Generation Z’s Behavioral Factors Affecting Purchase Intention Through Social Media Portals: A Hybrid PLS-SEM and ANN Modeling Approach

Abstract

The objective of this study is to ascertain the behavioral factors that influence the online purchase intention (PI) of Generation Z. For this, it extended the theory of planned behavior (TPB) by incorporating five new concepts. We collected data using non-probability convenience and purposive sampling approaches. We surveyed 551 from Generation Z using a standardized questionnaire. The study adopted a hybrid approach that combines SEM with artificial neural network (ANN) analysis to investigate real-world information. The outcomes demonstrated that attitude (PA), subjective norms (SN), behavioral control (PBC), live streaming (LS), celebrity endorsement (CE), promotional tools (PT), and online reviews (OR) significantly correlated with Gen-Z’s online PI employing SMPs. The ANN results revealed that CE had the most significant impact on the online PI of Gen-Z’s students with 97.5% normalized importance, followed by OR (74.4%), LS (57.1%), PA (54.7%), SN (53.9%), PBC (49.3%), and PT (46.5%). In addition, personal faith significantly reinforces the positive relationship between PA and PI and LS and PI. The outcomes offer several theoretical and practical implications, such as the interconnectedness of the constructs and the understanding of the TPB constructs of human behavior. Online businesses must encourage Gen-Z to increase their positive attitudes toward online purchasing. Enterprises may offer relevant information, precise pricing details, and efficient delivery services, which can effectively promote Gen-Z’s attitude to participate in online shopping. The study uniquely considers Gen-Z’s purchase intention (PI) through SMPs by applying expanded TPB constructs and adopting a hybrid approach that combines SEM with ANN analysis.

Keywords

Generation Z theory of planned behavior behavioral factors social media portals purchase intention PLS-SEM ANN

Introduction

Online shopping through social media portals (SMPs) is increasing enormously. Due to the COVID-19 pandemic, most businesses have been influenced, leading them to adopt various initiatives to enhance their internal effectiveness and performance (Donthu & Gustafsson, 2020; Sohrabi et al., 2020). Consequently, organizations have transitioned their diverse business operations to Internet platforms. In such circumstances, organizations perceive SMPs as the most convenient and prominent means of online marketing and facilitate customers to purchase products from virtual businesses rather than traditional physical stores. Well-known SMPs, like Facebook, Twitter, and Google, play a significant role in facilitating a wide range of user engagement operations, such as chatting, messaging, blogging, and gaming. Likewise, consumers spend much time on SMPs, creating and sharing content. As a result, organizations aware of these factors employ various strategies to involve customers, improve their brand visibility, and take advantage of the potential benefits of using SMPs. Hence, companies must implement strategic initiatives that align with consumer frameworks and brand principles to enhance social brand recognition and encourage consumers to purchase their products. Properly utilizing digital and social media advertising will enable organizations to achieve their marketing objectives at a very affordable expense (Ajina, 2019; Yadav, 2016).

Behavioral psychology has a wide-ranging impact beyond shaping an individual’s character and directly affects their daily behaviors (S. S. Hossain & Khan, 2021; Khan et al., 2024). Behavioral psychology plays a crucial role in understanding consumer behavior. The emergence of Generation Z (Gen-Z) has recently led to a heightened focus on behavioral psychology and its implications for consumer intentions, particularly concerning product purchases. As one of the largest generations, Gen-Z individuals can change existing social attitudes and are highly knowledgeable and updated in technology (Bejan, 2023). In addition, they value their behavioral elements by actively participating in social media platforms, which demonstrates their behavioral intention. The growing ubiquity of mobile phones and the extensive embrace of wireless technologies have led to innovative business trends (Khan & Roy, 2023). Due to busy lifestyles, customers experience time constraints, which prevent them from being able to participate in the process of purchasing necessary goods and services physically (Khan & Sarkar, 2024). Hence, Gen-Z prefers to procure a wide range of goods or services through internet shopping. They are becoming more familiar with online purchasing and its accompanying benefits. A particular subset of customers places greater importance on having access to superior options concerning information, advantages, and pricing alternatives (T. Pervin & Khan, 2022). Based on the above discussion, the following research questions may arise,

RQ1. Which behavioral factors are most effective for Gen-Z’s online purchasing intention?

RQ2. Which behavioral factors may directly or indirectly influence Gen-Z’s online purchasing intention from an emerging market?

RQ3. How do the behavioral factors affect Gen-Z’s online purchasing intention?

Therefore, this study aims to comprehend Gen-Z’s behavioral drivers behind online purchasing intents. After conducting a comprehensive analysis of the relevant literature on the effects of social media platforms (SMPs) on online purchasing, it is clear that many researchers have attempted to assess the impact of social media, including celebrity endorsements and online reviews (McCormick, 2016; Weismueller et al., 2020), on various aspects of consumer behavior, including purchasing decisions, customer satisfaction, and online shopping behavior (Armutcu et al., 2024; Pop et al., 2022; P. Verma et al., 2019). In addition, earlier researchers focused their investigations on examining the impact of SMPs on buy intention or decision, satisfaction, and loyalty (Fu et al., 2020; Gupta et al., 2020; A. Hossain et al., 2020; Jenefa, 2017). However, there was a lack of emphasis on the Generations basis rather than particular buyer groups (garment products, service, university students, etc.), and thus, this study aims to investigate Gen-Z’s behavioral factors affecting purchase intention (PI) through SMPs. In addition, to our knowledge, no research has adopted a hybrid approach combining SEM with ANN analysis on Gen-Z’s expanded TPB factors affecting purchase intention through social media portals.

The study employs the expanded TPB to reach the objectives. Multiple scholars have confirmed the usefulness and resilience of the TPB model in forecasting online PIs, and they have enhanced this model by suggesting further exogenous constructs (Roy, 2023b; X. Wang et al., 2019). Data was collected using non-probability convenience and purposive sampling approaches, using a standardized questionnaire to administer 551 responses from the Generation Z samples. The study adopted a hybrid approach that combines SEM with ANN analysis to investigate real-world information. The next section explains the relevant literature extensively and shows the evidence for developing hypotheses. In the third section, the study thoroughly explains the methodology of this research. After that, in the fourth section, the study highlights the data analysis and findings of the research, and the following section thoroughly discusses the outcomes and points out the uniqueness of the research by comparing earlier related research findings. Finally, the study continues with several research implications and ends with a conclusion, limitations, and future research directions.

Literature Review and Hypotheses Development

Theory of Planned Behavior (TPB)

The current study utilizes the well-established TPB model as the foundational framework to clarify the intention to make online purchases (Ajzen, 1991). The theory is pragmatic, based on robust principles, and reliably forecasts various expected behaviors. Additionally, it is effective for ascertaining and predicting consumer purchasing intentions. Given its wide-ranging applicability, it is a valuable framework for assessing consumer preferences on online purchasing platforms (Ha et al., 2019; Hanaysha, 2022; Roy, 2023e). The model comprises four fundamental elements: behavioral control, intention, attitude, and subjective norms. In the literature, intention is defined as an individual’s deliberate strategy, dedication, or determination to engage in an activity or achieve a particular goal (Harland et al., 1999). According to Harland et al. (1999), behavior is directly influenced. The primary determinant of a consumer’s future behavior is their intention to engage in a particular behavior. Ajzen (1991) elucidated in his TPB that intention can effectively predict behavior when behavior is within one’s control. Thus, behavior can be quantified with a clear intention to take action. Individuals can form and ultimately possess an intention if they are motivated by particular activities. The Theory of Reasoned Action (TRA), proposed by Fishbein and Ajzen (1977), serves as the foundation for the TPB model. The application of TRA is expected in the fields of psychology and sociology for predicting behavior (Waris et al., 2023). Both the TRA and TPB are employed to forecast consumer behavior. However, TPB is considered the superior theory for predicting customer behavior (Taylor & Todd, 1995), and it is widely utilized for this purpose (Knowles et al., 2012).

Gen Z, TPB, and Extended Behavioral Factors

Generation Z has encountered unique situations throughout their upbringing compared to prior generations. Although Gen-Z members have recently joined the workforce, they have already formed opinions and developed distinctive characteristics. The Gen-Z cohort is the most contemporary, culturally diverse, and populous generation ever studied. They live in an age of continuous updates and can absorb information faster than previous generations. The Gen-Z cohort comprises those born from the mid-1990s to 2010 (Ameen et al., 2023; T. Pervin & Khan, 2022). Gen-Z is the largest generation cohort, consisting of more than two billion persons. They have experienced a more significant change over their lifetime than their predecessors. They are not simply representatives of the future but actively influence and create it. With a population share of 18% worldwide, they are expected to take on global leadership within 20 years (X. Chen et al., 2023). A group of elderly persons who have recently finished college are now entering the workforce and making their livelihood (Dadić et al., 2022).

The main feature of this age group is its growth during the Great Recession in the middle of the crises of extremism and climate change. It is classified as global, technological, esthetically pleasing, and sociable (Subawa et al., 2020). This generation is distinguished by their exceptional level of schooling, broad network of acquaintances, and sophisticated refinement. They maintain and support the principles of inclusivity, equity, and equal treatment in the community and the media. This current generation displays robust optimism and is profoundly driven by their dreams. Generation Z places a high importance on achieving their full potential, being rewarded for their work, and having a pleasant work atmosphere (Munsch, 2021). Gen-Z can be psychologically described by their propensity for abundant productivity, interconnection, and reliance on online interactions. Their psyche is driven by the pursuit of material prosperity, instant financial gains, instantaneous satisfaction, and a lifestyle focused on consumption. They inhabit a networked domain that allows them to exchange messages and converse globally with a simple click smoothly (Dadić et al., 2022). The extraordinary progress of technology has granted them global access. They can uncover and explore virtually any piece of information. Experiencing unexpected people, customs, actions, varied communities, viewpoints, adventures, and discoveries—all made readily available by the wealth of information and ease of technology (Ameen et al., 2023). This generation demonstrates a greater capacity for acceptance, brotherhood, transparency, and openness to variety than their predecessors. Utilizing the Internet offers a strong foundation for developing empathy for others, understanding a wide-ranging and open society, and spreading different perspectives (Salleh et al., 2017).

Earlier researches were also conducted on Gen Z’s intention on tourism, household food leftovers, clothes purchasing, etc. (Putri & Akbari, 2021; Setiawan et al., 2024; Siddiqui et al., 2022, 2023). This study rigorously examines the behavioral elements from TBP theories (attitude, subjective norms, and perceived behavioral control) that influence purchase intention using social media platforms (SMPs) to gain insight into the purchasing psychology of Generation Z. In addition, four more behavioral factors (live streaming, celebrity endorsement, promotional tools, and online review) were added as behavioral factors of SMPs purchase intentions (Kashif et al., 2018; Miah et al., 2022; Patel et al., 2023; Rehman et al., 2019; Yahya et al., 2019; Yin, 2020).

Personal Attitude (PA)

Attitude greatly influences a person’s impression and behavior. Thus, it greatly influences an individual’s behavior intention. PA is a person’s subjective assessment of a behavior’s repercussions based on past experiences. In the context of TRA and TPB, the term attitude pertains to the result shaped by specific ideas held by an individual on the positive or negative consequences of a particular activity (Ajzen, 1991). Attitude significantly impacts decision-making across all domains (Fazio et al., 2000). Several scholars argue that attitude significantly influences intention and should not be neglected as a factor (Kashif et al., 2018). Previous empirical research has shown a significant association between consumer perceptions, buying intentions, and shopping behavior. (Patel et al., 2023; Rehman et al., 2019). Thus, the current study proposes a consequent hypothesis:

H1: PA has a significant impact on Gen-Z’s online purchase intentions.

Subjective Norms (SNs)

External elements, such as familial, social, and relational pressures, can impact an individual’s behavior and decision to purchase (Ajzen, 1991). SN, an integral component of TPB theory, pertains to an individual’s opinion of whether their friends, family, and kin support or discourage the targeted action. In the context of e-commerce, SN describes how consumers feel pressure from society to buy more things from online retailers. When someone lacks knowledge regarding the sources and methods of obtaining specific products, they may solicit the perspectives of those in their immediate social circle, such as relatives and close friends. Subsequently, most individuals will be influenced by their opinions and choose to adhere to their guidance (Cheng & Yee, 2014). Again, research demonstrates that individuals are susceptible to the influence of those they perceive as significant, who may urge or persuade them to engage in online purchasing (Ru et al., 2021; Sin et al., 2012). Recent findings supported that SN is significantly associated with consumers’ online purchase intentions (Ha et al., 2019; Ru et al., 2021). Therefore, the present study puts forth the following hypothesis:

H2: SN has a significant impact on Gen-Z’s online purchase intentions.

Perceived Behavioral Control (PBC)

PBC stands for an individual’s high level of self-confidence in their capacity to effectively do a specific behavior. As Ajzen (1991) defined, PBC refers to the individual’s subjective assessment of the level of ease or difficulty associated with a certain action, considering their personal experiences and anticipated obstacles. It evaluates an individual’s capacity to surmount hurdles during performance (V. K. Verma & Chandra, 2018). Perceived behavioral control refers to an individual’s belief in their ability to control their actions when deciding whether or not to make a purchase. Their level of competence influences this belief(Francis et al., 2004). It is a factor that enhances an individual’s ability to search for relevant information actively. It is similar to the conditions that make it easier for technology to be accepted and used and for individuals to express their opinions. PBC depends on the individual having the requisite abilities, resources, and a feeling of authority over the decision-making process (Gao & Bai, 2014). Online purchasing is significantly influenced by PBC, which plays a pivotal role in shaping purchase intention (Kashif et al., 2018; Rehman et al., 2019). Thus, the current study proposes a subsequent hypothesis:

H3: PBC has a significant impact on Gen-Z’s online purchase intentions.

Live Streaming (LS)

LS business is a new kind of social trading and has recently gained increasing popularity (Adoeng et al., 2019; Taobangdan, 2019). In traditional forms of Internet purchasing, customers are limited to textual descriptions and visual representations while seeking information about products and services. In contrast, LS purchasing facilitates a synchronous and instantaneous connection between sellers and consumers. It utilizes one or more pieces of instruments to provide the instantaneous transmission of visual and auditory content from remote places, enabling viewers to remotely observe and examine the comprehensive attributes and quality of the products (C.-C. Chen & Lin, 2018; Li, 2019; Wongkitrungrueng & Assarut, 2020) As a result, buyers ought to have greater faith in the companies and their offerings. Therefore, LS encourages people to buy items and positively impacts online purchase behavior (Yin, 2020). Hence, the present study puts forth the following hypothesis:

H4: LS has a significant impact on Gen-Z’s online purchase intentions.

Celebrity Endorsement (CE)

Celebrities utilize social media networks to disseminate diverse information on personal lives, events, offerings, or products (Ashfaq & Ali, 2017). Some superstars maintain online networks with a significant number of followers. On the other hand, customers (especially youngsters) follow celebrities on social media and emulate their manner of living, including their choice of restaurants, holiday destinations, clothes, makeup, and fashion. Followers believe celebrity statements are genuine and easily persuaded to purchase online (Wilcox & Stephen, 2013). Existing research suggests that CEs contribute to the appeal, credibility, and compatibility between celebrities and products. These factors favorably impact consumers’ attitudes toward products and brands and their desire to purchase (Miah et al., 2022). Therefore, the present study puts forth the following hypothesis:

H5: CE has a significant impact on Gen-Z’s online purchase intentions.

Promotional Tools (PT)

Advertising on the Internet is a commonly employed tactic by firms to develop marketing plans and promotional ideas, as well as to study customer purchasing patterns. Dehkordi et al. (2012) argued that e-commerce and e-marketing demonstrate the relative ease of Internet marketing compared to traditional marketing methods. Online advertising encompasses several forms, such as contextual ads, banner ads, digital classified advertising, social network promotion, multimedia ads, and promotional emails resembling spam (Eyre et al., 2020). It delivers written details on particular services or products to users via SMPs. As a result, social media advertisements facilitate the practical acquisition of product information, conserving energy and time. Trust, confidentiality, and safety are crucial factors in social media networking platforms (Shamout, 2016; Siddique & Hossain, 2018; Yahya et al., 2019). According to the prior debate, PTs are projected to favor Gen-Z’s purchase intention. Therefore, the present study puts forth the following hypothesis:

H6: PT has a significant impact on Gen-Z’s online purchase intentions.

Online Review (OR)

Online review is a novel recommendation for purchasing products (Helm et al., 2010; Miah et al., 2022). A product review website evaluates consumers’ perspectives and opinions regarding service systems, product quality, and the general marketplace. ORs, specifically comments and ratings, shape consumers’ perception of a product as positive or negative. Therefore, ORs are crucial in helping consumers make decisions because positive ratings from one customer may influence a subsequent buyer to buy goods (Gan & Wang, 2015; Lim et al., 2016). Numerous studies have shown that significant ORs can impact consumer decision-making on SMPs (Miah et al., 2022; M. Zhang et al., 2020). Zhu and Zhang (2010) argued that a higher quantity of OR favorably influences potential customers’ perception of unfamiliar products. As a result, ORs that express sentiment get greater attention from consumers and have a beneficial impact on their purchasing decisions (Miah et al., 2022). Therefore, the present study puts forth the following hypothesis:

H7: OR has a significant impact on Gen-Z’s online purchase intentions.

Personal Faith (PF)

Faith is essential for firms to establish customer communication and relationships (Abror et al., 2020; Bhalla, 2020). Trust is crucial in traditional retail and online commerce (Cheng & Yee, 2014; Rehman et al., 2019). In online purchasing, faith refers to the readiness to accept the possibility of adverse conditions to engage in transactions with online sellers, believing that they will prioritize the consumer’s best interests (M. K. O. Lee & Turban, 2001). Moreover, it relates to how consumers perceive online retailers’ behavior, including their competence, impartiality, and integrity (McKnight et al., 2002; McKnight & Chervany, 2001). Recently, customers’ use of SMPs has enabled them to establish interpersonal connections and share information, leading to increased levels of trust (Pop et al., 2022). Yahia et al. (2018) emphasized the significance of trust in SMPs and the need to examine and understand faith as a variable in SMPs (Cooley & Parks-Yancy, 2019). According to Hansen et al. (2018), SMPs impact business trust. Conversely, trust is a significant factor in consumers’ decisions about what to purchase (Sanny et al., 2020). Previous studies have demonstrated that trust is essential in determining the intention to buy goods online (Ha et al., 2019; Hanaysha, 2022; Jadil et al., 2022).

However, the TBP theory fails to consider several aspects significantly influencing a consumer’s online purchase decision. This framework fails to account for the influence of perceived trust, risks, customer satisfaction, commitment, consumer experience, personal variables, psychological factors, and economic issues (Rehman et al. (2019). An individual may initially intend to purchase, but their intention might be influenced by factors such as their aversion to risk and the perceived risks associated with online buying, which may change their decision (Sahi et al., 2016). Therefore, it is necessary to provide an additional variable that may enhance these relationships. In this study, the researcher is utilizing personal faith (PA) as a moderating variable to examine if it moderates the association between various factors and the online purchase intention of Gen-Z (Kiani et al., 2016; Teah et al., 2014). Therefore, the present study puts forth the following hypotheses:

H8: PF moderates the association between PA and Gen-Z’s online purchase intentions.

H9: PF moderates the association between SN and Gen-Z’s online purchase intentions.

H10: PF moderates the association between PBC and Gen-Z’s online purchase intentions.

H11: PF moderates the association between LS and Gen-Z’s online purchase intentions.

H12: PF moderates the association between CE and Gen-Z’s online purchase intentions.

H13: PF moderates the association between PT and Gen-Z’s online purchase intentions.

H14: PF moderates the association between OR and Gen-Z’s online purchase intentions.

Research Methodology

Research Model

The study aims to ascertain Gen-Z’s behavioral factors that drive them to make online purchases through SMPs. Furthermore, this study examines how personal faith influences the connection between the suggested factors and the intention to purchase online. Figure 1 depicts the research paradigm.

Figure 1.

Proposed research framework.

Participants

The investigation employed primary data to augment the presentation and boost the study’s credibility. Business investigations frequently utilize a questionnaire methodology to get insights and viewpoints from consumers and customers (Khan, Rana, & Hosen, 2022; Khan et al., 2024). So, the primary data was obtained through a survey and a well-constructed questionnaire. The study’s population of interest consists of Gen-Z in Dhaka City. In recent years, Gen-Z has traditionally been the primary demographic of SMPs (like Facebook, Instagram, WhatsApp, YouTube, and so on) users. The participants for this study were selected through a combination of non-probability convenience and purposive sampling techniques. Nonprobability sampling is employed due to its efficiency in terms of time and cost (Al Ahad & Rahat Khan, 2020). Among the several methods of non-probability sampling, convenience and purposive sampling methodologies were chosen due to their ease of accessibility, cost-effectiveness, and convenience.

Before completing the final survey, the research project conducted an initial survey with a sample size of 55. The researchers exclusively selected individuals with preexisting knowledge of internet purchasing. A total of 600 students were invited to partake in the research project. The final sample consisted of 551 Gen-Z communities after removing the incomplete questionnaires. The G*power software (version 3.1.9.4) was used in the present study to determine the minimal sample size requirement (Faul et al., 2009; Khan & Roy, 2023). A statistical power of 0.95 and an effect size of 0.05 were utilized to ascertain the required minimal sample size (Roy, 2023g). The software estimated that the lowest suitable sample size was 262. Therefore, the study’s sample size was sufficiently large for statistical evaluation.

Measures

The study included nine constructs: PA, SN, PBC, LS, CE, PT, OR, PF, and PI. Most of the items used in this research were taken from widely recognized scales commonly used in online purchase contexts. A few queries were modified to verify their relevance to the study’s context. In the previous studies, five indicators were adapted for PA (Roy, 2023d), four items were used for SN (Bhagat et al., 2023), and four observations were made for PBC (Sohn & Kwon, 2020). Similarly, four indicators for individual LS, CE, PT, and OR were taken from the work of Miah et al. (2022). Finally, three indicators were adopted from earlier research to measure individual PF and PI (Bhagat et al., 2023). This study utilized a seven-point Likert scale to assess Gen-Z’s purchase intention through SMPs. In this study, 1 = strongly disagree and 7 = strongly agree. Higher scores indicate a high rank in the constructs (Khan & Rammal, 2022; Roy et al., 2023). Construct indicators are included in Table 2.

Gen-Z’s Profile

The study sample consisted of Gen-Z from Dhaka city. Male participants constitute the majority, 69.30% of the total population, and 30.70% are female. The mean age was 22.53 (SD = 1.186 years), with a minimum and maximum of 19 and 25 years, respectively. Most were undergraduates (40.50%), and a few had a business (6.50%). Again, almost 70% of the participants had 2 to 3 years of online purchasing experience. In addition, the average yearly expense was 6,432.71 (SD = 4,202.556 taka) Bangladeshi takas. Table 1 displays the demographic information of the participants.

Table 1.

Gen-Z’s Descriptive Statistics.

Variables	Categories	Frequency	Percent
Gender	Male	382	69.30
Gender	Female	169	30.70
Occupation	Business	36	6.50
	Undergraduate	223	40.50
	Graduate	179	32.50
	Service	113	20.50
Online purchasing experience (in years)	1	40	7.30
	2	202	36.70
	3	181	32.80
	4	95	17.20
	More than 4	33	6.00
Expenditure per year (in Bangladeshi Taka)	0–5,000	300	54.40
	5,001–10,000	176	31.90
	10,001–15,000	55	10.00
	More than 15,000	20	3.60

Table 2.

Outcomes of the Measurement Model.

Variables	Items	Code	Sources	Mean	SD	VIF	λ	α	CR	AVE
Personal attitude (PA)	Purchasing through SMPs is a good idea	PA1	Roy (2023d)	4.878	1.445	2.353	0.826	.897	0.899	0.709
	Purchasing through SMPs is interesting.	PA2	5.005	1.445	3.088	0.887
	Purchasing through SMPs is beneficial.	PA3	5.031	1.516	2.208	0.813
	Purchasing through SMPs is wise.	PA4	4.989	1.474	2.640	0.859
	Purchasing through SMPs is favorable.	PA5	5.225	1.540	2.120	0.822
Subjective norm (SN)	People prefer to shop through SMPs	SN1	Bhagat et al. (2023)	4.103	1.678	2.723	0.891	.902	0.912	0.772
	Friends prefer to shop through SMPs	SN2	4.584	1.313	2.111	0.810
	Shop through SMPs has social acceptance.	SN3	4.122	1.624	3.582	0.914
	My family support to shop through SMPs	SN4	4.162	1.721	3.279	0.897
Perceived behavior control (PBC)	SMP purchase is entirely within my control	PBC1	Sohn and Kwon (2020)	4.279	1.314	1.823	0.828	.874	0.891	0.722
	Good command in handling SMP purchase	PBC2	4.510	1.376	2.911	0.877
	Adequate knowledge of SMP purchase	PBC3	4.552	1.477	2.801	0.827
	Adequate resources and abilities to SMPs purchase	PBC4		4.682	1.496	2.451	0.866
Live streaming (LS)	Social networking sites influence me in SMP purchase	LS1	Miah et al. (2022)	4.475	1.556	1.986	0.788	.864	0.873	0.709
	Hedonic Consumption	LS2	4.490	1.371	2.390	0.844
	Impulsive consumption	LS3	4.203	1.700	2.623	0.881
	cognitive inspiration or assimilation	LS4	4.096	1.765	2.277	0.852
Celebrity endorsement (CE)	Celebrity’s shared posts influence me	CE1	Miah et al. (2022)	4.642	1.499	1.704	0.769	.872	0.879	0.724
	Authentic celebrity posts influence me.	CE2	4.463	1.438	2.658	0.890
	CE fosters my positive feelings.	CE3	4.559	1.465	2.626	0.880
	Recognizable CEs influence me.	CE4	4.577	1.445	2.257	0.860
Promotional tools (PT)	Price discount	PT1	Miah et al. (2022)	4.361	1.628	2.710	0.828	.888	0.911	0.745
	Sales promotion	PT2	4.328	1.679	2.886	0.855
	Bogo technique	PT3	3.942	1.689	2.690	0.897
	Social surroundings	PT4	4.018	1.853	2.354	0.871
Online review (OR)	Good customer ratings	OR1	Miah et al. (2022)	5.129	1.443	2.075	0.824	.857	0.873	0.698
	Reviewers’ reputations	OR2	5.038	1.383	2.129	0.849
	ORs are reliable	OR3	5.120	1.397	1.980	0.826
	Argument quality helps to convince	OR4	4.904	1.275	1.774	0.842
Personal faith (PF)	I prefer online shopping	PF1	Bhagat et al. (2023)	3.918	1.717	2.715	0.910	.894	0.900	0.825
	I trust online shopping from online stores	PF2	4.100	1.723	3.381	0.933
	I feel more confident.	PF3	4.189	1.673	2.408	0.880
Purchase intention (PI)	I tend to visit online sites for purchases	PI1	Bhagat et al. (2023)	4.515	1.558	2.979	0.918	.914	0.914	0.854
	I mostly end up with SMPs purchasing	PI2	4.443	1.724	3.649	0.934
	I am willing to spend more on SMPs purchasing	PI3	4.468	1.759	3.082	0.920

Data Analysis and Results

Data was analyzed using SPSS (v 22) and SmartPLS (4.0.9.8 free version). The proposed research approach was evaluated using artificial neural networks (ANN) and partial least squares-structural equation modeling (PLS-SEM; Hair, 2009; Haykin, 2001). The integration of SEM and ANN approaches is expected to impact the field of social media and online buying significantly. Moreover, it can become a famous and widely used research approach (V.-H. Lee et al., 2020). The current study opted for PLS-SEM instead of the covariance-based Structural Equation Modeling (CB-SEM) technique due to its exploratory nature rather than confirmatory. The use of the PLS-SEM method was inspired by the sophistication of the recommended model and the numerous associated factors (Hair et al., 2021). PLS-SEM does not require the premise of normal distribution for the data. It can also manage direct and moderator analysis within a unified architecture (Becker et al., 2018). The constructs’ significance was assessed using a bootstrapping approach with 5,000 iterations. To accomplish this, we examined the constructs’ path coefficients (Kashyap & Agrawal, 2020; Roy, 2023e). The study used a two-step procedure following the PLS-SEM principles to analyze the findings. The measurement model was computed initially, while the structural model was determined in subsequent stages. Although PLS-SEM can handle non-normal distributions, it cannot analyze non-linear connections between various components. Thus, this study has combined the ANN approach with PLS-SEM to determine the relative significance of the important components (Al-Sharafi et al., 2023; Kalinić et al., 2021). The combined SEM-ANN technique detects and analyses linear and non-linear connections between variables, improving understanding of PI components.

Normality Test

The study utilized a web-based calculator (Z. Zhang & Yuan, 2018) to analyze the data for multivariate normality using Mardia’s (1970) test. For more accurate model prediction, it is crucial to have multivariate normality. The multivariate normality examination indicated that both multivariate skewness (β = 210.2698, p-value < .001) and multivariate kurtosis (β = 1,692.4435, p-value < .001) demonstrated multivariate non-normality (Mardia, 1970). Partial Least Squares Structural Equation Modeling (PLS-SEM) is particularly effective in dealing with data that does not follow a normal distribution (Hair et al., 2019). It provides extra motivation for utilizing PLS-SEM.

Multicollinearity Test

The study assessed multicollinearity to determine independent construct correlation. The calculated path coefficients were affected by the collinearity of the predictor variables. Variance inflation factor (VIF) scores exceeding 5 indicate the presence of collinearity amongst predictor variables (Hair et al., 2019). Table 3 demonstrates that all VIF values fall within a reasonable range. So, it was suggested that multicollinearity does not impact the exogenous constructs’ capacity to predict the endogenous construct accurately. In addition, to avoid the common method bias (CMB), this study applied Harman’s single factor to estimate common variance; the value was 26.859. The estimated value was less than the cut-off value of 50%. So, there are no issues related to common method bias.

Table 3.

Outcomes of Discriminant Validity and VIF.

Fornell-Larcker scores	PA	SN	PBC	LS	CE	PT	OR	PF	PI	VIF
PA										1.436
SN	0.217									1.581
PBC	0.180	0.248								1.107
LS	0.372	0.448	0.194							1.549
CE	0.539	0.469	0.184	0.485						1.711
PT	0.329	0.363	0.062	0.436	0.462					1.385
OR	0.421	0.221	0.146	0.283	0.390	0.131				1.236
PF	0.220	0.584	0.270	0.537	0.423	0.398	0.131			1.650
PI	0.460	0.481	0.263	0.541	0.655	0.435	0.419	0.474
HTMT scores
PA	0.842
SN	0.189	0.879
PBC	0.157	0.230	0.850
LS	0.318	0.411	0.183	0.842
CE	0.474	0.423	0.164	0.426	0.851
PT	0.294	0.346	0.054	0.397	0.417	0.863
OR	0.372	0.195	0.114	0.250	0.338	0.119	0.835
PF	0.198	0.529	0.252	0.478	0.381	0.366	0.124	0.908
PI	0.418	0.440	0.242	0.486	0.588	0.404	0.379	0.430	0.924

Note. On the diagonals of the previous table, the square roots of the AVEs are displayed in bold and italic. Off-diagonal values demonstrate relationships among the various components.

Evaluation of the Measurement Model

Initially, the measurement model was verified for accuracy and reliability. To accomplish this, we comprehensively analyzed multiple facets of the variables. For the internal consistency and reliability of the constructs, the recommended values for factor loadings (λλ), composite reliability (CR), and Cronbach’s alpha (α) are at least .70 (M. T. Pervin & Khan, 2025; Roy, 2022). All the λ, CR, and α values exceed the threshold value. Again, the convergent validity of the constructs was evaluated with the help of average variance extracted (AVE) scores. The recommended value for the AVE is 0.50 (Hair et al., 2019; Roy, 2023a). For this study, all the AVE values are more than 0.50. Therefore, the study confirmed the convergent validity. After that, discriminant validity was assessed. Two commonly used techniques for determining discriminant validity are the Heterotrait-Monotrait Ratio (HTMT) and the Fornell & Larcker criterion (Fornell & Larcker, 1981). The optimal threshold for the HTMT is below 0.80 (Kline, 2015; Roy, 2023d). All of the HTMT values are below the specified threshold. There are no difficulties with discriminant validity. The results are presented in Tables 2 and 3, respectively.

Direct Path Estimation

At this point, the investigators checked the structural model for evidence of the predicted connection (Khan et al., 2019; Khan, Roy, & Chowdhury, 2022; Roy, 2023c). We used PLS-SEM methods to find out how the predictor variables affected PI. The study analyzed the importance of the path coefficient (β) to assess the findings of the structural model. The results indicated that all of the direct assertions were accepted. The evaluation of the structural model is concisely presented in Table 4. The findings demonstrate that there is a substantial correlation between PI and PA (β = .118, p < .01), SN (β = .093, p < .05), PBC (β = .094, p < .01), LS (β = .159, p < .001), CE (β = .274, p < .001), PT (β = .102, p < .01), and OR (β = .144, p < .001). Therefore, all the hypotheses of direct relationships (H1, H2, H3, H4, H5, H6, and H7) were supported. See Figure 2. Structural equation outcomes revealed that CE has the most substantial impact on PI among these independent constructs since β = .274 (p < .001).

Table 4.

Outcomes of the Structural Model.

Hypotheses	Relationships	β	t-Value	p-Values	95% BC-CIs	Supported
Direct effects
H1	PA -> PI	.118	3.387	.001	[0.059; 0.185]	Yes
H2	SN -> PI	.093	2.437	.015	[0.014; 0.154]	Yes
H3	PBC -> PI	.094	2.979	.003	[0.016; 0.143]	Yes
H4	LS -> PI	.159	3.960	.000	[0.092; 0.250]	Yes
H5	CE -> PI	.274	6.996	.000	[0.199; 0.339]	Yes
H6	PT -> PI	.102	2.797	.005	[0.035; 0.152]	Yes
H7	OR -> PI	.144	4.349	.000	[0.049; 0.214]	Yes
Moderation effects
H8	PF × PA -> PI	.110	2.785	.005	[0.023; 0.169]	Yes
H9	PF × SN -> PI	−.043	0.974	.330	[−0.129; 0.054]	No
H10	PF × PBC -> PI	−.040	1.443	.149	[−0.087; 0.018]	No
H11	PF × LS -> PI	.086	1.998	.046	[0.000; 0.160]	Yes
H12	PF × CE -> PI	−.063	1.562	.118	[−0.143; 0.010]	No
H13	PF × PT -> PI	−.060	1.549	.121	[−0.106; 0.041]	No
H14	PF × OR -> PI	−.045	1.325	.185	[−0.106; 0.018]	No

Figure 2.

Findings from the structural equation modeling analysis.

Moderation Path Estimation

In the present analysis, the researchers expected that PF would have a moderate effect on the associations under consideration. The moderation analysis revealed that the connection between PA and PI (β = .110, p < .05) and LS and PI (β = .086, p < .05) were influenced by PF. So, hypotheses H8 and H11 were confirmed. This suggests that BP has a role in enhancing the beneficial relationship between PA and PI and LS and PI. Check Figures 3 and 4. The other moderation hypotheses (H9, H10, H12, H13, and H14) were found insignificant. So, PF did not moderate these relationships. Results are presented in Table 4.

Figure 3.

PF moderates the PA and PI connection.

Figure 4.

PF moderates the SL and PI connection.

Assessment of the Explanatory Power

The coefficient of determination (R²) was computed to assess the explanatory power of the structural model. The estimated score of R² for the PI was .509. Hence, the combination of the independent factors (PA, SN, PBC, LS, CE, PT, and OR) accounted for 50.9% of the variability of PI. As a result, the model demonstrated a substantial capacity to explain. After that, the study calculated the model’s predictive relevance (Q2; Geisser, 1974; Stone, 1974). If the Q2 score is more than 0, it means that the prediction is highly relevant. The model’s exceptional predictive significance was indicated by the Q2 score of 0.476 for PI (Chin et al., 2020).

Results of Artificial Neural Network (ANN)

Artificial Neural Networks (ANNs) are widely employed machine-learning algorithms in various research fields in the present era. Haykin (2001) states that it is a highly homogeneous distributed processor consisting of fundamental computing units with the intrinsic capability of storing experimental data and making it available. It can mimic the cognitive functions of the human brain. There are two phases, namely, the training and testing phases. The technique of discovering underlying correlations within a dataset is accomplished through the training phase, followed by the testing step to demonstrate the obtained knowledge. ANNs are employed to investigate complex relationships since they do not make any assumptions about the distribution of multivariate data (Liébana-Cabanillas et al., 2017). It has a significant advantage over traditional statistical methods such as regression or structural equation modeling (SEM). Conventional statistical methods are restricted to evaluating linear relationships between variables.

In contrast, ANNs have the ability to analyze non-linear relationships. ANNs generally comprise three layers: the input layer, the hidden layer, and the output layer. In a neural network, the activation function establishes the connections between each layer. A commonly used activation function in ANNs is the sigmoid function. Chiang et al. (2006) state that it is preferred because it can compress the original data effectively at both the higher and lower ends. Under the direction of a learning algorithm, the backpropagation neural network accomplishes its tasks. ANNs make extensive use of this method of calculation.

Neural Network (NN) Validation

We used SPSS (22) software to perform the NN model. A widely used training method for multilayer feed-forward backpropagation was employed in the present study. The sigmoid activation and multilayer perceptron algorithms were used to implement the input and hidden nodes. However, overfitting is a notable issue within the realm of ANNs. Therefore, we utilized a 10-fold cross-validation methodology to address this concern (Chong, 2013; Roy, 2023f). For this study, we used 90% of the data points in a training dataset and set aside 10% for testing (Roy, 2023e). The research protocol specifies using a single model for the ANN evaluation. Neurons with the labels PA, SN, PBC, LS, CE, PT, and OR made up the input layer of the ANN model, and the neuron with the name PI made up the output layer. Check Figure 5. The current study validated the results of the ANN assessment using the commonly used Root Mean Square Error (RMSE) criterion (Chong, 2013). It aids in detecting errors for both the training and testing datasets. The Root Mean Square Error evaluation is displayed in Table 5. The mean Root Mean Square Error values for the training and testing techniques are 0.131 and 0.126, respectively, indicating a relatively small error level. Hence, it may be inferred that the results of the ANN evaluation were very dependable, as evidenced by previous research (Chong, 2013).

Table 5.

RMSE Calculation.

Training		Testing		Total sample
N	RSME	N	RSME
488	0.127	63	0.131	551
497	0.136	54	0.119	551
489	0.137	62	0.126	551
484	0.126	67	0.120	551
490	0.132	61	0.124	551
493	0.131	58	0.119	551
491	0.134	60	0.145	551
499	0.133	52	0.120	551
490	0.125	61	0.141	551
494	0.131	57	0.110	551
Mean	0.131		0.126
SD	0.004		0.010

Figure 5.

Architecture of NN.

Sensitivity Analysis

Sensitivity testing in a model assessed how changes to exogenous constructs affected endogenous constructs. The present experiment calculated the average significance of PA, SN, PBC, LS, CE, PT, and OR as independent factors in predicting the PI. Table 6 displays the outcomes of the sensitivity assessment. Based on the results, it has been shown that CE has the most significant degree of effect as an independent factor in its association with PI due to its highest normalized importance of 97.5%, followed by OR (74.4%), LS (57.1%), PA (54.7%), SN (53.9%), PBC (49.3%), and PT (46.5%). Therefore, it can be deduced that the variable CE exhibits the most vital link with students’ online PI.

Table 6.

Sensitivity Assessment.

NN	PA	SN	PBC	LS	CE	PT	OR
1	0.091	0.108	0.104	0.164	0.269	0.080	0.184
2	0.111	0.199	0.101	0.061	0.149	0.185	0.194
3	0.137	0.114	0.128	0.131	0.230	0.085	0.175
4	0.163	0.121	0.098	0.150	0.232	0.105	0.130
5	0.053	0.123	0.108	0.101	0.271	0.142	0.203
6	0.156	0.089	0.098	0.180	0.189	0.120	0.167
7	0.106	0.087	0.126	0.145	0.265	0.105	0.166
8	0.146	0.163	0.112	0.160	0.178	0.077	0.165
9	0.148	0.114	0.131	0.110	0.260	0.079	0.158
10	0.124	0.098	0.142	0.104	0.267	0.083	0.182
Average importance	0.124	0.122	0.115	0.130	0.231	0.106	0.172
Normalized importance	54.7%	53.9%	49.3%	57.1%	97.5%	46.5%	74.4%

Discussion

The research sought to comprehend the behavioral factors influencing Gen-Z’s intention to purchase through SMPs. This research explored the direct influence of PA, SN, PBC, LS, CE, PT, and OR on PI. The results of this investigation support all the previously mentioned direct hypotheses. The outcomes advocate that the H₁: PA is significantly and positively connected with PI. So, PA plays a vital role in refining Gen-Z’s online PI. These explanations boost the understanding of the influence of PA on PI for Gen-Z. Earlier investigations also exposed a significant correlation among these variables (Ha et al., 2019; Rehman et al., 2019; Roy, 2023e). In a similar row, H₂: SN also positively correlated with PI. So, peers, friends, family, and relatives motivate Gen-Z’s intentions to purchase online, and the results were supported by earlier work (Ha et al., 2019; Rehman et al., 2019; Ru et al., 2021). Again, the other TPB construct, H₃: PBC, significantly predicts Gen-Z’s online PI. So, when students believe in their capabilities, they are motivated to buy online. The results were also coherent with other studies (Rehman et al., 2019). The above three hypotheses were based on the TPB model, and the following are the constructs of the extended TPB model. In H₄: The study also reveals that LS positively impacts online PI. So, LS of products and services motivates Gen-Z to purchase online. Contradictorily, Miah et al. (2022) found non-significant results. In addition, other social media variables in H₅, H₆, and H₇ (CE, PT, and OR) were positively associated with Gen-Z’s online PI. Previous studies also supported these results (Ahmed et al., 2015; Fu et al., 2020; Miah et al., 2022; Xiang et al., 2016; Zafar et al., 2021). However, in H₅, CE has the most decisive impact on Gen-Z’s online PI among these behavioral factors. This result was analogous to the research work of Miah et al. (2022). The outcome implies that when celebrities give positive feedback about products or goods, it strongly motivates students to purchase products online.

The observed results confirmed the hypothesized moderation impact of PF on the connection between PA and PI in H₈. Therefore, PF enhances the favorable correlation between PA and PI. That means if the Gen-Z community has trust in online and social media, it positively influences their favorable attitude toward online purchase intention. As a result, with positive attitudes and personal faith, Gen-Z shows more intention to purchase online using SMPs. In addition, in H₁₁, PF also moderates the positive relationship between LS and PI. So, when students believe online purchasing is trustworthy, their perception of LS positively affects their online purchasing. However, the moderating influence of PF on the other associations in H₉, H₁₀, H₁₂, H₁₃, and H₁₄ (SN, PBC, CE, PT, and OR) was statistically non-significant. A potential reason for the lack of significant results could be that individuals from Gen-Z who have a favorable perception of their ability to control their behavior (PBC) and perceive social norms (SN) may not need any extra incentive to participate in online buying activities. They have faith in their abilities and can easily be motivated by their peers or family members to purchase online. Similarly, students motivated by CE, PT, and OR already believe in social media, and this trust helps them to engage in online buying. So, they do not need extra factors to push them to purchase online. Although all PI predictors were shown to be statistically significant, the relative value of these variables concerning the PI is still not obvious.

Therefore, the research placed the PI predictors in order of relevance using ANN modeling. Conclusions from the ANN analysis validated those from the PLS-SEM investigation. The study focuses on an emerging economy (Bangladesh) where consumers have collective values and are mostly from modest socioeconomic backgrounds (Dadzie et al., 2017). As a result, this emerging market stands out from the vast mass market of impoverished buyers (Osei-Frimpong et al., 2019). Therefore, CE was the most important component of Gen-Z’s online PI based on the ANN outcomes. The outcomes of the ANN analysis supported the PLS conclusions (all the behavioral constructs of our validated-extended TPB model in PLS outcomes are very crucial to Gen-Z’s online PI). The above results are important for future theoretical models, research projects, and real-world applications.

Implications

Theoretical Implications

Online purchasing is rapidly gaining popularity. Many people are active on SMPs, such as YouTube, Instagram, WhatsApp, Facebook, etc. Gen-Z are not exempt from this. Gen-Z is viewing various adverts on SMPs, encouraging them to shop online. The outcomes of this investigation have significant theoretical consequences. The present study incorporates PA, SN, PBC, LS, CE, PT, and OR into an inclusive research framework, which clarifies the effect of these behavioral factors on Gen-Z online PI and then examines the vital reason behind Gen-Z’s propensity to purchase online products using social media. Thus, this research contributes significantly to the existing understanding by elucidating the interconnectedness of these key constructs. In addition, this research contributes considerable theoretical instincts by objectively inspecting the impact of four theoretical aspects, namely TPB theory, on the intention of Gen-Z to shop online.

Additionally, it improves the understanding of the TPB constructs of human behavior and confirms the existing knowledge baseline. Subsequently, this study investigates the role of PF as a moderator between TPB and social media characteristics concerning online PI among Gen-Z. The TPB is highly effective in predicting consumer behavior. However, it does have several shortcomings, including its failure to account for factors such as trust, perceived dangers, financial resources, customer satisfaction, commitment, and past consumer experiences (Rehman et al., 2019). Thus, this study employed PF within the framework of the TPB theory to ascertain the online PI of Gen-Z. This study first time presents PF as a moderating aspect of the relationship between TPB factors and students’ online PI.

Moreover, the moderated model has a deeper understanding and better prediction power than a straight correlation between independent variables and PI. A complete model analysis of PA, SN, PBC, LS, CE, PT, and OR’s direct and indirect relationships with PI offers a substantial theoretical contribution. Several studies have attempted this task separately; however, none have combined them. This study used a two-stage technique to determine Gen-Z’s online PI parameters. This unique methodology may be the first instance of its use in this specific context. The integrated method makes the research model more versatile than existing theories. This method connects theory with practice, creating a wider paradigm. The revelation may improve individuals’ knowledge of internet PI and enlarge their viewpoint.

Practical Implications

The findings of ongoing research have several practical consequences for businesses selling online. It shows that the TPB factors (PA, SN, and PBC) and the social media constructs (LS, CE, PT, and ORs) favorably correlate with the online PI of Gen-Z. According to the results, all the measured behavioral factors significantly impact Gen-Z online PI. Therefore, online businesses must encourage Gen-Z to increase their positive attitudes toward online purchasing. Enterprises may offer relevant information, precise pricing details, and efficient delivery services, which can effectively promote Gen-Z’s attitude to participate in online shopping. Again, friends and family members play a crucial role in online purchasing. They share their positive or negative feelings about products or services with the nearest one. When these people are satisfied with online products, their satisfaction assists them in persuading their family members to buy products online. In addition, the perceived ease of a business website also stimulates people to make online purchases. When Gen-Z sees the websites as user-friendly and controllable (such as being able to order or pay for things quickly), and the related marketing activities (such as advertising, promotions, pricing, distribution strategies, etc.) are campaigned based on their measured behavior factors, it ultimately stimulates them to engage with the online purchasing. Hence, online merchants may craft their business strategies based on Gen-Z’s measured behavioral factors, enhancing their intentions to purchase online. Also, this study’s results indicated that personal trust plays a crucial and favorable role in Gen-Z’s intentions toward online purchasing. So, An e-commerce platform must establish consumer trust to encourage online purchases instead of physical transactions. Marketers are advised to prioritize the development of captivating advertisements that can elicit emotional responses from consumers, thereby strengthening their connection with the service provider and enhancing brand loyalty.

Conclusion, Limitations, and Future Directions

Gen-Z are heavy users of online platforms. They effortlessly gain knowledge about recently presented products through social media and make essential purchases using SMPs. Various facets of social media and interpersonal elements, such as attitude and behavioral control, influence students’ online buying choices. Thus, this research aims to examine the impact of behavioral factors on the online purchasing habits of Gen-Z in Bangladesh by applying extended TPB theory. This study concludes that PA, SN, PBC, LS, CE, PT, and OR positively and significantly influence Gen-Z’s online PI. In addition, PF significantly moderates the relationships between PA, PI, LS, and PI. Thus, the study offers practical recommendations for internet-based enterprises on how to utilize SMPs for targeted advertising and promotional attempts efficiently. In addition, the study uniquely considers Gen-Z’s purchase intention (PI) through SMPs by applying expanded TPB constructs and adopting a hybrid approach that combines SEM with ANN analysis.

Certain constraints remain that persist and necessitate consideration in future research. This study collected data using non-probability convenience and purposive sampling. Nonprobability sampling can produce a non-representative population sample and be subject to selection and response bias. Therefore, future research may use probability sampling to represent an accurate population. Furthermore, the study only sampled Gen-Z of Dhaka City, although different geographic areas of Bangladesh or even comparing separated zones could affect PI concerns. It may limit the study’s generalizability. Future studies should use more diverse and representative populations.

Similarly, the present study is limited to one culture and has not considered cross-cultural differences; future studies should include more cross-cultural assessments. The present study relies on quantitative data, but the outcomes may vary when it includes qualitative data. Subsequent investigations should consist of a blend of qualitative and quantitative analysis. Likewise, the present investigation utilized only direct relationships and PF as moderating variables. Future research actions may contemplate mediating effects by considering mediator variables, such as electronic satisfaction or online security. Other unmeasured factors like peer influence or brand loyalty also play a role in purchase intention; therefore, future research may consider these factors to extend the TPB model for online PI.

Footnotes

ORCID iDs

Sanjoy Kumar Roy

Udayshankar Sarkar

Md. Rahat Khan

Abdullah Al Ahad

Ethical Considerations

The researchers confirm that the research was performed in accordance with relevant guidelines/regulations applicable when human participants are involved (e.g., Declaration of Helsinki or similar).

Consent to Participate

Informed consent was obtained from all participants in this study.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

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.

Data Availability Statement

The data that support the findings of this study are not openly available due to securing the anonymity of the respondents and their institutions. However, anonymized data sets are available from the corresponding author upon reasonable request.

References

Abror

Patrisia

Engriani

Evanita

Yasri

Dastgir

(2020). Service quality, religiosity, customer satisfaction, customer engagement and Islamic bank’s customer loyalty. Journal of Islamic Marketing, 11(6), 1691–1705.

Adoeng

Kalangi

J. B.

Wangke

S. J. C.

(2019). A comparative analysis of E-advertisement between jd. Id and shopee customers in Manado. Jurnal EMBA, 7(3), 3379–3388.

Ahmed

Seedani

Ahuja

Paryani

(2015). Impact of celebrity endorsement on consumer buying behavior. SSRN 2666148.

Ajina

A. S.

(2019). The perceived value of social media marketing: An empirical study of online word-of-mouth in Saudi Arabian context. Entrepreneurship and Sustainability Issues, 6(3), 1512.

Ajzen

(1991). The theory of planned behavior. Organizational Behavior and Human Decision Processes, 50(2), 179–211.

Al Ahad

Rahat Khan

. (2020). Exploring the mediating effect of demographic attributes on the relationship between employee engagement and organizational citizenship behavior. International Journal of Management and Sustainability, 9(1), 11–23.

Al-Sharafi

M. A.

Al-Emran

Iranmanesh

Al-Qaysi

Iahad

N. A.

Arpaci

(2023). Understanding the impact of knowledge management factors on the sustainable use of AI-based chatbots for educational purposes using a hybrid SEM-ANN approach. Interactive Learning Environments, 31, 7491–7510.

Ameen

Hosany

Taheri

(2023). Generation Z’s psychology and new-age technologies: Implications for future research. Psychology and Marketing, 40(10), 2029–2040. https://doi.org/10.1002/mar.21868

Armutcu

Ramadani

Zeqiri

Dana

L. P.

(2024). The role of social media in consumers’ intentions to buy green food: Evidence from Türkiye. British Food Journal, 126(5), 1923–1940.

10.

Ashfaq

Ali

(2017). Impact of celebrity endorsement on consumer buying behavior in FMCG sector of Pakistan. Arabian Journal of Business and Management Review (Oman Chapter), 7(1), 1–12.

11.

Becker

J. M.

Ringle

C. M.

Sarstedt

(2018). Estimating moderating effects in PLSSEM and PLSc-SEM: Interaction term generation* data treatment. Journal of Applied Structural Equation Modeling, 2(2), 1–21.

12.

Bejan

B. M.

(2023). How does self-esteem impact generation Z’s behavioral intentions? Ecoforum, 12(1), 1–8. http://www.ecoforumjournal.ro/index.php/eco/article/view/1725

13.

Bhagat

Chauhan

Bhagat

(2023). Investigating the impact of artificial intelligence on consumer’s purchase intention in e-retailing. Foresight, 25(2), 249–263.

14.

Bhalla

(2020). Demystifying the key antecedents of consumer trust in online shopping and testing the mediating role of consumer trust: An empirical study. Journal of Marketing Management, 19(1), 7–22.

15.

Chen

C.-C.

Lin

Y.-C.

(2018). What drives live-stream usage intention? The perspectives of flow, entertainment, social interaction, and endorsement. Telematics and Informatics, 35(1), 293–303.

16.

Cheng

B. L.

Yee

S. W.

(2014). Factors influencing consumers’ online purchase intention: A study among university students in Malaysia. International Journal of Liberal Arts and Social Science, 2(8), 121–133.

17.

Chen

Yuan

(2023). Impacts of built environment on travel behaviors of Generation Z: A longitudinal perspective. Transportation, 50(2), 407–436. https://doi.org/10.1007/s11116-021-10249-6

18.

Chiang

W. Y. K.

Zhang

Zhou

(2006). Predicting and explaining patronage behavior toward web and traditional stores using neural networks: A comparative analysis with logistic regression. Decision Support Systems, 41(2), 514–531.

19.

Chin

Cheah

J.-H.

Liu

Ting

Lim

X.-J.

Cham

T. H.

(2020). Demystifying the role of causal-predictive modeling using partial least squares structural equation modeling in information systems research. Industrial Management & Data Systems, 120, 2161–2209.

20.

Chong

A. Y. L.

(2013). A two-staged SEM-neural network approach for understanding and predicting the determinants of m-commerce adoption. Expert Systems with Applications, 40(4), 1240–1247.

21.

Cooley

Parks-Yancy

(2019). The effect of social media on perceived information credibility and decision making. Journal of Internet Commerce, 18(3), 249–269.

22.

Dadić

Jerčić

Dadić

(2022). Behavior of Generation Z. Proceedings of IAC in Vienna 2022 (pp. 10–16).

23.

Dadzie

K. Q.

Amponsah

D. K.

Dadzie

C. A.

Winston

E. M.

(2017). How firms implement marketing strategies in emerging markets: An empirical assessment of the 4A marketing mix framework. Journal of Marketing Theory and Practice, 25(3), 234–256.

24.

Dehkordi

G. J.

Rezvani

Rahman

M. S.

Nahid

F. F. N.

Jouya

S. F.

(2012). A conceptual study on E-marketing and its operation on firm’s promotion and understanding customer’s response. International Journal of Business and Management, 7(19), 114.

25.

Donthu

Gustafsson

(2020). Effects of COVID-19 on business and research. Journal of Business Research, 117, 284–289.

26.

Eyre

De Luca

Simini

(2020). Social media usage reveals recovery of small businesses after natural hazard events. Nature Communications, 11(1), 1629.

27.

Faul

Erdfelder

Buchner

Lang

A. G.

(2009). Statistical power analyses using G*Power 3.1: Tests for correlation and regression analyses. Behavior Research Methods, 41(4), 1149–1160.

28.

Fazio

R. H.

Ledbetter

J. E.

Towles-Schwen

(2000). On the costs of accessible attitudes: Detecting that the attitude object has changed. Journal of Personality and Social Psychology, 78(2), 197–210.

29.

Fishbein

Ajzen

(1977). Belief, attitude, intention, and behavior: An introduction to theory and research. Penn State University Press.

30.

Fornell

Larcker

D. F.

(1981). Evaluating structural equation models with unobservable variables and measurement error. Journal of Marketing Research, 18(1), 39–50.

31.

Francis

Eccles

M. P.

Johnston

Walker

A. E.

Grimshaw

J. M.

Foy

Kaner

E. F. S.

Smith

Bonetti

(2004). Constructing questionnaires based on the theory of planned behaviour: A manual for health services researchers. Centre for Health Services Research, University of Newcastle upon Tyne.

32.

Manogaran

Cao

Jiang

Yang

(2020). Intelligent decision-making of online shopping behavior based on Internet of things. International Journal of Information Management, 50, 515–525.

33.

Gan

Wang

(2015). Uses and gratifications of social media: A comparison of microblog and WeChat. Journal of Systems and Information Technology, 17(4), 351–363.

34.

Gao

Bai

(2014). A unified perspective on the factors influencing consumer acceptance of Internet of things technology. Asia Pacific Journal of Marketing and Logistics, 26(2), 211–231.

35.

Geisser

(1974). A predictive approach to the random effect model. Biometrika, 61(1), 101–107.

36.

Gupta

Agarwal

Singh

P. B.

(2020). To study the impact of instafamous celebrities on consumer buying behavior. Academy of Marketing Studies Journal, 24(2), 1–13.

37.

Hair

J. F.

(2009). Multivariate data analysis. Pearson Prentice Hall.

38.

Hair

J. F.

Jr Hult

G. T. M.

Ringle

C. M.

Sarstedt

(2021). A primer on partial least squares structural equation modeling (PLS-SEM). Sage Publications.

39.

Hair

J. F.

Risher

J. J.

Sarstedt

Ringle

C. M.

(2019). When to use and how to report the results of PLS-SEM. European Business Review, 31(1), 2–24.

40.

Hanaysha

J. R.

(2022). Impact of social media marketing features on consumer’s purchase decision in the fast-food industry: Brand trust as a mediator. International Journal of Information Management Data Insights, 2(2), 100102.

41.

Hansen

J. M.

Saridakis

Benson

(2018). Risk, trust, and the interaction of perceived ease of use and behavioral control in predicting consumers’ use of social media for transactions. Computers in Human Behavior, 80, 197–206.

42.

N. T.

Nguyen

T. L. H.

Nguyen

T. P. L.

Nguye

T. D.

(2019). The effect of trust on consumers’ online purchase intention: An integration of TAM and TPB. Management Science Letters, 9(9), 1451–1460.

43.

Harland

Staats

Wilke

H. A. M.

(1999). Explaining proenvironmental intention and behavior by personal norms and the theory of planned behavior. Journal of Applied Social Psychology, 29(12), 2505–2528.

44.

Haykin

S. S.

(2001). Neural networks: A comprehensive foundation. Tsinghua University Press.

45.

Helm

Eggert

Garnefeld

(2010). Modeling the impact of corporate reputation on customer satisfaction and loyalty using partial least squares. In Esposito Vinzi

Chin

Henseler

Wang

(Eds.), Handbook of partial least squares: Concepts, methods and applications (pp. 515–534). Springer. https://doi.org/10.1007/978-3-540-32827-8_23

46.

Hossain

Chowdhury

H. K.

Hasan

Shamsuzzaman

Fahim

A. Y.

Khan

M. Y. H.

(2020). Banking service in Bangladesh: The impact of service marketing mix on purchase intention of university students. Strategic Change, 29(3), 363–374.

47.

Hossain

S. S.

Khan

M. R.

(2021). Perception of distance learning in Bangladeshi tertiary education: Prospects and obstacles in the Covid-19 era. Journal of Social, Humanity, and Education, 1(3), 197–207.

48.

Jadil

Rana

N. P.

Dwivedi

Y. K.

(2022). Understanding the drivers of online trust and intention to buy on a website: An emerging market perspective. International Journal of Information Management Data Insights, 2(1), 100065.

49.

Jenefa

(2017). Impact of digital advertisement on garments buying behavior. International Journal of Transformation in Operational and Marketing Management, 1, 1.

50.

Kalinić

Marinković

Kalinić

Liébana-Cabanillas

(2021). Neural network modeling of consumer satisfaction in mobile commerce: An empirical analysis. Expert Systems with Applications, 175, 114803.

51.

Kashif

Zarkada

Ramayah

(2018). The impact of attitude, subjective norms, and perceived behavioural control on managers’ intentions to behave ethically. Total Quality Management & Business Excellence, 29(5-6), 481–501.

52.

Kashyap

Agrawal

(2020). Scale development and modeling of intellectual property creation capability in higher education. Journal of Intellectual Capital, 21(1), 115–138.

53.

Khan

M. R.

Pervin

M. T.

Arif

M. Z. U.

Hossain

S. M. K.

(2024). The impact of technology service quality on Bangladeshi banking consumers’ satisfaction during the pandemic situation: Green development and innovation perspective in banking service. Innovation and Green Development, 3(2), 100120.

54.

Khan

M. R.

Rammal

H. G.

(2022). The role of trade unions in promoting CSR in Bangladesh’s ready-made garments sector: The post-pandemic scenario. Management Decision, 60(10), 2786–2800.

55.

Khan

M. R.

Rana

Hosen

M. I.

(2022). Impact of trustworthiness on the usage of m-banking apps: A study on Bangladeshi consumers. Business Perspectives and Research, 10(2), 234–250.

56.

Khan

M. R.

Roy

Chowdhury

(2022). The impact of work-family balance on manufacturing employees’ mental health during pandemic. Организационнаяпсихология, 12(4), 29–40.

57.

Khan

M. R.

Roy

S. K.

(2023). Moderating effect of M-banking apps users’ demographic variables on the relationship between the ease of use and brand trust. European Journal of Business Science and Technology, 9(2), 249–265.

58.

Khan

M. R.

Roy

S. K.

Hossain

S. M. K.

(2019). Factors affecting garments employees’ perception on job performance: Evidence from Bangladesh. International Journal of Management and Sustainability, 8(1), 32–47.

59.

Khan

M. R.

Sarkar

(2024). Business leadership in the digital age: IKEA’s story in the retail and consumer sector. Green Technologies and Sustainability, 2, 100081.

60.

Kiani

Laroche

Paulin

(2016). Development of market Mavenism traits: Antecedents and moderating effects of culture, gender, and personal beliefs. Journal of Business Research, 69(3), 1120–1129.

61.

Kline

R. B.

(2015). Principles and practice of structural equation modeling. Guilford Publications.

62.

Knowles

S. R.

Hyde

M. K.

White

K. M.

(2012). Predictors of young people’s charitable intentions to donate money: An extended theory of planned behavior perspective. Journal of Applied Social Psychology, 42(9), 2096–2110.

63.

Lee

M. K. O.

Turban

(2001). A trust model for consumer internet shopping. International Journal of Electronic Commerce, 6(1), 75–91.

64.

Lee

V.-H.

Hew

J.-J.

Leong

L.-Y.

Tan

G. W.-H.

Ooi

K.-B.

(2020). Wearable payment: A deep learning-based dual-stage SEM-ANN analysis. Expert Systems with Applications, 157, 113477.

65.

C.-Y.

(2019). How social commerce constructs influence customers’ social shopping intention? An empirical study of a social commerce website. Technological Forecasting and Social Change, 144, 282–294.

66.

Liébana-Cabanillas

Marinković

Kalinić

(2017). A SEM-neural network approach for predicting antecedents of m-commerce acceptance. International Journal of Information Management, 37(2), 14–24.

67.

Lim

Y. J.

Osman

Salahuddin

S. N.

Romle

A. R.

Abdullah

(2016). Factors influencing online shopping behavior: The mediating role of purchase intention. Procedia Economics and Finance, 35, 401–410.

68.

Mardia

K. V.

(1970). Measures of multivariate skewness and kurtosis with applications. Biometrika, 57, 519–530.

69.

McCormick

(2016). Celebrity endorsements: Influence of a product-endorser match on Millennials attitudes and purchase intentions. Journal of Retailing and Consumer Services, 32, 39–45.

70.

McKnight

D. H.

Chervany

N. L.

(2001). What trust means in e-commerce customer relationships: An interdisciplinary conceptual typology. International Journal of Electronic Commerce, 6(2), 35–59.

71.

McKnight

D. H.

Choudhury

Kacmar

(2002). Developing and validating trust measures for e-commerce: An integrative typology. Information Systems Research, 13(3), 334–359.

72.

Miah

M. R.

Hossain

Shikder

Saha

Neger

(2022). Evaluating the impact of social media on online shopping behavior during COVID-19 pandemic: A Bangladeshi consumers’ perspectives. Heliyon, 8(9), e10600.

73.

Munsch

(2021). Millennial and generation Z digital marketing communication and advertising effectiveness: A qualitative exploration. Journal of Global Scholars of Marketing Science, 31(1), 10–29. https://doi.org/10.1080/21639159.2020.1808812

74.

Osei-Frimpong

Donkor

Owusu-Frimpong

(2019). The impact of celebrity endorsement on consumer purchase intention: An emerging market perspective. Journal of Marketing Theory and Practice, 27(1), 103–121.

75.

Patel

A. K.

Singh

Rana

N. P.

Parayitam

Dwivedi

Y. K.

Dutot

(2023). Assessing customers’ attitude towards online apparel shopping: A three-way interaction model. Journal of Business Research, 164, 113917.

76.

Pervin

M. T.

Khan

M. R.

(2025). The impact of social media marketing on consumers’ buying decisions in the restaurant industry: Mediating role of trustworthiness. Jindal Management Journal, 14, 28–48. https://doi.org/10.1177/22786821241235741

77.

Pervin

Khan

(2022). The impact of organizational career management and supporting system on generation Y employees’ career satisfaction and intention to stay in the textile industry. SCMS Journal of Indian Management, 19(3), 34–50.

78.

Pop

R.-A.

Săplăcan

Dabija

D.-C.

Alt

M.-A.

(2022). The impact of social media influencers on travel decisions: The role of trust in consumer decision journey. Current Issues in Tourism, 25(5), 823–843.

79.

Putri

I. S.

Akbari

A. D.

(2021). Extended theory of planned behavior (TPB) to analyze the batik purchase intention of Indonesian millennials and gen Z. International Journal of Industrial Engineering and Engineering Management, 3(2), 97–104.

80.

Rehman

S. U.

Bhatti

Mohamed

Ayoup

(2019). The moderating role of trust and commitment between consumer purchase intention and online shopping behavior in the context of Pakistan. Journal of Global Entrepreneurship Research, 9(1), 1–25.

81.

Roy

S. K.

(2022). The impact of age, gender, and ethnic diversity on organizational performance: An empirical study of Bangladesh’s banking sector. International Journal of Financial, Accounting, and Management, 4(2), 145–161.

82.

Roy

S. K.

(2023a). E-learning portal success in higher education organizations: A multi-group comparison. Journal of Social, Humanity, and Education, 3(3), 197–218.

83.

Roy

S. K.

(2023b). Factors influencing undergraduate students’ green entrepreneurial intentions: Evidence from an emerging economy. Entrepreneurship Education, 6, 365–398.

84.

Roy

S. K.

(2023c). Green factors and green behavioral intentions: A hybrid two-stage modeling approach. SSRN 4462203.

85.

Roy

S. K.

(2023d). Impact of green factors on undergraduate students’ green behavioral intentions: A hybrid two-stage modeling approach. Heliyon, 9, e20630.

86.

Roy

S. K.

(2023e). Impact of SMS advertising on purchase intention for young consumers. International Journal of Financial, Accounting, and Management, 4(4), 427–447.

87.

Roy

S. K.

(2023f). Influencing factors for pursuing agriculture as a career for agriculture undergraduates: A two-stage approach. Entrepreneurship Education, 6(2), 169–203.

88.

Roy

S. K.

(2023g). YouTube’s influential factors for academic achievement: A two-stage approach. Telematics and Informatics Reports, 10, 100060.

89.

Roy

S. K.

Khan

M. R.

Shanto

N. I.

(2023). Bankers’ job stress, job performance, and job commitment trajectories during the COVID-19 pandemic. Journal of Entrepreneurship, Management and Innovation, 19(3), 203–228.

90.

L. J.

Kowang

T. O.

Long

C. S.

Fun

F. S.

Fei

G. C.

(2021). Factors influencing online purchase intention of Shopee’s consumers in Malaysia. International Journal of Academic Research in Business and Social Sciences, 11(1), 761–776.

91.

Sahi

G. K.

Sekhon

H. S.

Quareshi

T. K.

(2016). Role of trusting beliefs in predicting purchase intentions. International Journal of Retail & Distribution Management, 44(8), 860–880.

92.

Salleh

M. S. M.

Mahbob

N. N.

Baharudin

N. S.

(2017). Overview of “generation Z” behavioural characteristic and its effect towards hostel facility. International Journal of Real Estate Studies, 11(2), 59–67.

93.

Sanny

Arina

A. N.

Maulidya

R. T.

Pertiwi

R. P.

(2020). Purchase intention on Indonesia male’s skin care by social media marketing effect towards brand image and brand trust. Management Science Letters, 10(10), 2139–2146.

94.

Setiawan

Purwanto

Ikasari

W. S. D.

Suryadi

(2024). An inclusive extension of the theory of planned behavior for explaining household food leftover reduction intention among gen Z. Journal of Social Marketing, 14, 328–346. https://doi.org/10.1108/jsocm-09-2023-0210

95.

Shamout

M. D.

(2016). The impact of promotional tools on consumer buying behavior in retail market. International Journal of Business and Social Science, 7(1), 75–85.

96.

Siddique

M. Z. R.

Hossain

(2018). Sources of consumers awareness toward green products and its impact on purchasing decision in Bangladesh. Journal of Sustainable Development, 11(3), 9–22.

97.

Siddiqui

Bano

Al Rousan

(2023). Understanding intention of gen Z Indians to visit heritage sites by applying extended theory of planned behaviour: A sustainable approach. Journal of Cultural Heritage Management and Sustainable Development. Advance online publication. https://doi.org/10.1108/jchmsd-03-2022-0039

98.

Siddiqui

Bano

Hamid

(2022). Travelling to tourism destinations through the lens of sustainability: An extended TPB model to predict behavioural intention of gen Z consumers. Journal of Tourism, Sustainability and Well-being, 10(3), 172–188.

99.

Sin

S. S.

Nor

K. M.

Al-Agaga

A. M.

(2012). Factors affecting Malaysian young consumers’ online purchase intention in social media websites. Procedia - Social and Behavioral Sciences, 40, 326–333.

100.

Sohn

Kwon

(2020). Technology acceptance theories and factors influencing artificial intelligence-based intelligent products. Telematics and Informatics, 47, 101324.

101.

Sohrabi

Alsafi

O’Neill

Khan

Kerwan

Al-Jabir

Iosifidis

Agha

(2020). World Health Organization declares global emergency: A review of the 2019 novel coronavirus (COVID-19). International Journal of Surgery, 76, 71–76.

102.

Stone

(1974). Cross-validation and multinomial prediction. Biometrika, 61(3), 509–515.

103.

Subawa

N. S.

Widhiasthini

Pika

P. A. T. P.

Suryawati

P. I.

Astawa

I. N. D.

(2020). Generation Z behavior and low price products in the era of disruption. International Journal of Social Sciences and Management, 3(3), 1–12.

104.

Taobangdan

T. L. S.

(2019). Taobao live streaming ecological development report. Report.

105.

Taylor

Todd

(1995). Assessing IT usage: The role of prior experience. MIS Quarterly, 19, 561–570.

106.

Teah

Lwin

Cheah

(2014). Moderating role of religious beliefs on attitudes towards charities and motivation to donate. Asia Pacific Journal of Marketing and Logistics, 26(5), 738–760.

107.

Verma

Miglani

Brar

(2019). Impact of fashion advertising through social media portals on the purchase decision of young adults. International Journal of Social Science Research, 9(8), 187–199.

108.

Verma

V. K.

Chandra

(2018). An application of theory of planned behavior to predict young Indian consumers’ green hotel visit intention. Journal of Cleaner Production, 172, 1152–1162.

109.

Wang

Pacho

Liu

Kajungiro

(2019). Factors influencing organic food purchase intention in developing countries and the moderating role of knowledge. Sustainability, 11(1), 209.

110.

Waris

Hameed

Ali

(2023). Predicting household sign up for solar energy: An empirical study based on the extended theory of planned behavior. International Journal of Energy Sector Management, 17(3), 455–473.

111.

Weismueller

Harrigan

Wang

Soutar

G. N.

(2020). Influencer endorsements: How advertising disclosure and source credibility affect consumer purchase intention on social media. Australasian Marketing Journal, 28(4), 160–170.

112.

Wilcox

Stephen

A. T.

(2013). Are close friends the enemy? Online social networks, self-esteem, and self-control. Journal of Consumer Research, 40(1), 90–103.

113.

Wongkitrungrueng

Assarut

(2020). The role of live streaming in building consumer trust and engagement with social commerce sellers. Journal of Business Research, 117, 543–556.

114.

Xiang

Zheng

Lee

M. K. O.

Zhao

(2016). Exploring consumers’ impulse buying behavior on social commerce platform: The role of parasocial interaction. International Journal of Information Management, 36(3), 333–347.

115.

Yadav

B. S.

(2016). The role of social media communication in the branding of educational hubs. IUP Journal of Soft Skills, 10(4), 51.

116.

Yahia

I. B.

Al-Neama

Kerbache

(2018). Investigating the drivers for social commerce in social media platforms: Importance of trust, social support and the platform perceived usage. Journal of Retailing and Consumer Services, 41, 11–19.

117.

Yahya

S. F. H.

Hashim

N. A.

Bahsri

Dahari

N. A.

(2019). The effect of sales promotion strategy on online fashion shopping behavior among employee of Sahawan Sdn Bhd. Global Business and Management Research: An International Journal, 11(2), 1–13.

118.

Yin

(2020, July 19–24). A study on the influence of E-commerce live streaming on consumer’s purchase intentions in mobile Internet [Conference session]. HCI International 2020–Late Breaking Papers: Interaction, Knowledge and Social Media: 22nd HCI International Conference, HCII 2020, Copenhagen, Denmark, Proceedings 22 (pp. 720–732).

119.

Zafar

A. U.

Qiu

Wang

Shahzad

(2021). The impact of social media celebrities’ posts and contextual interactions on impulse buying in social commerce. Computers in Human Behavior, 115, 106178.

120.

Zhang

Qin

Wang

G. A.

Luo

(2020). The impact of live video streaming on online purchase intention. Service Industries Journal, 40(9–10), 656–681.

121.

Zhang

Yuan

K. H.

(2018). Practical statistical power analysis using Webpower and R. ISDSA Press.

122.

Zhu

Zhang

(2010). Impact of online consumer reviews on sales: The moderating role of product and consumer characteristics. Journal of Marketing, 74(2), 133–148.