged 15 and older, chosen for its notable Internet penetration rate.
According to Cimigo's 2011 report, over 50% of the population in urban Vietnam has accessed the internet, with Ho Chi Minh City surpassing this average at a notable 62%.
- Survey is conducted in this research to gather information among people who has browsed online shopping pages (visitors) as well as those who has experience in online shopping (purchasers)
According to ACNielsen (2008), the top items purchased online worldwide include books (41%), clothing/accessories/shoes (36%), and video/DVD/games and airline ticket reservations (24%) In Vietnam, clothing, shoes, and cosmetics dominate online purchases at 62%, followed by technology products at 35%, household items at 32%, air tickets at 25%, and books and stationery at 20%.
Source: Survey of E-commerce and Information Technology Agency 2013
Figure 1.3: Popular online products on e-commerce website purchased
Clothing, footwear, electronics, books, and stationery were the most frequently sold items on e-commerce websites, making up 79% of the surveyed platforms This statistic reflects a strong alignment between supply and demand, with 62% of consumers opting to buy these products online Consequently, the study focused on clothing, electronics, and books as key product categories.
Source: Data gathered by E-commerce and Information Technology Agency 2013
Figure 1.4: The types of products or services introducing on e-commerce websites
Organization of the Study
This research is organized as five chapters
Chapter 1 offers an introduction to the study, including background to the study, statement of the problem, research objectives, research questions, the significance of the study, and scope of the study
Chapter 2 makes a brief reference to the theories that have been used These literatures summarize briefly the knowledge of recent studies, describes the conceptual model, and hypotheses
Chapter 3 presents research methodology, including the research design, population and sampling plan, the instruments, procedures and methods of data analysis
Chapter 4 presents the description and analysis of the data collected
Finally, Chapter 5 reports in detail research findings and implications Further, this chapter concludes with discussion on the results derived, limitations and future research.
Literature Review
Introduction
This chapter aims to review pertinent literature to provide an overview of the research area, enabling the researcher to develop a conceptual model for assessing the factors that influence consumers' online purchase intentions.
Behavioral Intention Models
The Theory of Reasoned Action (TRA), the Theory of Planned Behavior (TPB), the Technology Acceptance Model (TAM), and the Diffusion of Innovations are among the most widely used frameworks for understanding behavioral intentions towards technological products These theories provide valuable insights into the factors influencing user acceptance and the adoption of new technologies.
Innovation Most of these theories have been developed from the Theory of Reasoned
The Diffusion of Innovation Theory, proposed by Fishbein and Ajzen (1975), primarily examines the adoption of the Internet rather than online purchasing behaviors Consequently, this research focuses on three prevalent theories—Technology Acceptance Model (TAM), Theory of Planned Behavior (TPB), and Theory of Reasoned Action (TRA)—to better understand the relationships among consumer beliefs, attitudes, and behavioral intentions when buying products online.
2.2.1 Theory of Reasoned Action (TRA)
The theory of reasoned action (TRA), developed by Fishbein and Ajzen (1975,
The Theory of Reasoned Action (TRA), proposed by Ajzen in 1980, asserts that individual behavior is primarily influenced by behavioral intentions This model facilitates the prediction of both attitudes and behaviors, highlighting the distinction between behavioral intention and actual behavior This separation enables a deeper understanding of the factors that may limit the impact of attitudes on behavior.
According to Fishbein (1980), behavioral intentions encompass an individual's attitude towards a specific behavior and the subjective norms influencing that behavior The actual adoption of an innovation relies on the individual's intention to use it, which is shaped by their positive or negative feelings towards the behavior Additionally, the subjective norm reflects an individual's perception of whether significant others believe that the behavior should be executed.
A person's voluntary behavior is influenced by their attitude towards that behavior and their perception of how others would react to it This combination of personal attitude and social norms shapes an individual's intention to act, as outlined by Fishbein and Ajzen (1975).
Figure 2.1: Theory of reasoned action model
The Theory of Reasoned Action (TRA) has limitations in fully explaining the mechanisms behind innovation use and individual behavioral intent Critics argue that it overlooks the significance of social factors that can influence behavior in real-life scenarios (Grandon & Peter, 2004) Despite this, TRA's inclusion of subjective norms is beneficial in certain contexts Research indicates that TRA effectively predicts consumer behavioral intentions across various products, including fashion, beer, toothpaste, dog food, mineral water, and facial tissue (Grandon & Peter, 2004).
2.2.2 Theory of Planned Behavior (TPB)
The theory of planned behavior, proposed by Ajzen in 1991, enhances the predictive power of the theory of reasoned action by incorporating perceived behavioral control This theory explores the connections between beliefs, attitudes, behavioral intentions, and actual behaviors across various domains, including advertising, public relations, and healthcare Perceived behavioral control highlights the significance of individuals' perceptions regarding the ease or difficulty of executing a specific behavior.
According to the Theory of Planned Behavior (TPB) proposed by Ajzen in 1985, an individual's behavior is influenced by their intention and perceived behavioral control Ajzen (1991) emphasized that the key element of TPB is the person's intention to engage in a specific behavior, which reflects the motivational factors that drive actions This intention serves as an indicator of an individual's willingness to perform the behavior.
Figure 2.2: Theory of planned behavior
The Theory of Planned Behavior has been applied in various research areas, including weight loss, sexual behavior, waste recycling, student class attendance, spreadsheet software usage, and information technology, as noted by Richard & Joop de Vries (2000) and Taylor & Todd.
1995) Nevertheless, the Theory of Planned Behavior lacks sufficient scale development for studying online shopping behavior
The Technology Acceptance Model (TAM) extends the Theory of Reasoned Action (TRA) as proposed by Ajzen and Fishbein (1980), with Davis (1989) and Davis, Bagozzi, and Warshaw (1989) introducing TAM to elucidate user acceptance of information technology This model illustrates how external variables shape beliefs, attitudes, and intentions toward technology use Central to TAM are two cognitive beliefs: perceived usefulness, which refers to the degree to which a user believes that using a particular technology will enhance their performance, and perceived ease of use, which reflects the degree to which a user feels that using the technology will be free of effort.
Perceived usefulness refers to an individual's belief that utilizing a specific system can improve their job performance (Davis, 1989) Meanwhile, perceived ease of use is defined as the extent to which a person feels that operating the system requires minimal effort (Davis).
1989) TAM also suggests that the external factors affecting intention and actual use through mediated effects on perceived usefulness and perceived ease of use (see Figure 2.3)
The Technology Acceptance Model (TAM) effectively explains and predicts user behavior, emphasizing two key constructs: perceived ease of use and perceived usefulness Research consistently demonstrates that these constructs exhibit high reliability and validity, making TAM a valuable framework for understanding technology adoption (Adam, Nelson, & Todd).
The Technology Acceptance Model (TAM) is widely recognized as the leading theory for examining the behavioral intentions behind the use of technological products, as highlighted in numerous studies and literature reviews Its robustness has been validated through various experiments, organizational surveys, and research across multiple domains, including microcomputers, software, spreadsheets, email, and the World Wide Web Furthermore, TAM has been effectively tested in diverse countries such as the United States, Canada, Taiwan, China, India, Thailand, Malaysia, and Iran.
Singapore (Jayawardhena, Wright, & Dennis, 2007) Hence, TAM seems to be a suitable model to develop a conceptual model to examine the factors affecting online purchase intention in Vietnam
This study proposes a preliminary theoretical model based on the Technology Acceptance Model (TAM) to examine the factors influencing online purchasing behavior Empirical research consistently demonstrates that TAM effectively accounts for variations in behavioral intentions and actual behaviors across various information technologies Recognized for its robustness and simplicity, this model sets the foundation for the unique preliminary study model introduced here Additionally, specific constructs and measurement scales will be established to evaluate the theoretical model concerning the determinants of online purchasing.
Overview of the Preliminary Model
The Technology Acceptance Model (TAM) has been widely applied across various contexts, backed by numerous empirical studies While the original model included attitudes as a key component, many researchers have since opted to exclude these constructs from their adaptations of the model.
The removal of attitudes from the Technology Acceptance Model (TAM) is supported by three key reasons First, previous studies indicated that attitudes have a non-significant effect on behavioral intention, with perceived usefulness being a more crucial determinant (Davis et al., 1989) Second, simplifying the model by excluding attitudes could enhance its predictive capability without significantly reducing its effectiveness (Mathieson, 1991; Davis, 1989) Finally, the TAM operates on the premise that attitudes are inherently reflected in perceived usefulness, suggesting that individuals may adopt technology based on its utility rather than their feelings toward it (Davis et al., 1989) Consequently, attitudes are excluded from the structural model proposed in this study.
In the Technology Acceptance Model (TAM), perceived usefulness (PU) is identified as the primary factor influencing behavioral intention, while perceived ease of use (PEOU) indirectly affects this intention through PU (Davis et al., 1989) Research has demonstrated that both PU and PEOU exhibit high reliability and validity (Adam et al., 1992), and they significantly account for the variance between behavioral intentions and actual behaviors (Kamarulzaman, 2007) Therefore, this study will retain the constructs of perceived usefulness and perceived ease of use as foundational elements of the original TAM.
While the Technology Acceptance Model (TAM) has been widely adopted, Lee, Kozar, and Larsen (2003) suggest that it should be integrated and extended for a deeper understanding of IT adoption According to Amin (2007), there are three main approaches to enhance TAM: incorporating factors from related models (Dai, Forsythe, & Kwon, 2014), adding alternative belief factors (Moon & Kim, 2001), and exploring antecedents and moderators of perceived usefulness and ease of use (Kamarulzaman, 2007) Many studies may utilize one or a combination of these approaches to further develop the original TAM framework (Amin, 2007; Chang, 2004).
The Technology Acceptance Model (TAM) has undergone significant evolution, leading to the development of TAM2, which enhances the original framework by incorporating factors such as social influence, cognitive instrumental processes, and user experience to better explain perceived usefulness and intentions to use technology This revised model has been validated in both voluntary and mandatory contexts, demonstrating a strong explanatory power for user adoption, accounting for 60 percent of the variance in technology acceptance (Venkatesh & Davis, 2000).
The Technology Acceptance Model (TAM) has been widely utilized in various studies to understand consumers' online purchase intentions across different countries Researchers have expanded TAM by incorporating factors such as gender, product category, brand orientation, perceived risk, online trust, and prior online purchase experience These extensions enhance the model's applicability in analyzing consumer behavior in the digital marketplace.
Thamizhvanan & Xavier, 2013; Brown et al., 2003), social influence (Zamri & Idris, 2013; Abadi, Hafshejani, & Zadeh, 2011)
Numerous studies indicate that behavioral intention significantly influences usage and can effectively predict actual behavior in real-world scenarios (Igbaria et al., 1995; Taylor & Todd, 1995; Mathieson, 1991) Consequently, this research examines purchase intention as a key predictor of actual purchasing behavior The following section provides a comprehensive overview of the preliminary model, including its constructs and hypotheses.
Conceptual Framework and Proposed Hypotheses
This study utilizes the Technology Acceptance Model (TAM) as a foundational framework to examine the factors influencing consumers' online purchase intentions By leveraging the reliability and validity of perceived usefulness (PU) and perceived ease of use (PEOU) from TAM, the research aims to enhance the understanding of online consumer behavior in Ho Chi Minh City To achieve this, additional constructs such as brand orientation, perceived risk, and prior online purchase experience are incorporated to improve the explanatory power and predictive capabilities of the model.
Perceived usefulness (PU) refers to the belief that using a specific system can enhance job performance (Davis, 1989) This concept is crucial for understanding consumer behavior, particularly in online shopping, where potential buyers may react positively or negatively based on their perception of a system's usability and usefulness The effectiveness of Internet retailing hinges on the benefits it provides compared to traditional retail, emphasizing the importance of PU in influencing purchasing intentions Although research on Internet retailing from the Technology Acceptance Model (TAM) perspective is limited, the PU construct has received substantial support across various technological applications, including a positive correlation between PU and intention in Intranet media (Dai et al., 2014).
(2007); Moon and Kim (2001); Venkatesh and Davis (2000); Igbaria et al (1995);
Mathieson (1991); Davis et al (1989) reported that perceived usefulness is significant and positively influences the behavioral intention Hence, it is expected that:
H1: Perceived usefulness has a significant positive impact on the consumer’s online purchase intention
2.4.2 Perceived ease of use (PEOU)
Perceived ease of use, as defined by Davis (1989), is the extent to which individuals believe that utilizing a system requires minimal effort This concept highlights the convenience of online shopping compared to traditional shopping, which often involves challenges such as limited time, anxiety, parking issues, crowds, and traffic (Yulihasri, Islam, & Daud, 2011) Amin (2007) found that ease of use directly influences consumers' intentions to purchase online, reinforcing that consumers view online shopping as beneficial when it is user-friendly Additionally, Peng, Wang, and Cai (2008) emphasize that perceived ease of use significantly affects consumers' willingness to shop online, suggesting that characteristics such as controllability, flexibility, ease of learning, clarity, skill acquisition, and overall usability are crucial Yulihasri et al (2011) further assert that ease of use impacts consumers' online purchasing intentions, leading to the hypothesis that enhancing usability can promote online shopping engagement.
H2: Perceived ease of use has a significant positive impact on the consumer’s online purchase intention
The concept of perceived risk, introduced by Bauer in 1960, refers to the uncertainties consumers encounter when considering the purchase of a new product or service (Taylor, 1974) In online shopping, these perceived risks can be heightened due to the lack of physical access to products and sales personnel, making customers more apprehensive about their buying decisions (Kwek et al., 2010).
In the online retail environment, customers often encounter various risks, including security, privacy, and product risks (Chen & Barnes, 2007) Security plays a crucial role in establishing customer trust, as it pertains to the protection of information technology systems and sensitive financial data, such as credit card information (Bart et al., 2005) Additionally, perceived privacy is defined as consumers' ability to control the dissemination of their information during online transactions (Dai et al.).
Research indicates that consumers perceive a higher level of product risk when shopping for apparel online compared to traditional stores (Goldsmith & Goldsmith, 2002) This heightened product uncertainty can adversely impact online shopping intentions (Bhatnagar, Misra, & Rao, 2000) Various studies (Dai et al., 2014; Lee et al., 2003; Miyazaki & Fernandez, 2000; Warrington, Abgrab, & Caldwell, 2000) have consistently demonstrated that perceived risk significantly negatively affects the intention to purchase online Therefore, it can be hypothesized that perceived risk influences online shopping behavior.
H3: Perceived risk has a significant negative impact on the consumer’s online purchase intention
A brand is a unique identifier, such as a name or symbol, that distinguishes a seller's products or services from competitors (Aaker, 1991) In the online marketplace, brand identity serves as a cognitive anchor, helping customers navigate uncertainty (Julie, Anthony, & Dena, 2006) Customers often rely on trusted corporate and brand names as substitutes for product information when making online purchases (Julie et al., 2006) The concept of brand orientation refers to a product's alignment with its brand, which evolves in response to market intelligence (Aaker, 1991).
Research by Kamins and Marks (1991) highlights that a strong brand image and familiarity significantly enhance consumer attitudes towards the brand and their willingness to purchase Subodn and Srinivas (1998) further emphasize that brand image comprises the information and expectations consumers associate with a product, service, or company, directly influencing their purchase decisions and perceptions of product quality Additionally, numerous studies indicate that brand loyalty has a substantial positive effect on purchase intentions within traditional online retail environments.
(Thamizhvanan & Xavier, 2013; Kwek et al., 2010) Studies carried out by Ling et al
(2010); Jayawardhena et al (2007) conclude that brand orientation is positively related to the customer online purchase intention Thus, we propose:
H4: Brand orientation has a significant positive impact on the consumer’s online purchase intention
2.4.5 Prior online purchase experience (POPE)
Helson (1964) posited that an individual's judgment in decision-making is influenced by their past experiences, contextual background, and the stimuli presented Despite the growing popularity of online shopping, many consumers still view it as riskier compared to traditional shopping methods (Laroche et al., 2005) As a result, online shoppers heavily rely on the quality of their past experiences, which can only be gained through previous purchases.
Repeated actions lead to reduced time spent on future iterations, highlighting the impact of prior experiences on future behaviors In online shopping, customers assess their experiences based on various factors, including service quality, perceived risks, payment processes, product information, visual appeal, privacy, personalization, security, navigation, delivery terms, and overall enjoyment (Burke, 2002).
Customer experience significantly influences the growth of Internet shopping, as highlighted by Elliot and Fowell (2000) The overall buying process is crucial for consumers, who are more inclined to purchase online after prior positive experiences with products, according to Kwek et al (2010) Seckler (2000) notes that as individuals gain confidence and skills through initial small purchases, they become more willing to engage in larger online transactions Ultimately, consumers' responses to marketing messages, advertising strategies, website design, and brand interactions play a vital role in shaping their online shopping behavior.
Customers with prior online shopping experiences tend to feel more confident and less uncertain about making purchases online (Solomon et al., 2007) In contrast, those who have never shopped online are often more risk-averse (Lee et al., 2003) Positive past experiences can encourage customers to continue shopping online, fostering future purchases (Shim et al., 2001) Conversely, negative evaluations of previous online shopping experiences can lead to reluctance in engaging in future online transactions.
Research indicates that a positive online purchase experience significantly influences customers' future intentions to shop online, as supported by studies from Solomon et al (2007), Laroche et al (2005), and Shim et al (2001).
H5: Prior online purchase experience has a significant positive impact on the consumer’s online purchase intention
Previous research on information asymmetry categorizes products and services into three types: search, experience, and credence goods (Brush & Artz, 1999) Search goods are characterized by features that can be easily assessed prior to purchase In contrast, experience goods have attributes, such as quality or price, that are challenging to evaluate beforehand but can be determined through consumption Lastly, credence goods are those for which the consumer finds it difficult or impossible to assess their utility These categories form a continuum from "easy to evaluate" search goods to "difficult to evaluate" credence goods.
According to Dimoka, Hong, and Pavlou (2012), information signals play a crucial role in reducing product uncertainty, particularly when consumers can trust a brand without knowing its name, relying instead on detailed product descriptions and third-party assurances Higher prices are often perceived as indicators of better quality, although this perception can differ by product category (Lichtenstein & Burton, 1989) The internet allows consumers to assess the quality of search goods and services at minimal cost, meaning that excessive pricing can deter purchase intentions In contrast, evaluating experience goods requires actual usage, while credence goods necessitate additional information for value assessment For these types of products, a personalized approach from providers is essential, as it limits customers' ability to compare prices effectively (Brush & Artz).
1999) Books, clothing and electronics, used in the study, represented three product types— search, experience and credence qualities goods
Chapter summary
This chapter outlines the significance of understanding the factors that influence consumers' online purchase intentions, particularly focusing on the moderating role of product categories A literature review reveals that the Technology Acceptance Model (TAM) serves as a robust foundation for this study due to its theoretical strength, reliability, and widespread acceptance Key constructs from TAM, namely perceived usefulness and perceived ease of use, significantly explain variations in behavioral intentions Consequently, a preliminary research model is proposed, incorporating these TAM constructs along with three additional factors: prior online purchase experience, perceived risk, and brand orientation, to create a tailored model for consumers in Ho Chi Minh City, Vietnam Notably, the study posits that product category acts as a moderating variable affecting the relationship between these factors and online purchase intentions, leading to the formulation of six testable hypotheses.
Research Method
Introduction
This chapter outlines the research design of the study, detailing the measurement and questionnaire design, the instruments utilized for data collection, the methods employed for gathering data, and the analytical techniques applied for data analysis.
Research design
This study employed a quantitative research design, utilizing the survey method for its effectiveness in gathering data on human attitudes, behaviors, and characteristics, as noted by Gray (2009) Surveys are widely recognized in social, business, and information science for their ability to quickly and cost-effectively collect substantial amounts of data, making them an ideal choice for this research (Zikmund et al., 2010).
To conduct the survey research, a literature review established the conceptual foundation, followed by the design and pre-testing of a self-administered questionnaire to enhance clarity and minimize non-response (Gray, 2009) The finalized questionnaire was then distributed to the target population through face-to-face and web-based methods after pilot testing Data collected were imported into SPSS software for various analyses, including exploratory factor analysis, reliability analysis, correlation analysis, and multiple regression The findings and their implications were subsequently presented and discussed based on the results obtained.
Measurement and Questionnaire Design
The instrument was developed by selecting measures from validated questionnaires utilized in previous research Multi-scaled items to assess the constructs were gathered from various studies, with Table 3.1 presenting the measurements and their sources These measurements will be clarified and modified to align with the objectives of this study.
I intend to purchase online in the future
Chou & Kimsuwan (2013); Wen, Prybutok, & Xu (2011); Kwek, Tan, & Lau (2010) ; Moon & Kim (2003)
I will keep using online shopping in the future
I think it would be very good to purchase products through Internet in addition to traditional methods
I would strongly recommend others to use online shopping
I would frequently use “online shopping”
It is easier to use the Internet to find products that I want to buy
Wen, Prybutok, & Xu (2011); Moon & Kim
It is easy for me to learn how to use online shopping, even as the first time
It will be impossible to make an online transaction without expert help
Using online shopping do not requires a lot of mental effort
It is easy for me to become skillful at using online shopping
Using online shopping enables me to search and buy something more quickly
Wen, Prybutok, & Xu (2011); Moon & Kim
Use online shopping enables me to save my money Using online shopping helps me to make better purchase decisions (efficiency and effectiveness)
Use online shopping makes it easier for me to obtain goods
Overall, I believe “online shopping” is useful and advantageous
If I buy products/services from a web-retailer that I am familiar with, I would prefer to buy well—known brand name
Thamizhvanan & Xavier (2013); Kwek, Tan, & Lau (2010); Ling, Chai, & Piew
It is important for me to buy products/services from the web retailer with well-known brand names
Once I find a brand I like through web-shopping, I stick with it
I think paying with a credit card to shop online is risky
Dai, Forsythe, & Kwon, (2014) ; Ling, Chai, & Piew (2010); Broekhuizen & Huizingh, (2009)
I think online shopping would put my privacy at risk
I am worried whether I can get a product on time
I am worried that product quality may not meet my expectations
I am concerned that I may NOT receive the item purchased
I may buy the same product at a lower price from somewhere else
It is DIFFICULT for me to compare the quality of similar products
After doing an online shopping, I know reliable retailer's web sites Thamizhvanan &
After doing an online shopping, I become more skillful
I am experienced with the use of the retailer's web site
The questionnaire was structured into three key sections: the first section gathered demographic information about the respondents, the second section explored their online purchasing behaviors, and the final section outlined the independent and dependent variables to be analyzed in the survey.
The questionnaire was developed in English and Vietnamese All items in the questionnaires adopted five-point Likert scales, where 1 = “strongly disagree” and 5 “strongly agree”
Before conducting the main survey, a pretest was performed to assess the clarity and effectiveness of the questionnaire and survey process A convenience sample of 25 university students from the International School of Business (ISB) participated in this pilot test Participants were asked to complete the initial survey questions on-site, with the understanding that their responses would not be used for data analysis Feedback on the clarity and ease of understanding of the items led to modifications in the questionnaire The reliability of the questionnaire was evaluated using SPSS, resulting in a Cronbach’s alpha of 0.854, with individual variable values exceeding 0.7, indicating high reliability according to Sweet and Grace-Martin (2008) Consequently, no changes were made to the questionnaire, and the detailed versions are available in Appendix A1 (English).
Data Collection Method
The target population of this study was focused on the Internet users aged between
16 and 45 including online shopping visitors (non-shoppers) and purchasers (shoppers) in
Ho Chi Minh City, Vietnam In this study, purchasers are those consumers who had been involved in the online transaction in the last six months
This research utilized a self-administered questionnaire survey, combining both web-based and face-to-face methods The web-based survey was chosen for its ability to gather large, geographically diverse samples efficiently, yielding faster response rates and simplifying data processing, as responses can be directly imported into spreadsheets or analysis software at a lower cost (Gunn, 2002) Additionally, it offers the advantages of a mail survey, allowing respondents to thoughtfully consider and answer questions at their own pace (Sekaran).
Web-based surveys offer the advantage of allowing respondents to reference websites they visited or purchased from prior to answering questions Conversely, face-to-face surveys provide significant benefits, including a clear structure, adaptability, and flexibility through personal interaction These surveys also enhance data quality, as they allow for the use of physical stimuli and enable the observation of respondents (Alreck & Settle, 2004).
This research utilized convenience sampling to efficiently gather information at a low cost, as noted by Sekaran (2005) An invitation email containing a survey link was distributed to 600 individuals, including colleagues and friends on platforms like Facebook and Google Circle, inviting them to participate in a study on consumer Internet shopping behaviors The questionnaire employed a forced-answer format, ensuring that participants could not submit their responses without complying with the rules Data was securely stored in Google Drive, accessible only to the researcher, and all responses were collected anonymously without requiring any personal identification or detailed demographic information from participants.
A total of 600 questionnaires were distributed through a web-based survey, yielding 242 responses, of which 174 were usable, resulting in a response rate of 42% Additionally, 160 on-the-spot questionnaires were completed by university students and building officers in Ho Chi Minh City, selected for their high internet and online shopping usage (VECITA, 2014) From 185 on-the-spot surveys distributed, 173 were collected, but 28 were deemed unusable due to incompleteness or invalidity Ultimately, 145 valid responses from the on-the-spot survey were included, leading to a combined sample size of 319 for testing and analysis.
Data Analysis Method
The data analysis was conducted using SPSS version 20, beginning with descriptive statistics to summarize the variables Factor analysis was then performed to group the variables, utilizing principal components extraction with Varimax rotation To ensure internal consistency, the reliability of each measure was assessed using Cronbach’s alpha Finally, correlations among the variables were calculated, and multiple regression analysis was employed to explore the relationships represented in each hypothesis.
Once the data had been collected and prepared for analysis, basic statistical and descriptive analysis was developed for the study
In this study, the author utilized Exploratory Factor Analysis (EFA) to validate scale items that were either adapted from existing research or created by the researchers themselves EFA is typically employed when there are no predefined hypotheses regarding the underlying factor structure of the measures used.
Effective factor analysis (EFA) yields more precise results when each factor is represented by multiple measured variables Research indicates that having a minimum of 3 to 5 measured variables per factor enhances the accuracy of the analysis (Hogarty et al., 2005).
Correlation matrix analysis is essential in determining the significance of loadings, as outlined by Hair, Anderson, Tatham, and Black (1995), who established a guideline where loadings of ±0.30 are considered minimal, ±0.40 important, and ±0.50 practically significant If no correlations exceed the ±0.30 threshold, researchers should reevaluate the suitability of factor analysis as their chosen statistical method (Tabachnick & Fidell, 2007; Hair et al., 1995).
Kaiser-Meyer-Olkin (KMO) Measure of Sampling Adequacy/Bartlett's Test of
Sphericity Several tests should be used to assess the suitability of the respondent data for factor analysis These tests include Kaiser-Meyer-Olkin (KMO) Measure of
Sampling adequacy is assessed using the Kaiser-Meyer-Olkin (KMO) index and Bartlett's Test of Sphericity The KMO index, which ranges from 0 to 1, indicates that values of 0.50 or higher are deemed appropriate for conducting factor analysis Additionally, for factor analysis to be valid, the Bartlett's Test of Sphericity must yield a significant result, with a p-value less than 0.05.
The researcher employed Principal Component Analysis (PCA) for factor extraction, as it is widely recognized as the default method in numerous statistical programs, making it the most frequently used technique in Exploratory Factor Analysis (EFA) (Thompson, 2004) Additionally, Henson and Roberts (2006) recommended utilizing PCA to develop preliminary solutions in EFA This study implemented various extraction rules and approaches, including Kaiser’s criterion, which focuses on eigenvalues greater than one.
1 rule) (Kaiser, 1960), the Scree tests (Cattell, 1966), the cumulative percent of variance extracted, and parallel analysis (Horn, 1965)
Rotation analysis enhances the interpretability of data by maximizing high item loadings and minimizing low item loadings, leading to a simplified solution (Hogarty et al., 2005) In this study, the author employed varimax rotation, a widely used method for orthogonal rotation that ensures all factors remain uncorrelated.
Following the factor analysis, the data underwent reliability testing, focusing on internal consistency The reliability of the scale was assessed using Cronbach’s alpha coefficient, with values exceeding 0.7 indicating high reliability and strong internal consistency.
The questionnaire utilized a multiple-item method, resulting in an average score for each construct This average score was subsequently applied in further analyses, including correlation and multiple regression analysis (Park & Kim, 2003).
In order to investigate the relationship between two variables, Pearson correlation coefficient (r) was used to measure the strength of the relationship (Sweet & Grace-Martin,
2008) The correlation coefficient ranges from -1 to +1 The correlation of 0.30 is considered a “good” correlation, and a correlation above 0.40 is considered “strong”
(Garson, 2011; Sweet & Grace-Martin, 2008) However, Garson (2011) recommends that the correlation coefficient should not exceed 0.80 in order to avoid multicollinearity
Multiple regression analysis is a valuable tool in both social and natural science research for examining the relationship between independent and dependent variables This method allows researchers to identify a set of independent variables that significantly explain the variance in a dependent variable, as determined by the significance test of R-square Furthermore, multiple regression analysis offers insights into the strength and nature of these relationships, making it an essential technique for data-driven investigations.
The "net strength" of each independent variable's relationship to the dependent variable can be assessed by analyzing the weight of beta coefficients This allows researchers to evaluate the impact of each independent variable on the dependent variable effectively (Garson, 2011).
Before performing multiple regression analysis, the following assumptions had been tested and qualified to provide reliable results (Garson, 2011):
(1) Normality The normality of the residuals can be diagnosed by applying a histogram of residuals, as well as using normal probability plot
(2) Linearity If the relationship is linear, the plot of residuals does not form any curve (“n” shape or “u” shape) (Hair, Black, Babin, Anderson, & Tatham, 2006)
The independence of the error term is assessed using the Durbin-Watson statistic, which tests for autocorrelation (Garson, 2011) According to Norusis (2005), a Durbin-Watson statistic close to 2 suggests no correlation between error terms Additionally, Garson (2011) notes that a Durbin-Watson coefficient falling between 1.5 and 2.5 indicates the independence of error terms.
(4) Absence of multicollinearity The simplest way of assessing multicollinearity is through the examination of the correlation matrix for the independent variables If the
A Pearson correlation coefficient greater than 0.80 indicates significant multicollinearity among variables To assess multicollinearity, researchers also utilize the tolerance statistic and the variance inflation factor (VIF) Hair et al (2006) suggest that a tolerance value below 0.10 signals potential multicollinearity issues, while Garson (2011) recommends that VIF values should exceed 1.0 to mitigate these problems.
(5) Absence of heteroscedasticity In this study, the author uses scatterplot diagram, which shows the variation in error, variance of the residuals
(6) Absence of outlier and influential observations The author use Normal P-P Plot of Regression Standardized Residual to identify this assumption
To establish a moderation effect involving a third variable (M) on the relationship between two variables (X and Y), it is essential to demonstrate that this relationship varies with changes in the moderating variable (Sharma, Durand, & Gur-Arie, 1981) Utilizing multiple regression analysis is advantageous due to its versatility in handling categorical variables (Cohen, Cohen, West, & Aiken, 2003) Following specific procedural steps will enhance the clarity and accuracy of the analysis.
1 Standardize all variables to make interpretations easier afterwards and to avoid multicolliearity, dummy code categorical variables and manually create product terms for the predictor and moderator variables
2 Fit a regression model (block 1) predicting the criterion from the predictors and the moderator Both effects and the model in general (R 2 ) should be significant
3 Add the interaction effect to the previous model (block 2) and check for a significant R 2 change as well as a significant effect by the new interaction term
If both are significant, then moderation is occurring
When analyzing regression results, it is crucial to focus on the unstandardized (B) coefficients instead of the standardized (β) coefficients This is particularly important in models that include interaction terms, as the β coefficients for these terms are not accurately standardized and therefore lack interpretability (Cohen et al., 2003; Aiken & West, 1991).
Chapter summary
This chapter outlines the primary research methodology employed in the study, detailing the research design and the selected data collection and analysis techniques A pretest survey was administered to academic personnel, research experts, colleagues, and friends, utilizing a questionnaire based on previous studies Data was gathered through both online and face-to-face questionnaires For data analysis, exploratory factor analysis, reliability testing using Cronbach’s alpha coefficient, and multiple regression with SPSS were planned The findings from the field research and the analysis results will be presented in the following chapter.
Research Results
Introduction
This chapter presents the analysis of survey data using previously outlined methodologies It begins with a description of the respondents' demographic profiles, followed by an examination of the validity and reliability of the constructs through factor analysis and reliability analysis Finally, multiple linear regression analysis was employed to test the developed hypotheses.
Profile of Respondents
The survey participants were Internet users aged between 16 and 45 including online shopping visitors (non-shoppers) and purchasers (shoppers) in Ho Chi Minh City A total of
A total of 242 online and 173 face-to-face questionnaires were collected, but 96 (68 online and 28 face-to-face) were deemed unusable due to incompleteness or invalidity, likely indicating a lack of respondent engagement Additionally, responses from participants who were not online shopping visitors or purchasers in the relevant product categories were also considered invalid Consequently, only 319 usable questionnaires, representing 76.87 percent of the total, were analyzed using SPSS software version 20.
Table 4.1 presents the demographic profile of the surveyed respondents, including key details such as gender, age group, marital status, education, monthly income, occupation, and internet experience, with further information available in Appendix C.
The survey respondents were comprised of 50.8% males and 49.2% females, with a significant 77.8% aged between 21 and 30 years Notably, 75.2% identified as single, and 66.8% held a bachelor's degree or professional qualification A majority reported an income of less than VND 5,000,000 (30.4%), while 27.0% earned between VND 5,000,000 and 10,000,000, reflecting the presence of university students among the respondents Additionally, over half (53.6%) were employed in IT-related occupations, and the data revealed that 63.9% of respondents had over six years of internet usage experience, with 27.3% having 4 to 6 years of experience.
Source: Analysis of field data
According to the data presented in Table 4.2, a significant majority of respondents (62.7%) reported spending over four hours daily on the Internet Notably, only 17.9% of participants indicated they had never made an online purchase, suggesting a high adoption rate of 82.1% for online shopping Among those with online shopping experience, 53.0% had made purchases 1 to 2 times, while 28.2% shopped more than five times, and 18.8% had purchased between three to five times The most popular categories for online purchases included clothing (38.9%) and electronics.
(33.2%) were the two most common items purchased or visited by the respondents
Hours spent on the Internet (daily)
Frequency of purchase in last 06 months
Kind of product visiting or purchasing
Source: Analysis of field data
Descriptive statistics
The collected data was transformed for easier interpretation by the researcher, with means and standard deviations for each model variable reported in Appendix D Utilizing a five-point Likert scale, where a score of 5 indicated strong agreement and 1 indicated strong disagreement, the means for nearly all variables (23 items) exceeded the neutral midpoint of 2.5 This indicated a strong consensus among respondents regarding the survey statements The subsequent section provides a detailed discussion of the descriptive statistics for the items measuring each construct.
Perceived usefulness (PU) There were 5 measurement items for this construct
Respondents rated the ability to search and purchase items quickly (PU1), the usefulness and advantages of online shopping (PU5), and the ease of obtaining goods (PU4) highly, with means exceeding 3.6 In contrast, the factors related to saving money (PU2) and making better purchase decisions (PU3) received lower ratings compared to other items Nonetheless, all items scored above the neutral mean, indicating general agreement among respondents Summary statistics, including means and standard deviations for perceived usefulness, are detailed in Table 4.3, organized from highest to lowest mean.
Table 4.3: Means and standard deviations of items measuring perceived usefulness
No Perceived usefulness (PU) Mean Std Deviation
PU1 Use online shopping enables me to search and buy something more quickly 4.03 0.774
PU5 Overall, I believe “online shopping” is useful and advantageous 3.61 0.804
PU4 Use online shopping makes it easier for me to obtain goods 3.61 0.853
PU2 Use online shopping enables me to save my money 3.18 0.959
PU3 Use online shopping helps me to make better purchase decisions (efficiency and effectiveness) 3.14 1.007
Source: Analysis of field data
Perceived ease of use (PEOU) There were 4 measurement items for this construct
Respondents expressed a strong agreement regarding the ease of making online transactions without expert assistance (PEOU2), with a mean score exceeding 4.2 Additionally, the simplicity of learning to use online shopping for the first time (PEOU1) and the ability to become proficient in it (PEOU4) received high ratings of 3.93 and 3.87, respectively Conversely, the aspect of online shopping that does not require significant mental effort (PEOU3) received the lowest mean score at 3.62 Despite this, all items scored above the neutral mean, indicating overall agreement among respondents Detailed summary statistics, including means and standard deviations, can be found in Table 4.4.
Table 4.4: Means and standard deviations of items measuring perceived ease of use
No Perceived ease of use (PEOU) Mean Std Deviation
PEOU2 It will be POSSIBLE to make an online transaction without expert help 4.23 0.743
PEOU1 It is easy for me to learn how to use online shopping, even at the first time 3.93 0.861
PEOU4 It is easy for me to become skillful at using online shopping 3.87 0.860
PEOU3 Use online shopping do not requires a lot of mental effort 3.62 0.979
Source: Analysis of field data
Perceived risk (PR) was assessed through three measurement items, with privacy risk concerns (PR2) receiving high ratings from respondents, averaging over 3.6 Additionally, participants expressed agreement regarding the risks associated with receiving products that do not meet expectations during online shopping (PR3) and the perceived dangers of credit card payments (PR1), both averaging above 3.4 Importantly, all measurement items scored above the neutral mean, indicating a general awareness of perceived risks in online transactions Summary statistics, including means and standard deviations for the perceived risk items, are detailed in Table 4.5.
Table 4.5: Means and standard deviations of items measuring perceived risk
No Perceived risk (PR) Mean Std Deviation
PR2 I think online shopping would put my privacy at risk 3.69 0.911
PR3 I am worried that product quality may not meet my expectations 3.55 0.943
PR1 I think paying with a credit card to shop online is risky 3.48 0.967
Source: Analysis of field data
Brand orientation (BO) was assessed through three key variables, with respondents rating the items highly The statements included in BO3 emphasized loyalty to preferred brands found through online shopping, while BO2 highlighted the importance of purchasing from well-known brand names, and BO1 indicated familiarity with web retailers as a significant factor Summary statistics, including means and standard deviations for the brand orientation items, are detailed in Table 4.6.
Table 4.6: Means and standard deviations of items measuring brand orientation
No Brand orientation (BO) Mean Std Deviation
BO3 Once I find a brand I like through web-shopping, I stick with it 3.95 0.840
BO2 It is important for me to buy products/services from the web retailer with well-known brand names 3.73 1.004
BO1 If I buy products/services from a web-retailer that I am familiar with, I would prefer to buy well—known brand name
Source: Analysis of field data
Prior online purchase experience (POPE) significantly influences consumers' online shopping decisions, with three key measurement items identified Respondents expressed confidence in recognizing reliable retailer websites (POPE3) and reported improved skills in online shopping (POPE2) However, experience with the retailer's website (POPE1) received the lowest average rating of 3.73, although all measurement items maintained scores above the neutral mean Detailed summary statistics, including means and standard deviations for these variables, can be found in Table 4.7.
Table 4.7: Means and standard deviations of items measuring prior online purchase experience
No Prior online purchase experience (POPE) Mean Std Deviation
POPE3 After doing an online shopping, I know reliable retailer's web sites
POPE2 After doing an online shopping, I become more skillful 3.90 0.761
POPE1 I am experienced with the use of the retailer's web site 3.73 0.754
Source: Analysis of field data
The study on online purchase intention (OPI) revealed that respondents rated their intention to shop online highly, with scores exceeding 3.8 for several items Participants expressed a strong preference for online shopping over traditional methods (OPI3) and indicated a commitment to continue using it (OPI2) and plan to use it in the future (OPI1) Additionally, they showed a positive inclination towards recommending online shopping to others (OPI4) and frequently using it themselves (OPI5), with mean scores of 3.63 and 3.47, respectively.
Obviously, all of the measurement items had means higher than the neutral mean Table 4.8 reported the summary statistics of means and standard deviations of these variables as below
Table 4.8: Means and standard deviations of items measuring online purchase intention
No Online purchase intention (OPI) Mean Std Deviation
OPI3 I think it would be very good to purchase products through Internet in addition to traditional methods
OPI2 I will keep using online shopping in the future 3.85 808
OPI1 I intend to shop online in the future 3.84 795
OPI4 I would strongly recommend others to use online shopping 3.63 848
OPI5 I would frequently use “online shopping” 3.47 903
Source: Analysis of field data
Factor analysis
The purpose of the factor analysis was to identify the underlying structure among the analyzed variables (Hair et al., 2006) To evaluate the validity of the constructs within the questionnaires, a total of 23 items were examined using Principal Components analysis.
The extraction process utilized Varimax rotation, as detailed in Appendix E Table 4.9 presents the Kaiser-Meyer-Olkin (KMO) measure of sampling adequacy and Bartlett's test of sphericity for both independent and dependent variables The KMO values were 0.809 for independent variables and 0.819 for the dependent variable, both exceeding the threshold of 0.5 Additionally, Bartlett’s test yielded a p-value of less than 0.001, indicating high significance These results demonstrate that the relationships among the variables are statistically significant, making them suitable for exploratory factor analysis to identify a parsimonious set of factors (Tabachnick & Fidell, 2007).
Table 4.9: KMO and Bartlett's Test
Kaiser-Meyer-Olkin Measure of Sampling Adequacy 0.809 0.819
Bartlett's Test of Sphericity Approx Chi-Square 1680.179 1037.999 df 153 10
Source: Analysis of field data
The Rotated Component Matrix, presented in Table 4.10, displays the factor loadings for each item, sorted by size The results indicate that all 18 items are appropriately aligned with the construct, grouped into five distinct components, ranging from Factor 1 to Factor 5 Notably, each item has a loading value exceeding 0.5 and is associated with only one factor following the rotation.
The analysis revealed three key factors influencing user perceptions: the first factor, labeled Perceived Usefulness (PU), included five items related to the perceived benefits of the product The second factor, named Perceived Ease of Use (PEOU), encompassed items reflecting the ease with which users can interact with the product Lastly, the third factor, Prior Online Purchase Experience, consisted of three items that captured users' previous experiences with online shopping.
The analysis identified three key factors influencing consumer behavior: Brand Orientation (BO), which encompasses the impact of brand names, and Perceived Risk (PR), which relates to the associated risks of purchasing decisions Each factor consists of closely related items that collectively represent broader evaluative dimensions in consumer evaluation (Hair et al., 2006).
The analysis revealed a strong correlation between each item and its respective factor, with the "usefulness" construct clearly aligning under Factor 1 (Perceived Usefulness) and showing no correlation with Factors 2 to 5, thereby confirming the validity of the scale items Additionally, as shown in Table 4.11, all factors exhibited eigenvalues greater than 1 (4.633, 1.939, 1.731, 1.589, and 1.119), indicating their significance, with further details available in Appendix E, section E.1.
PU5 Overall, I believe “online shopping” is useful and advantageous .730
Use online shopping helps me to make better purchase decisions (efficiency and effectiveness)
PU2 Use online shopping enables me to save my money .705
PU4 Use online shopping makes it easier for me to obtain goods .655
PU1 Use online shopping enables me to search and buy something more quickly .605
PEOU3 Use online shopping do not requires a lot of mental effort .811
PEOU1 It is easy for me to learn how to use online shopping, even as the first time .687
PEOU4 It is easy for me to become skillful at using online shopping .685
PEOU2 It will be IMPOSSIBLE to make an online transaction without expert help .561
POPE1 I am experienced with the use of the retailer's web site .830
POPE3 After doing an online shopping, I know reliable retailer's web sites .312 752
POPE2 After doing an online shopping, I become more skillful .396 706
It is important for me to buy products/services from the web retailer with well-known brand names
If I buy products/services from a web-retailer that I am familiar with, I would prefer to buy well—known brand name
BO3 Once I find a brand I like through web- shopping, I stick with it .719
PR2 I think online shopping would put my privacy at risk .845
PR1 I think paying with a credit card to shop online is risky .813
PR3 I am worried that product quality may not meet my expectations .692
Extraction Method: Principal Component Analysis
Rotation Method: Varimax with Kaiser Normalization a Rotation converged in 7 iterations
Source: Analysis of field data
The analysis revealed that each factor explained more variance than individual items, with the first factor accounting for 14.648%, the second for 12.989%, and the third for 11.865% Collectively, the five components explained 61.171% of the variance, indicating that 61.171% of the total variance could be attributed to the 18 items represented by these factors.
Component Initial Eigenvalues Rotation Sums of Squared Loadings
Total % of Variance Cumulative % Total % of Variance Cumulative %
Source: Analysis of field data
The analysis indicates that each item set is unidimensional, as the items are closely related and collectively represent a single concept under one factor Furthermore, the high loading of each scale item on this singular factor confirms the construct validity (Hair et al., 2006).
Reliability Test
The reliability of the scale was assessed using Cronbach’s alpha coefficient, with the reliability analysis for each construct—Perceived Usefulness (PU), Perceived Ease of Use (PEOU), Perceived Risk (PR), Behavioral Intention (BO), Perceived Online Purchase Experience (POPE), and Online Purchase Intention (OPI)—detailed in Table 4.12 For comprehensive results, please refer to Appendix F.
As observed, the Cronbach’s alpha coefficient for almost constructs, except
Perceived Risk ranged from 0.715 (Perceived Ease of Use) to 0.890 (Online Purchase
The intention scores exceeded 0.7, indicating strong reliability, while the Perceived Risk value of 0.692 was also deemed acceptable These findings suggest that each scale item effectively measures the same construct, demonstrating high reliability and good internal consistency (Field, 2005; Sweet & Grace-Martin, 2008).
No Items Cronbach's Alpha Cronbach's Alpha Based on Standardized Items
Source: Analysis of field data
The values under the column “corrected item – total correlation” pointed out the correlations between each item and the rest of the values from every construct (Field, 2005)
For effective correlation among items, a score exceeding 0.3 is essential As shown in Appendix F, all items for each construct achieved corrected item-total correlation scores ranging from over 0.3 to under 0.8, with the exception of item OPI3 ("I will keep using online shopping in the future"), which scored 0.838 Consequently, the construct reliability for this dataset is confirmed.
The study identified perceived risk, perceived ease of use, perceived usefulness, brand orientation, and prior online purchase experience as key factors influencing consumers' behavioral intention to shop online Consequently, the research model and hypotheses outlined in the Literature Review were retained for further analysis.
To perform further analyses such as: correlation and multiple regression, proxy variables were computed by getting average of related items Accordingly, Perceived
The Usefulness (PU) variable is calculated as the average of PU1 to PU4, while the Perceived Ease of Use (PEOU) variable is derived from the mean of PEOU1 to PEOU4 Additionally, the Perceived Risk (PR) variable is determined by averaging PR1 to PR3, and the Band Orientation (BO) variable is based on the mean of BO1 to BO3 Finally, the analysis incorporates the influence of Prior Online Purchase behavior.
Experience (POPE) variable was mean of POPE1 to POPE3; and Online Purchase Intention (OPI) variable was mean of OPI1 to OPI5.
Correlation Analysis
Table 4.13 reveals that all variable pairs were significant at the 0.01 level, with the exception of the relationship between Perceived Risk and the other independent variables Nevertheless, nearly all proposed hypotheses demonstrated significant relationships.
Chapter 2 were found to be statistically significance at level 0.01, except perceived risk correlated significantly with online purchase intention at level 0.05 It meant that the PU (r
The study found significant positive correlations between various factors and Online Purchase Intention (OPI), with perceived usefulness (PU) showing the strongest correlation at r = 0.638 Other positive correlations included perceived ease of use (PEOU) at r = 0.421, perceived online privacy expectations (POPE) at r = 0.596, and brand orientation (BO) at r = 0.270 Conversely, perceived risk (PR) demonstrated a significant negative correlation with OPI at r = -0.127 These findings highlight the varying degrees of influence that these factors have on consumers' intentions to make purchases online.
PU PEOU PR BO POPE OPI
** Correlation is significant at the 0.01 level (2-tailed)
* Correlation is significant at the 0.05 level (2-tailed)
Source: Analysis of field data
Multiple Regression
Before conducting regression analysis, it is essential to address six key assumptions: normality, linearity, independence of the error term, absence of multicollinearity, absence of heteroscedasticity, and the exclusion of outliers and influential observations (Field, 2005) Detailed results for these assumptions in the context of multiple regression can be found in Appendix G.
The histogram generated from SPSS demonstrated a unimodal and symmetric distribution, indicating normality Additionally, the scatterplot analysis confirmed that the assumptions of linearity and homoscedasticity were satisfied, as the residuals were randomly scattered without any discernible patterns when plotted against the predicted values.
The analysis confirmed the independence of the error term, with a Durbin-Watson value of 2.043, indicating good independence Multicollinearity was assessed through VIF values, which were all below 10 and above 1.0, while tolerance statistics exceeded 0.2, demonstrating no multicollinearity among the independent variables, as the correlation coefficients remained below 0.80 Additionally, the normal p-p plot analysis revealed no outliers or influential observations, as the normal probability plot appeared as a straight line and the residuals displayed a uniform spread against the predicted values.
In Chapter 2, the proposed hypotheses were evaluated through multiple linear regression analysis, which allowed for an examination of the impact and significance of independent variables on their relationship with the dependent variables.
The R-Square value of approximately 0.544 indicates that 54.4% of the variability in Online Purchase Intention (OPI) can be explained by the combined influence of the independent variables: Perceived Usefulness (PU), Perceived Ease of Use (PEOU), Perceived Risk (PR), Brand Orientation (BO), and Prior Online Purchase Experience (POPE) This relatively high value demonstrates a strong linear relationship between these predictors and the dependent variable.
R R Square Adjusted R Square Std Error of the Estimate Durbin-Watson
537 a 544 536 47077 2.043 a Dependent Variable: Online Purchase Intention
Source: Analysis of field data
According to Field (2005), hypotheses are supported when the beta coefficients are significant, indicated by a p-value of less than 0.05 and a t-value greater than 2.0 A smaller p-value and a larger t-value enhance the predictive power of the independent variable Furthermore, the standardized beta coefficient (β) reveals the nature of the relationship—whether positive or negative—between the dependent and independent variables.
The ANOVA table from the multiple regression analysis (Table 4.15) indicates a p-value of less than 0.001, significantly lower than the 0.05 threshold This suggests that at least one of the five independent variables effectively contributes to modeling the dependent variable, Online Purchase Intention.
Squares df Mean Square F Sig
Total 152.026 318 a Dependent Variable: OPI Online Purchase Intention b Predictors: (Constant), POPE Prior Online Purchase Experience, PR Perceived Risk,
BO Brand Orientation, PEOU Perceived Ease of Use, PU Perceived Usefulness
Source: Analysis of field data
Table 4.16 displays the results of the multiple regression analysis conducted for the research model outlined in Chapter 2 The findings support Hypothesis H1, indicating a positive relationship between Perceived Usefulness and customers' intention to make online purchases in Ho Chi Minh City.
Perceived Usefulness and Online Purchase Intention (β = 0.427 > 0, t = 9.665> 2, p < 0.001) was statistically significant This hypothesis had the most significant relationship among the others Therefore, the hypothesis H1 was supported
Hypothesis H2 suggested a positive relationship between Perceived Ease of Use and customers' online purchase intention The findings, as shown in Table 4.16, confirmed this relationship with significant results (β = 0.105, t = 2.392, p = 0.017) Therefore, the study supports Hypothesis H2.
Table 4.16: Result of Multiple Regression Analysis a
Standardized Coefficients t Sig Collinearity Statistics
B Std Error Beta Tolerance VIF
Source: Analysis of field data
Hypothesis H3 examined the relationship between consumers' risk perception and their intention to shop online in Ho Chi Minh City Initially, the perceived risk (PR) showed a weak statistical association with online purchase intention (OPI) in bivariate analysis; however, it became significant in multivariate analysis when other independent variables were considered The results indicated a p-value of 0.014, confirming a statistically significant relationship at a 98.6% confidence level, demonstrating that perceived risk influences customers' online shopping intentions Additionally, the β value of -0.094 (t = -2.464) indicated a negative relationship, leading to the acceptance of hypothesis H3.
Hypothesis H4 posited that increased brand orientation correlates with higher online purchase intentions among consumers The findings revealed a statistically significant positive relationship (β = 0.106, t = 2.700, p = 0.007), indicating that brand orientation serves as a significant independent variable in predicting online shopping intentions, thus supporting hypothesis H4.
Hypothesis H5 investigated the connection between consumers' online purchase experiences and their intention to shop online The results, as shown in Table 4.16, revealed a statistically significant relationship with a p-value of less than 0.05 (p < 0.001) and a t-value of 6.560, exceeding the acceptable threshold Additionally, the β value of -0.342 indicated a positive relationship, confirming that hypothesis H5 was supported in this study.
To examine the moderating effect of the product category variable, a hierarchical moderator regression analysis (HMRA) was conducted following the methodology outlined by Sharma et al (1981) For a comprehensive overview of the moderating test, please refer to Appendix H The analysis involved a series of specific steps to ensure accurate results.
Step 1: Center the continuous predictor variables (Centered Perceived Usefulness (CPU), Centered Perceived Ease of Use (CPEOU), Centered Perceived Risk (CPR),
Centered Brand Orientation (CBO), and Centered Prior Online Purchase Experience
(CPOPE)) and dummy code moderator variable (DCLO, DELECS) to eliminate multicollinearity effects between the predictor and moderator, and the interaction terms
Step 2: Enter the predictors and moderators main effects (hierarchical)
Step 3: Enter the interactions between predictors and moderators
The analysis presented in Table 4.17 indicates that the product categories DCLO and DELECS do not exhibit a significant correlation with either the independent or dependent variables at a 0.05 significance level This finding suggests that the product categories DCLO and DELECS are not significantly related to the dependent variable or its predictors.
Table 4.17: Correlation between Moderator, Dependent and Independent Variables
DCLO DELECS CPU CPEOU CPR CBO CPOPE OPI
** Correlation is significant at the 0.01 level (2-tailed)
* Correlation is significant at the 0.01 level (2-tailed)
Source: Analysis of field data
The change in R-squared was used to assess the predictive power gained by adding another variable in step 2 However, Model 2, which included interactions among perceived usefulness, perceived risk, perceived ease of use, brand orientation, prior online purchase experience, and product category, showed insignificant results This indicates that there is no significant moderation effect between the predictors and product category on consumers' online purchase intentions.
Table 4.18: Model Summary c with Moderator
Std Error of the Estimate
R Square Change F Change df1 df2 Sig F Change
2 756 b 571 547 46549 021 1.473 10 301 149 a Predictors: (Constant), DELECS, CPEOU, CPR, CBO, CPU, DCLO, POPE; b Predictors: (Constant), DELECS, CPEOU, CPR, CBO, CPU, DCLO, CPOPE, DELECS_CPR, DCLO_CBO, DELECS_CPU, DCLO_CPEOU, DCLO_CPR,
DELECS_CBO, DCLO_CPOPE, DCLO_CPU, DELECS_CPEOU, DELECS_CPOPE; c Dependent Variable: OPI
Source: Analysis of field data
Chapter summary
This chapter details the data analysis conducted for the research, which achieved a response rate of 76.87 percent through web-based and on-the-spot surveys Initially, descriptive statistics were presented, followed by factor and reliability analyses to assess the validity and reliability of the data Correlation analysis was then employed to examine the relationships between the proposed hypotheses Finally, multiple regression analysis was performed after verifying six assumptions to ensure the reliability of the results.
The multiple regression analysis tested six hypotheses derived from the literature review, revealing that perceived usefulness, perceived ease of use, perceived risk, brand orientation, and prior online purchase experience significantly influence consumers' online purchase intentions Additionally, the product category was found to moderate these relationships The subsequent chapter will discuss the findings' implications, while the research's limitations and recommendations for future studies will be presented at the conclusion.
