Drivers and Impact of Intention to Adopt Online Purchases: The Moderating Effects of Cultural Values

This study examines the factors determining the adoption of online purchases among customers and the influence of cultural variables in an African context. The research is based on a combination of IS/IT theories and models about the adoption of online purchases. The hypotheses are tested using structural equation modeling (PLS-SEM) on a sample of 446 individuals. The findings show that rational perception variables drive users' intention to adopt online purchases; cultural values in Africa affect online purchases. Customers who value physical interaction with the product are more likely to make online purchases, primarily for hedonic reasons, and the relationship between utilitarian expectation and purchase intention depends on the level of conformity to the group and the effect of intention to adopt online purchases on repeat purchase intention is positive and stronger among consumers with low group conformity. The implications and limitations of the research are explained.


Introduction
The rapid expansion of Internet-related technologies and services has significantly altered the distribution landscape, with new and innovative approaches to doing business, such as e-commerce.Online shopping is, therefore, a key element of the modern shopping environment (Lu & Wang, 2018).This business model is growing rapidly, especially in developing African countries with low rates of technology adoption (Statista, 2021).A large proportion of African populations do not yet utilize this technology because it is new and was developed for Western consumers, with specific cultural characteristics (Faqih, 2016;Hofstede, 1993;Van Slyke et al., 2010).
In Cameroon, despite promising statistics (Statista, 2021), online retailers face two main challenges.The first is the logistic challenge of digitalizing economic and social practices.A large percentage of the population lives in a so-called "popular" economy (Bikanda & Mefoute Badiang, 2017); financial inclusion indicators for Cameroon are very low (Demirgüç -Kunt et al., 2018;Hootsuite, 2020).Poor logistics make the product delivery process difficult, as there is no fixed address system in the country.The second challenge relates to the sociocultural characteristics rooted in the environment.Past studies have stressed the importance of transactions based on interpersonal interactions; since the oral channel is the most common medium of negotiation, retailers that use it well have greater power of persuasion (Assande, 1996;Bikanda & Mefoute Badiang, 2017).However, online retailers often adopt successful models from the West as part of a standardization strategy, ignoring the fact that traditions strongly influence the behavior of Africans.According to Diop & Merunka (2013), Africa is unique in that local traditions and norms influence consumer behavior.Understanding online shopping adoption, therefore, necessitates a holistic approach that includes both universal and cultural aspects of human behavior (De Maricourt & Ollivier, 1990).
Accordingly, recent studies suggest the need to integrate cultural dimensions into the technology adoption model (Chen et al., 2015;Zhang et al., 2012).Cultural values in different countries could explain the varying levels of acceptance of new technologies (Warkentin et al., 2015;Yuen et al., 2015;Zhao et al., 2014).Until now, studies that integrate cultural aspects have compared several countries from a multicultural perspective (Hofstede, 1993) within the framework of international marketing.These cross-cultural studies have prioritized samples from North America, Western Europe, and Asian countries (Mazaheri et al., 2011).Significant research interest has been shown in the behavioral dynamics of consumers in western or Asian markets (Rehman et al., 2020;Shin et al., 2007).However, academic literature on cultural aspects in Africa is scant (Virkijika, 2018).The conclusions emanating from certain research studies are based on propositions that require empirical verification (Bikanda & Mefoute Badiang, 2017).The complexity of consumer behaviors and expectations in Africa needs to be better considered by online retailers.Consumers in Africa are influenced by both global and local consumption cultures (Diop & Merunka, 2013).Through this study, we examine the relationships between consumers' behavioral intention in adopting online shopping and its antecedents.More precisely, we examine the moderating impact of cultural aspects on these relationships, especially in an environment where this distribution model is at an early stage.
We begin by presenting the theoretical foundation and the hypotheses underlying the model.Then, we present the research method, and discuss the empirical findings.We conclude by stating the limitations of the study and the avenues for future research.

Literature Review
Various models have explained technology adoption.Most studies have focused on IS/IT adoption theories and models (Babones, 2015;Chatterjee, 2019;Diegmann et al., 2017).These include the technology acceptance model (TAM) (Davis, 1989;Venkatesh & Davis, 2000), the unified theory of acceptance and use of technology (UTAUT) (Venkatesh & Davis, 2000;Venkatesh et al., 2003;Venkatesh et al., 2012), the information systems success model (ISSM) (Baabdullah et al., 2019;DeLone & McLean, 2003;Zhou, 2013), and the protection motivation theory (PMT) (Charki et al., 2011;Floyd et al., 2000;Rogers, 1975).The TAM is one of the most popular models (Dauda & Lee, 2015;Shaikh & Karjaluoto, 2015).However, it mainly considers two factors of technology adoption: perceived usefulness (PU) and perceived ease of use (PEOU).The TAM does not include other external factors, limiting its potential to explain consumer behavior (Venkatesh et al., 2003).To address these limitations, Venkatesh et al. (2003) combined eight (08) prominent theories into a single model called the unified theory of acceptance and use of technology.However, Hofstede (2001) argues that these popular theories may not entirely apply to other cultures as most of them were developed within the North American culture.Indeed, technological capabilities must correspond to the consumer's lifestyle to facilitate the adoption and use of the technology (Zhu et al., 2009).Considering this, cross-cultural studies suggest an adaptation of existing models to accommodate new constructs (Malhotra & McCort, 2001;Sawang et al., 2006).Malaquias & Hwang (2019) show that while in developed countries, such as the United States, social influence plays a critical role in technology adoption, perceived ease of use is the main determinant in emerging economies such as Brazil.Research by Singh et al. (2020) shows that ease of use, perceived usefulness, perceived risk, and attitude are the determining factors in India.In the African context, different factors seem to be at play; for example, Matemba & Li (2018) found that South Africans were more likely to adopt innovative technologies that are trustworthy, secure, and mindful of privacy.In Cameroon, technology adoption is driven by hedonic motivations, social influence, trust, facilitating conditions, and perceived risk (Virkijika, 2018).Song et al. (2015) suggest three categories of constructs which influence users' intention to adopt online technology.These are rational perception, behavioral control, and social influence.Our study adopts this model, which, in our opinion, better matches the characteristics of the context of African countries.

Rational Perception
Our study uses as its foundation the existing literature on technological innovations (Kamdjoug et al., 2021;Song et al., 2015) to determine the key rational perception factors in the intention to adopt e-commerce.We retained the following variables: utilitarian expectation (UE), hedonic expectation (HE), and status gains (SG).
Utilitarian expectation (UE) is the user's perception of the functional benefits of an online purchase plateform.Utilitarian value reflects the basic expectations of consumers about online shopping; it is also considered one of the main drivers of consumption (Batra & Ahtola, 1991) and technology adoption (Sullivan Mort & Drennan, 2007).We therefore hypothesize that utilitarian expectation will influence adoption intention: H1.Utilitarian expectation has a positive influence on intention to adopt online purchases.Hedonic expectation (HE) is -the extent to which making an online purchase is pleasurable for the user" (Alalwan et al., 2017;Song et al., 2015;Venkatesh et al., 2012).This concept is also known as hedonic value (Sullivan Mort & Drennan, 2007) or emotional value in western studies (De Marez et al., 2007;Nysveen et al., 2005).Pursuing pleasurable and joyful experiences is known to be one of the fundamental personal desires (Song et al., 2015), and hedonic expectation was further identified as a an important determinant of consumer choices.We can anticipate that hedonic expectations may affect the intention to adopt online purchases.Therefore, the following hypothesis is proposed: H2a.Hedonic expectation has a positive influence on intention to adopt online shopping.
Research on innovation adoption found that hedonic expectation influences utilitarian expectation (Hong and Tam, 2006;Starbuck and Webster, 1991).Hence the following hypothesis: H2b.Hedonic expectation has a positive effect on utilitarian expectation.
Status gains (SG).The diffusion of innovation theory emphasizes that the desire to acquire status is an essential reason for adopting innovation (Rogers, 2004).Past studies have reported that socially acceptable practices of innovation adoption improve the social image of individuals (Venkatesh et al., 2012).Therefore, perceived status gains are assumed to influence intention to adopt innovation: H3a.Status gains have a positive influence on intention to buy online.
Status gains is, therefore, instrumental in enhancing non-functional benefits to individuals.They can also lead to intrinsic satisfaction since different levels of pleasure are believed to arise from an improved social image (Bao et al., 2003).Therefore, status gains is a potential determinant of both utilitarian and hedonic expectations.Accordingly, the following hypotheses are formulated: H3b.There is a positive relationship between Status gains and utilitarian expectation.

Behavioral Control and Security
Empirical evidence suggests that the adoption and use of new technologies is a function of both individual (internal) factors and external factors.These factors usually play a pivotal role in a person's willingness to adopt a technology.Since online purchase is still at an early stage in developing countries, this study considers internal factors such as habit (HA), concern for quality (CQ), and privacy perception (PP).
Habit (HA) is a second nature acquired through learning and expressed by sustained behaviors over a long period of time (Baptista & Oliveira, 2015;Farah et al., 2018;Venkatesh et al., 2012).It facilitates the continual use of technologies (Baabdullah et al., 2019;Ramí rez-Correa et al., 2019).Repeated practice of online purchases can potentially increase the tendency to use it as the sole mode of shopping (Baabdullah et al., 2019;Baptista & Oliveira, 2015;Farah et al., 2018).We therefore zhypothesize that: H4.Habit has a positive influence on intention to buy online.Quality Concern (QC) is "the extent to which a person perceives a poor quality that prevents them from using a technology" (Krafft et al., 2017;Song et al., 2015).In other words, it is the power of poor quality to discourage technology use or adoption.Therefore, concern for quality must be considered to improve technology adoption and use in developing countries (Mehta et al., 2018;Song et al., 2015).Song et al. (2015) found that concern for quality is the primary motivation for adopting and using technology in China.Consequently, the following hypothesis is formulated:

H5. Concern for quality has a negative influence on intention to adopt online purchases.
Perceived privacy concern (PPC).It is the perceived risks associated with the online purchase; these risks include the uncertainty of the product quality, the privacy and security of personal information, and the safety of online transaction systems.Shaw & Sergueeva (2019) identify several elements that constitute privacy concerns for users, including perceived privacy concern.Therefore, we can formulate this hypothesis: H6.Perceived privacy concern negatively affects intention to adopt online shopping.

Intention to adopt (INAD).
Intention is considered the primary predictor of the actual behavior (Montaño & Kasprzyk, 2015).The TAM and UTAUT models consistently used the constructs of perceived rationality and behavioral control to predict the adoption and use of technology (Natarajan et al., 2017;Ramí rez-Correa et al., 2019;Song et al., 2015).Given its predictive role and theoretical heritage, this research considers intention to adopt online purchase as a relevant variable of interest.The construct occurs before the online purchase and constitutes the motivational factor that determines customer behavior (Armitage & Conner, 2001).For the purpose this research, intention to adopt is conceptualized as consumer willingness to purchase a product online.

Repurchase intention (RI)
. This is the intention declared by the consumer to repurchase the same brand or revisit the same retailer.Researchers seem to support measures of loyalty that incorporate both behavioral and attitudinal components.However, Lichtlé & Plichon (2008) suggest clearly distinguishing between purchase intentions and repurchase intention.Only intention to repeat purchases is a measure of behavioral loyalty.Oliver (1997) distinguishes four stages in the loyalty process: cognitive loyalty (based on brand beliefs), affective loyalty (attitude toward the brand), conative loyalty (behavioral intention), and fidelity action, which concerns the conversion of intentions into action, accompanied by a willingness to overcome obstacles to such action (Harris & Goode, 2004).Only the passage to the last stage (action loyalty) allows behavioral loyalty.However, the individual must first pass successively through the three previous stages.Loyalty occurs when favorable attitudes accompany repeated purchasing behavior (Assael, 1992).We therefore formulate the following hypothesis: H7. Intention to adopt online purchases positively influences intention to repurchase online.

Online Purchases and Cultural Variables
Hofstede ( 2001) understands culture as the collective programming of the mind that differentiates group members from others.Cultural variables affect consumers' motivations, attitudes toward choices, intentions, and behavior, and make it possible to predict beliefs, attitudes, and behaviors (Zhang & Jolibert, 2003).They are important in the African context, as in all societies, where customs and habits determine consumer behavior (Diop & Merunka, 2013).Indeed, the culture is omnipresent in consumer behavior in Africa (Bikanda & Mefoute Badiang, 2017;Hernandez, 2007;Kamdem & Mutabazi, 2017).Socializing in a market space or a shopping center is an integral part of consumer culture (Bikanda & Mefoute Badiang, 2017;Diallo et al., 2022).
Cultural variables in Africa fit into the concept of Ubuntu (Biwolé Fouda, 2020;Diallo et al., 2022).Ubuntu is the philosophy behind the Xhosa proverb "Umntu ngumntu ngabantu", which describes a typically African conception of the person; the proverb means that a person is a person through other people (Karsten & Illa, 2005).This concept is based on relationships, community, loyalty, and fundamental African characteristics (Lutz, 2009;Xing et al., 2014).This principle confirms the interpretations of Metz (2007) and the societal challenges of positive relationships with others for self-realization, solidarity with the group, and the role of humanity in the social world (Schwartz, 1994).These fundamental cultural principles allow us to retain three characteristics of cultural variables: community, conformity to the group, and physical interaction.
Community.The African context is characterized by a strong presence of collectivism rooted in traditions and communalism, which values time spent with the group (Diallo et al., 2022).Communalism is the choice to conform to a group rather than taking an individual position.In other words, it is a subordination of individual goals in favour of those of the group, emphasizing sharing and harmony, family, interdependence, and sociability (Diallo et al., 2022).These elements can impact how consumers perceive the use of the Internet for e-commerce in contexts where a collectivist culture is prevalent (Chau et al., 2002).This leads us to formulate the following hypotheses: H8a.There is a positive relationship between communalism and intention to adopt online shopping.H8b,c,d.The relationships between rational perception and consumer intention to purchase online decrease when consumers have a positive community orientation.

H8e. The relationship between consumer intention to adopt online shopping and repurchase intention decreases when consumers have a positive community orientation.
Group conformity (GC).According to social learning theory, an individual's behavior is greatly influenced by the behavior of reference group members with whom the individual associates.This variable corresponds to social influence in its normative component, and it is often mobilized in research on intention to adopt technological innovations (Song et al., 2015).Numerous studies have shown that conformity to the group is one of the significant factors in consumers' acceptance of innovations (Nasco et al., 2008).In Africa, Diop & Merunka (2013) reveal that consumption behaviors are more linked to societal conformity than to oneself.Consumers in Senegal, for instance, wear traditional attires to conform to social norms and gain respect or consideration from their peers.They thus express their conformity to the dominant social values and their belonging to the social group (Diop & Merunka, 2013).We therefore formulate the following hypothesis:

H9a. Conformity to the group positively influences intention to adopt online purchases.
Previous research indicates that group conformity could have a moderating effect on adoption of innovation (Fazio & Zanna, 1981;Spreng & Page, 2001), particularly on online purchases (Lee et al., 2006).Group conformity can be considered a moderating variable that strengthens the relationship between rational perception, behavioral control, confidentiality, and purchase intention, as well as the relationship between behavioral intention and repurchase intention.Hence, the following assumptions: H9b,c,d.The relationships between rational perception and consumer intention to purchase online diminish when consumers have a positive group conformity orientation.
H9e.The relationship between the consumer's intention to adopt online shopping and their intention to repurchase decreases when group conformity is positive.
Physical interaction (PI).Researchers have found that customers have a preference for offline channels (Avery et al., 2012).Indeed, offline channels (e.g.physical stores) offer consumers several opportunities, including the ability to touch merchandise, talk to sellers face to face, and get advice, which helps customers to reduce the perceived risk associated with their choices.Despite their advantages, online channels come with potential risks, including privacy issues and a lack of personal assistance in resolving issues or conflicting situations.These risks generally discourage customers who are very sensitive to them.In Africa, the distribution channel is more than a space for purchasing goods or services; it is a place of meeting, exchange, familiarity, and civility.It is a space where social cohesion is formed and perpetuated.Interaction with the product, punctuated by verbal exchanges with the seller on tastes, fashion, family and friends, news, and jokes, goes beyond the object to find affinity links and reassure the consumer (Bikanda & Mefoute Badiang, 2017;Diop & Merunka, 2013).Physical contact with the product or the seller thus makes it possible to mitigate the risk associated with the purchase.Hugon (1999) evokes a survival strategy among consumers in Africa, where the weight of everyday life forges a very strong preference for the present.Instability, low life expectancy, and precariousness induce a preference for immediacy.In such a perspective, immediate property acquisition reduces uncertainty (Bhatnagar et al., 2000;Dachyar & Banjarnahor, 2017).We therefore formulate the following hypotheses: H10a.Absence of interaction negatively influences the consumer's intention to buy online.H10b, c, d.The relationships between rational perception and the consumer's intention to purchase online decrease when interaction with the product is low.

H10d. The relationship between the consumer's intention to adopt online purchases and their intention to repurchase decreases when interaction with the product is low.
The research model (see Fig. 1) depicts the hypotheses formulated from the relationships between the variables in the study.

Sampling and Data Collection
Data was collected through a questionnaire that included three main sections: (1) filter and general questions; (2) measurement of constructs, and (3) sociodemographic variables.The target population included all consumers who had ever made an online purchase.In the absence of a sampling frame for this population, convenience sampling was used to accumulate a sample (San Martí n & Herrero, 2012).The survey instrument was created with Microsoft Word.It was subsequently formatted into Google Forms, enabling us to generate an electronic link.The link generated was sent online (by email and whatsApp) to approximately 600 respondents so that they could complete the questionnaire through a suitable connected device.Putting the questionnaire on Google Forms enabled to reach as many users as possible, to carry out several checks, and to obtain answers ready to be analyzed in the form of a CSV extension in an Excel file.A total of 466 completed questionnaires remained after removing those that were incomplete or contained errors.All entrants were at least 18 years old and wholly or partially responsible for online product purchases.

Constructs
Based on the existing literature, scales with good psychometric properties were used whenever possible (see Table 1).The questionnaire was designed in line with the research model, and it consisted of 19 constructs and 84 items.To measure our items, we used a Likert scale with seven levels (1 = Totally disagree, 2 = Disagree, 3 = Slightly disagree, 4 = Neither agree nor disagree, 5 = Slightly agree, 6 = Agree, 7 = Totally agree).This scale is commonly used to measure attitude (Kavota et al., 2020;Wanko et al., 2019).All metrics were taken from previous research and adapted to the online shopping context.During my online shopping, I fear that my banking details will be stolen PPC3 I am concerned about the amount of information provided on this/these online retailer(s) PPC4 I am concerned about the amount of information provided on this/these online retailer(s) because it could be used for dishonorable purposes PPC5 When using this/these online retailer(s), I have the feeling that my actions are being spied on and traced.

PPC6
When using this/these online retailer(s), I feel I am being observed Quality concern (QC): α = 0.976; CR = 0.983; AVE = 0.934 QC1 When using this/these website(s), I feel the company collects a lot of my personal information When using this/these online retailer(s), I fear the company could use my personal information for other purposes than those for which I provided the information QC3 When using this/these online retailer(s), I fear the company will share my personal information with third parties QC4 When using this/these online retailer(s

GC2
People who have an influence on my behavior think that I should use these online sales websites during my shopping GC3 People whose opinion I value would like that I do my shopping through these online sales websites GC4 I often ask for advice from people to help me choose from the best online sales websites GC5 To look like certain people, I often try to buy the same brands that they buy GC6 I often identity with other people by buying from the same online retailer(s) that they buy from, or the same products and brands that they buy Physical interaction (PI): α = 0.931; CR = 0.942; AVE = 0.732 PI1 It is difficult to judge the quality of a product without touching it Bikanda & Mefoute Badiang (2017) PI2 I prefer bargaining with the seller before buying PI3 I like having my product immediately after buying PI4 I intend to be sure that the product meets my expectations

Hypothesis Testing
After data collection, the data was cleaned, screened, and prepared for analysis.Since the conceptual model includes 12 conditional effects (moderations), a partial least squares approach of structural equation modeling (SEM-SmartPLS 4) was adequate to test the model (Chaouali & Souiden, 2019;Koay et al., 2020).The SEM method includes of two phases: confirmatory factor analysis (CFA) and structural model analysis (Gallagher and Brown, 2013).These steps are discussed next.

Validity of the Measurements
CFA assesses the validity of measurements in terms of reliability, convergent validity, and discriminant validity.A summary of the reliability and convergent validity results is displayed in Table 1.The Cronbach's alpha coefficients, the composite reliability, and the average variance extracted (AVE) are greater than their required thresholds, namely 0.7 for reliability and 0.5 for AVE (Malhotra et al., 2017).The results (see Table 1) show that the items retained in the final model are reliable and have convergent validity. .Meaning, the items considered in this study adequately measures of their respective constructs.Regarding discriminant validity, the Fornell-Larcker criterion (Fornell & Larcker, 1981) and the heterotrait-to-monotrait correlation ratio are evaluated.Fornell & Larcker (1981) assess discriminant validity by comparing the highest pair of correlations with the square root of the AVE of each construct (Malhotra et al., 2017).Correlations less than the square root of the AVE suggest high discriminant validity between the constructs, as is the case in this study (see Table 3).(Hair et al., 2020;Henseler et al., 2015;Koay et al., 2020).4), all coefficients are less than the standard threshold of 0.85 (Hair et al., 2020;Koay et al., 2020;Ramí rez-Correa et al., 2019).Notes: CO = collectivism, GC = group conformity, HE = hedonic expectation, HA = habit, INAD = intention to adopt, LOY = repeat purchase intention, PPC = perceived privacy concern, SG = status gains, QC = quality concern, UE = utilitarian expectation, PI = physical interaction.

Structural Model Analysis
Since the aim of the PLS-SEM method is to appraise the predictive power of the research model (Hair et al., 2020;Koay et al., 2020), the coefficient of determination (R² ) was provided for all dependent variables in the model.Results show (see Table 5) that all dependent variables have a coefficient of determination higher than 0.25, except utilitarian expectation (UE).This low explanatory power may be because the variable of status gains (SG) alone is not enough to entirely explain the utilitarian expectation (UE) of individuals using online purchases.Furthermore, with an R² of 0.532, we can conclude that all proposed predictors selected explain intention to adopt online shopping well.Online repeat purchase intention (R² = 0.516) was also well explained by intention to adopt online shopping.Therefore, it is evident from the results that the power of the research model is good, as all coefficient of determination thresholds recommended by Hair Jr. et al. (2020) were met (see Table 5).As recommended by Hair et al. (2020), Stone-Geisser's Q² and Koay et al.'s (2020) values were evaluated in our study.The aforementioned coefficient was used to cross-validate the significance of the partial least squares path model.The findings presented in Table 5 indicate that our research model possesses strong predictive relevance, as evidenced by the Q² values exceeding zero.(Hair et al., 2020;Koay et al., 2020).The has moderate predictive power on utilitarian expectation, as R 2 is close to 25% (24.3%), and Q 2 (0.065) is greater than 0 Notes: HE = hedonic expectation, INAD = intention to adopt, LOY = repeat purchase intention, UE = utilitarian expectation.
However, the variables related to behavioral control (HA and QC) and terminal security (PPC) have no influence on intention to adopt online purchases (p ≥ 0.05).Thus, hypotheses 4 to 6 are not supported.The results also suggest that intention to adopt online purchases directly influences repeat purchase intention to purchase online (β = 10.77;P < 0.001).Therefore, hypothesis 7 is supported.
We hypothesized a positive moderating effect of collectivism on the relationship between rational perception variables and intention to adopt online purchases and between intention to adopt online purchases and repeat purchase intention to purchase online.As with some online purchase studies (Jaradat & Al Rababaa, 2013;Ramzy & Eldahan, 2016), no moderating effect was found.Thus, hypotheses 8a to 8e are not supported.
Regarding the moderating role of group conformity, the results demonstrate that group conformity strengthened the impact of utilitarian expectation on intention to make online purchases (β = 0.012; P < 0.05).This result supports the 9b hypothesis.The prevalence of group comparisons (Festinger, 1954) causes consumers to continually compare their views to those of others.According to the findings, this relationship is stronger and more positive among consumers with low group conformity than among consumers with high group conformity.
In other words, consumers with low group conformity are more likely than those with high group conformity to make online purchases because of the utilitarian benefits.According to Fig. 3, the relationship between utilitarian expectation and intention to make online purchases depends on the level of group conformity.According to the findings, the effect of intention to adopt online purchases on repeat purchase intention is positive and stronger among consumers with low group conformity than among consumers with high group conformity.In other words, consumers with low group conformity are more likely than consumers with high group conformity to see their repeat purchases increase more rapidly as a result of their decision to adopt online purchases.This result supports hypothesis 9e.According to Fig. 4, group conformity determines the strength and the direction of the relationship between intention to adopt online purchases and repeat purchase intention.We found a significant interaction effect on the relationship between hedonic expectation and intention to adopt online purchases (β = 0.012; P < 0.05), which implies that being exposed to experiences by other individuals during purchases strengthens the relationship between these variables.This outcome supports hypothesis 10b.The relationship between physical interaction, hedonic expectation, and intention to adopt online purchases is illustrated in Figure 5.

Discussion
As we have postulated, the constructs of rational perception (UE, HE, SG) significantly influence intention to adopt online purchases.Consistent with theories of social behavior, positive perceptions are a powerful force guiding the actions of individuals (Bandura, 1977).These results are in line with previous studies that have adopted IT/IS models (Baabdullah et al., 2019;Venkatesh et al., 2012).The regression weights of the three constructs, as indicated by the β values, suggest that utilitarian expectation holds the greatest sway over consumer adoption intention.Status gains positively affect the user's utilitarian expectation and hedonic expectation.Thus, people's behavior will be positively affected when they think that purchasing online improves their status or is actively promoted by other users (Song et al., 2015).However, the relationships between the constructs of behavioral control (HA) and terminal security (PPC) and intention to adopt online shopping are not significant, contrary to previous work (Song et al., 2015).Admittedly, users remain sensitive to notions of security related to privacy, risk of loss, and fraud.However, the online shopping model applied by local retailers allows consumers to pay on delivery, and it helps mitigate this type of risk (Baabdullah et al., 2019;Matemba & Li, 2018;Song et al., 2015;Venkatesh et al., 2012).
The impact of cultural variables on online purchases is one of the most interesting results of this research.The impact of collectivism on online shopping was not found to be significant (P ≥ 0.1).This could be attributed to the fact that the transactions carried out through this channel do not correspond to an activity conducive to socializing or sacrifice for group members (Bandura, 1977).By contrast, consumers who value physical interaction with products and sellers are more likely to shop online, but primarily for hedonic reasons (P < 0.05).
The findings show that for respondents who do not value physical interaction (low scores), their intention to make online purchases does not increase as their hedonic expectation increases.By contrast, those who value physical interaction (high scores) have a higher intention to make online purchases, which tends to increase with their hedonic expectation.Simply put, consumers who value physical interaction are more likely to make online purchases for hedonic reasons than consumers who do not value physical interaction.Regarding utilitarian expectation, the relationship between it and online purchase intention depends on the level of conformity to the group (P < 0.05).This relationship is stronger and more positive among consumers with low group conformity than among consumers with high group conformity.Group conformity can lead individuals to align their choices with those of others to construct or enhance their desired social identity (Chan et al., 2012).Consumers with low group conformity are more likely than those with high group conformity to shop online due to utilitarian benefits.

Implications and Limitations
This research provides a better understanding of the determinants of intention to adopt online purchases in an African context.It empirically examines the role of cultural variables, including collectivism, group conformity, and physical interaction.Following the preliminary review of studies on adoption of technologies, we employed a research model comprising three emergent constructs (rational perception, behavioral control, and safety) to formulate our research hypotheses.

Implications
On a theoretical level, this study contributes to the literature in two ways.Firstly, we extend the IS/IT model by providing a comprehensive theoretical explanation of the role of cultural values in determining consumers' behavioral intention toward use of online shopping services.Secondly, this research contributes to cross-cultural studies, and it is the first study that empirically validates the moderating effects of characteristically African cultural variables on the relationships between online shopping adoption and its determinants.Thirdly, the results of this research show an orientation toward traditional cultural values of African consumers in the online shopping process.This orientation is based on conformity to the group, as well as physical interaction with the product during the transaction.This research confirms previous studies, notably Diop & Merunka (2013), which found that both global and local consumer cultures influence African consumers.Hence the need to overcome this opposition between values of "modernity" and traditional values, which has been highlighted in previous studies (Diallo et al., 2022).We advocate for scholars to consider both forms of values, as, despite their strong anchoring in traditional values, consumers in Cameroon seem just as attached to modernity, particularly to technological innovations.Our results also complement previous studies conducted in other contexts (Diallo et al., 2015;Mai & Smith., 2012), and they indicate how traditional values influence consumer behavior in countries that welcome innovative western concepts.
Our findings have significant implications for future marketing strategies and campaigns of international and national companies operating in Cameroon.Firstly, companies must ensure that they position their offerings to account for important influencing factors.Secondly, leaders and managers involved in e-commerce need to understand better how collectivistic values may diverge from western values.Insufficient understanding of culture can become an obstacle to understanding social relations, leading to business failure.Taking into account a local cultural aspect could, in the eyes of consumers, be more important than the prices of products or other benefits (Ramzy & Eldahan, 2016;Santos, 2003).

Limitations
Our research has some limitations, which need to be borne in mind when interpreting the results.Firstly, use of an online survey could have excluded a segment of the population that is not online or computer literate, such as elderly people.This considerably restricts our target population and could hinder generalization of our results.Although this limitation, the results of the study cannot be dismissed, as users of online shopping are generally young and educated (Hong & Tam, 2006).Secondly, the study was conducted among people who have already made an online purchase, and, as such, it does not provide information about potential users who have not yet made an online purchase.Thirdly, the scales used for measurement of cultural variables were drawn from other contexts.Although those contexts are similar to the African context, we believe that measurement scales validated in the African context would be more appropriate.
The journal's policies adhere to the Core Practices established by the Committee on Publication Ethics (COPE).

Provenance and peer review
Not commissioned; externally double-blind peer reviewed.

Figure
Figure 1.The conceptual model 3. Methodology /these online retailer(s) during my online shopping is entertainingKamdjoug et al. (2021) andSong et al. (2015) HE2Use of this/these online retailer(s) during my online shopping is easy and niceHE3Use of this/these online retailer(s) during my online shopping gives me pleasure HE4Use of this/these online retailer(s) during my online shopping is amusing HE5Use of this/these online retailer(s) during my online shopping makes me feel good Intention to adopt (INAD): α = 0.954; CR = 0.964; AVE = 0.844 INAD1 I will readily browse this/these online retailer(s) INAD2I will encourage my friends to do their shopping through this/these online retailer(s)INAD3I am constantly watching out for special offers from this/these online retailer(s)INAD4In the future, I will continue doing my shopping through this/these online retailer(s)INAD5I will take advantage of more deals with this/these online retailer(s) in the coming years Kamdjoug et al. (2021) and Song et al. (2015) Repeat purchase intention (LOY): α = 0.934; CR = 0.958; AVE = 0.883 LOY1 I would recommend use of this/these online retailer(s) to people LOY2 I intend to continue using this/these online retailer(s) for my shopping Kamdjoug et al. (2021) and Song et al. (2015) LOY3 I prefer using this/these online retailer(s) to other shopping channels Perceived privacy concern (PPC): α = 0.976; CR = 0.980; AVE = 0.891 PPC1 I fear that the information I provide via this/these online retailer(s) will be used for other purposes Kamdjoug et al. (2021) and Song et al. (2015) PPC2 Kamdjoug et al. (2021),Krafft et al. (2017)andSong et al. (2015  QC2 Notes: CO = collectivism, GC = group conformity, HE = hedonic expectation, HA = habit, INAD = intention to adopt, LOY = repeat purchase intention, PPC = perceived privacy concern, SG = status gains, QC = quality concern, UE = utilitarian expectation, PI = physical interaction.Hair et al. (2020), Koay et al. (2020), and Ramí rez-Correa et al. (2019) further recommend assessing the homogeneity of the research model through the heterotrait-monotrait (HTMT) ratio of correlations.According to the results (see Table

Figure 3 .
Figure 3. Moderating effect of group conformity on the relationship between utilitarian expectation and intention to adopt online purchases

Figure 4 .
Figure 4. Moderating effect of group conformity on the relationship between intention to adopt online purchases and repeat purchase intention

Figure 5 .
Figure 5. Moderating effect of physical interaction

Table 1 .
Constructs and measurement items

): α = 0.971; CR = 0.976; AVE = 0.872 CO1
Yoo & Donthu (2002)crifice their personal interest for that of the group CO2 Individuals must remain with the group even in hard timesCO3The well-being of the group is more important than individual success CO4 Individuals must pursue their personal objectives only after they have considered the well-being of the group CO5Faithfulness to the group must be encouraged even if personal objectives must beYoo & Donthu (2002)and Yoo et al.

Table 2 .
Table2presents the profile of the interviewees.The data indicates a balanced sample, and it more or less reflects Cameroon's population that shops online.Descriptive results of the demographics indicate that most of the respondents were aged 18 to 35 years (50.7%), have attended secondary school or a first-or second-cycle university (62.8%), and earn less than $700 (87.2%) a month.Demographics of the sample

Table 3 .
Assessment of discriminant validity using the Fornell-Larcker criterion

Table 5 .
R squared and the predictive relevance of the endogenous variables