Abstract
This study investigates the impact of review quality (a situational stimulus) on consumers' risk perception and purchase intention in cross-border e-commerce based on the Stimulus-Organism-Response (SOR) model. In doing so, quantitative research involving 400 Hungarian respondents was performed. The data were analysed using composite-based structural equation modelling (SEM). The study concludes that an experience created through highly qualified online reviews of previous consumers has a significant effect on mitigating consumers' risk perception while increasing their purchase intentions. The study also differentiates two aspects of risk, including perceived risk and affective risk, and reveals the two-fold mechanism of the decision-making journey. These results enrich the existing literature by supporting the use of the SOR model and introducing review quality as a situational stimulus to explain consumers' risk perception and purchase behaviours in cross-border e-commerce. Additionally, the study also provides valuable guidance in website design that can stimulate purchasing while lowering online perceived risk.
1 Introduction
Cross-border e-commerce is defined as online transactions between sellers and customers located in different countries to exchange products and services via e-commerce platforms (Hsiao et al. 2017). Accordingly, the significance of this type of e-commerce is to enable global traders and consumers to engage in cross-border online transactions. The share of cross-border e-commerce in total global e-commerce obtains 22% in 2022 (Statista 2023a). Although cross-border e-commerce contributes to promoting business opportunities, it is obvious that this type of e-commerce is far more complicated and risky compared to domestic e-commerce (Guo et al. 2018). Notably, cross-border e-commerce is not widely preferred in several markets, such as Europe. According to the report by CMI (2023), most European shoppers prefer purchasing from domestic sellers (81%). In Hungary, the majority of online shoppers also prefer Hungarian web shops. Only around 37% of total Hungarian e-shoppers purchase in cross-border e-commerce. Therefore, inquiring about the perception and behaviour of Hungarian e-consumers in cross-border e-commerce may be significant.
Current studies have provided various results in perceived risk and consumer behaviour. For instance, Daragmeh et al. (2021) show that perceived risk diminishes Hungarian consumers' intentions to adopt a mobile payment system. According to Gáti and Simay (2019), perceived risk negatively impacts attitudes and intentions towards social mobile e-commerce in Hungary. However, limited studies apply the Stimulus-Organism-Response (SOR) theoretical model to explain the risk perception and behaviour of the consumer. There are two reasons why the SOR paradigm is preferred in this study. First, the SOR framework is a valuable theoretical perspective that enhances descriptive analysis and provides a critical lens through which to comprehend individual behaviour (Zhu et al. 2020). Second, the SOR framework enables an understanding of consumers' risk perception and purchase intention in cross-border e-commerce through various path-dependent mechanisms by which a stimulus impacts individuals' cognition and behaviours. Additionally, studies by the majority of scholars demonstrate that marketing factors (Chafidon et al. 2022; He et al. 2022), internal psychographic factors (Mohseni et al. 2018; Nilashi et al. 2022), and website factors (Pham et al. 2022) are major stimuli driving consumers' perceptions and behaviours. The aspect of situational factors rooted in society has not been frequently examined. Therefore, this study aims to address the following research questions:
RQ1
Does risk perception (perceived risk and affective risk) impact Hungarian consumers' purchase intentions in cross-border e-commerce?
RQ2
Does review quality impact Hungarian consumers' risk perception and purchase intention in cross-border e-commerce according to the SOR model?
This study focuses on two major issues. First, it emphasises the significant impact of perceived risk on consumers' purchase intentions in cross-border e-commerce. Additionally, it reveals two aspects of risk, including perceived risk and affective risk. Second, it investigates the effects of review quality as a situational stimulus on consumers' risk perceptions (organism) and behaviours (response) in cross-border e-commerce. Based on this, companies can enhance purchase intention by incorporating effective social interaction tools and relevant IT software into their platforms.
2 Literature review and hypothesis development
2.1 The stimuli-organism-response (SOR) model
The SOR model posits that the stimuli exert an impact on the organism, often affecting the cognitive and affective processes of a consumer, ultimately leading to their behavioural response (Mehrabian – Russell 1980). The SOR model is a viable theoretical framework for addressing consumer behaviour in online environments (Zhu et al. 2020). According to Kimiagari and Asadi Malafe (2021), two stimuli impact consumers' behaviour, external and internal. External stimuli include factors beyond the customer's control (e.g., the product, sellers, society, and website features). In contrast, internal stimuli are rooted in consumers, such as demographics and psychographic characteristics. The existing body of literature has used the SOR model to examine the impact of environmental stimuli on consumers' perceived risk and purchase intentions in the online context. For instance, Chang et al. (2019) found that the quality of a branded website mitigates consumers' perceived risks while improving consumers' trust and purchase intention. Further, perceived risk directly lowers consumers' trust in e-commerce. According to He et al. (2022), product authenticity diminishes consumers' risk perception while increasing their purchase intention. The research by Kim and Lennon (2013) reveals two types of stimuli influencing consumers' perceived risk and positive emotions, including website quality and the reputation of the brand. Generally, previous studies have mainly focused on investigating stimuli associated with websites, products, sellers, and consumers' psychographic characteristics. The aspects of the situational factor have not been frequently investigated.
2.2 Two aspects of risk and their impacts on purchase intention
The theories of decision under risk and uncertainty reveal that risk perception is understood in two fundamental models, including risk-as-analysis (i.e., a logical and cognitive reaction to risk) and risk-as-feeling (i.e., intuitive reactions to danger) (Slovic et al. 2004; Slovic – Peters 2006). However, current consumer behaviour literature in marketing mostly highlights the significance of risk perception as a cognitive process that entails analytical, rational, reason-based, and deliberative forms of information processing. Accordingly, perceived risk is the potential loss that a consumer perceives when making an online purchase (Ariffin et al. 2018). Limited attention was given to the aspect of emotion, also called affective risk, although this aspect has been identified in several theories of decision-making under risk and uncertainty. The affect-as-information model suggests that people will interpret their situations differently, relying on positive or negative feelings (Tompkins et al. 2018). Prospect theory explains the asymmetries in consumer behaviour between perceived losses and gains, elucidating the role of emotions and risk perception (Chiu et al. 2014). It highlights the more significant negative emotional impact of losses compared to the positive emotional impact of equivalent gains (Abdellaoui et al. 2007).
2.2.1 The aspect of the perceived risk
The current literature reveals two basic research streams on perceived risk, including uni-dimension and multi-dimension. The former approach has been applied frequently in various studies, such as Alimamy and Gnoth (2022) and Malak et al. (2021). Accordingly, perceived risk is considered a single conceptual construct without identifying any specific risk types. By contrast, the multidimensional theory based on the studies of Jacoby and Kaplan (1972) suggests that consumers perceive several types of risk in purchasing situations. On a conceptual level, these dimensions can be considered functionally independent. Therefore, each type of risk can reflect a different aspect of perceived risk. Although conceptualising perceived risk as a multi-dimensional construct has been preferred by several scholars, they have mainly measured these dimensions as lower-lever variables. This implies that relationships between sub-dimensions of perceived risk are independently investigated. The failure to consider these dimensions as indicators has raised the question concerning the construct clarity of perceived risk in e-commerce.
The efficacy of a high-order construct is predicated on a combination of theoretical and empirical factors. Scholars showed that higher-order constructs allow for more theoretical parsimony and reduce model complexity (Wetzels et al. 2009). Prior studies have modelled perceived risk as a second-order superordinate construct with effect indicators, such as studies by Pillai et al. (2022), Qalati et al. (2021) and Rosillo-Díaz et al. (2019). However, according to Johnson et al. (2011), indicators of superordinate constructs share more overlap than those of aggregate constructs, which affects internal consistency among indicators.
According to Yang et al. (2016), perceived risk in e-commerce derives from consumers' concerns about potential financial risk and product risk. Financial risk refers to the possibility of losing money or incurring additional charges when engaging in e-commerce purchases. Product risk is the potential loss incurred when goods do not perform as expected. From the principal-agent perspective, the potential fraud risk in e-commerce may arise from the scepticism of online shoppers regarding the potential impact of undisclosed information and undisclosed actions in transactions. Accordingly, fraud risk refers to the possibility of a seller's unreliability in e-commerce (Pavlou et al. 2007). Furthermore, perceived risk in e-commerce is also associated with the processes of information dissemination and delivering purchased items. Customers are exposed to several information risks, such as a lack of information and information disorganisation, which increases the level of perceived risk. Therefore, information risk refers to the possibility of asymmetric information about sellers and products. Additionally, the issues relative to delivery and logistics of the purchased products also increase consumers' perceived risk. The delivery risk is conceived as the possibility of not receiving the product on time (Ariffin et al. 2018). According to Pentz et al. (2020), time loss and process inconvenience are also the sources of online consumers' perceived risk. Process inconvenience is understood as the possibility of a consumer's inconvenience exceeding the expected process of online shopping. Zhang et al. (2012: 3) defined privacy risk as “the potential loss of control over personal information when the information is used without permission”. Hereby, this perceived risk derives from the consumers' concerns that an e-commerce website may abuse their financial and personal information and disclose it to a third party or that the e-commerce website fails to safeguard their information from cybersecurity attacks (Alrawad et al. 2023).
Consequently, this study synthesises seven potential dimensions and considers perceived risk as a second-order composite with causal indicators. The study hypothesises:
H1
Perceived risk is a higher-order composite formulated by fraud risk, delivery risk, financial risk, product risk, process & time loss risk, privacy risk, and information risk.
Previous studies revealed the significant impact of perceived risk in the pre-purchase stage. Hereby, many scholars found that perceived risk lowered consumers' purchase intention in e-commerce (He et al. 2022; Pratesi et al. 2021). Online purchase intention is defined as the desire of customers to acquire a product or service on an e-commerce platform (Meskaran et al. 2013). According to Pratesi et al. (2021), perceived risk negatively relates to buying intention in online shopping. Therefore, the study hypothesises:
H2
Perceived risk negatively impacts purchase intention.
2.2.2 The aspect of emotion – affective risk
As mentioned previously, the conventional perspective on risk within the marketing and consumer behaviour domains revolves around assessing the probability of adverse outcomes. However, it is contrasted to a normal individual's judgement of risk. According to Finucane et al. (2000), an emotion often has a significant impact, as it acts as a cognitive signal for the formation of judgements. The “risk-as-feelings” hypothesis also revealed that cognitive risk perception and anticipatory emotion functioned as antecedents to an individual's behavioural intention (Loewenstein et al. 2001). Consequently, perceived risk and affective risk are distinct. Perceived risk captures the cognitive beliefs or calculations involved in a risky event, while affective risk captures a person's emotions. Affective risk may be immediate because it requires less brain processing, suggesting less cognitive calculation (Sha 2018). In this study, affective risk refers to immediate anticipatory emotions (e.g., anxious, afraid, and worried) that individuals experience when purchasing online (Loewenstein et al. 2001).
Éthier et al. (2006) revealed that consumers' risk perception can generate emotions such as fear or frustration during the online shopping experience. Furthermore, perceived risks, such as monetary loss and privacy invasion, will generate affective risk and prevent consumers from pursuing further business transactions (Sha 2018). Therefore, the study postulates that:
H3
Perceived risk positively impacts affective risk.
Furthermore, negative emotions such as anxiety, depression, and loneliness can directly influence consumers' purchase behaviour in e-commerce (Luo et al. 2018). Affective risk is found to prevent consumers from checking out their shopping carts or drive them to switch intentions (Sha 2017; Verhagen – Dolen 2011). According to Li et al. (2011), the initial emotions of consumers form their overall impression of a new e-commerce vendor and influence their behaviour regarding online information disclosure. Specifically, initial negative emotions, such as fear, can alert consumers to potential issues. Therefore, the study postulates that:
H4
Affective risk is negatively related to purchase intention.
2.3 Role of review quality in e-commerce
2.3.1 Social interaction and consumer behaviour
Social interaction refers to a reciprocal exchange between two or more individuals (Lho et al. 2022). Accordingly, consumers gain knowledge, make informed decisions, and reduce their perceived risk when making purchases via interactions with their communities (Lee – Bell 2013). According to Wang and Yu (2017), social interactions are an essential precursor to purchasing behaviour. Consumers perceive the data obtained through online social interactions as more reliable and authentic than traditional marketing channels (Park et al. 2007). Social interactions are classified into two primary groups: opinion-based and behaviour-based social interactions, in which online reviews are considered an aspect of opinion-based social interaction (Chen et al. 2011). According to Zhang et al. (2010), the quality of reviews is essential in shaping consumers' acceptance of online reviews. Consumers value reviews because they help them assess the quality of products more accurately (Wang – Chang 2013). Scholars agree that high-quality reviews may influence customers' purchasing decisions and that customers are more likely to trust reviews from reliable sources (Zhang et al. 2014).
2.3.2 Review quality, risk, and purchase intention
Review quality is defined as the extent to which consumers find that online reviews accurately reflect users' experiences with products (Zhang et al. 2014). The quality of information available on a product is essential in e-commerce (Kollmann – Suckow 2012). Previous studies found that buyers generally perceive online product reviews as credible and valuable for making purchasing decisions (Wang 2010). Scholars suggested that e-consumers prefer to acquire product reviews from other consumers to evaluate product quality, decrease risk, feel safe about the website (Xu et al. 2020), and motivate themselves to purchase products online (Zhang et al. 2014). Therefore, the study postulates that:
H5a
Review quality is negatively related to perceived risks.
H5b
Review quality is negatively related to affective risk.
H5c
Review quality is positively related to purchase intention.
The research model is presented in Fig. 1.
3 Methodology
3.1 Research design
The study conducts quantitative research using an online survey. The target population was adult online shoppers in Hungary. Despite a small and less-noticed e-commercial market, the Hungarian e-commercial growth rate is significant. Furthermore, around 74.3% of Hungarians are online shoppers, which is approximately 7.4 million people in 2022 (Statista 2023b). However, as mentioned previously, only a minority of Hungarian online shoppers prefer buying from cross-border sellers (CMI 2023). Therefore, selecting Hungary as a research site may be meaningful. Accordingly, the survey was designed using a purposive sampling technique to collect the primary data. The respondents' selection criteria were: first, they must be Hungarian adults, and second, they should have personal experience with purchases on an e-commerce website.
n = sample size,
P = proportion of the sample
MOE = margin of error, a 5% margin of error is usually selected
N = Population size
= The critical value of the normal distribution at a (for a confidence interval level of 95%, is 0.05 and the critical value is 1.96)
According to Etikan (2019), the sample size does not change much for a population larger than 20,000. Consequently, based on the calculation, a minimum sample size of 385 was necessary for this study.
The survey contains three major tasks. In task 1, the participants were required to answer questions associated with their demographic information. In task 2, the participants were required to imagine that they would buy a pair of shoes as a birthday gift for their partners. They were required to visit the following website: https://knight.mysapo.net/ (see Appendix 1). During browsing, the participants were also required to read online reviews prepared in advance. After browsing the website, the participants would select items that they wanted to purchase, put the items in their carts and start placing the order. In task 3, respondents were asked to indicate their level of agreement or disagreement with items on a 5-point Likert scale ranging from 1 “strongly disagree” to 5 “strongly agree”.
To measure perceived risk, 7 sub-dimensions adapted from Naiyi (2004) were applied. Affective risk was measured by the scale adapted from Petrova et al. (2023) and Sha (2018). Purchase intention was measured by the scale from Zhang et al. (2012). Review quality was measured by the scale from Park et al. (2007) (see Appendix 2).
The data was analysed using Structural Equation Modelling (SEM). This study applied the variance-based SEM (PLS-SEM) technique for a number of reasons. It supports achieving a high level of statistical power even with small sample sizes. PLS-SEM can also handle non-normally distributed data. It is significant, as the data collected for this study has a non-normal distribution. Furthermore, PLS-SEM can work with a complex model, including formative and reflective constructs simultaneously (Afthanorhan et al. 2020). The study applied the ADANCO software, as this software is equipped with a composite-based SEM technique. Additionally, it corrects for attenuation when modelling higher-order constructs (i.e., perceived risk). Further, to ensure the appropriation of the sample size for SEM, the authors applied an A-priori Sample Size Calculator for SEM by Soper (2023) to calculate the minimum sample size to detect an effect. The results showed that a minimum sample size of 190 is recommended to run the SEM model.
3.2 Data collection and survey participant
The survey was developed on the Survey Monkey platform. After the survey items were set and activated online, the link was sent to respondents across Hungary. 963 respondents were invited, and 400 valid responses were selected for the analysis. The response rate is 41.5%. The number of respondents is relevant because it exceeds the minimum sample size required.
Table 1 indicates the demographic characteristics of respondents. Regarding gender distribution, the ratio of males and females is 50:50. With age distribution, the authors focused on collecting data in three age groups, including (1) 18–25, (2) 26–35, and (3) above 36 corresponding to the ratios of 35.00%, 38.50%, and 26.50%. These age groups are appropriate for this study's purpose because the majority of Hungarians prefer online shopping due to its convenience. Since the study seeks to investigate online adult consumers' risk perception and purchase intention, it was prudent to sample online consumers who were 18 years of age and older. The percentage of students and working people is equal. The employment status was selected as a demographic characteristic, as the authors believe that financial independence may encourage an individual to purchase online. Regarding expenditure for online shopping, approximately 64.25% of the total sample spends under 100 USD per month on online shopping, whereas 19.75% spend from 100 to 500 USD monthly, and 16% spend over 500 USD for online shopping monthly.
Demographic characteristics of respondents
| Number | Percentage (%) | ||
| Sample Size | 400 | 100.00 | |
| Gender | Male | 200 | 50.00 |
| Female | 200 | 50.00 | |
| Age groups | 18–25 | 140 | 35.00 |
| 26–35 | 154 | 38.50 | |
| 36 | 106 | 26.50 | |
| Employment status | Student | 200 | 50.00 |
| Working | 200 | 50.00 | |
| Monthly expenditure on online shopping | <100 USD | 257 | 64.25 |
| 100-500 USD | 79 | 19.75 | |
| >500 USD | 64 | 16.00 | |
Source: authors.
4 Data analysis
4.1 Measurement model analysis
The measurement model was analysed using the CFA analysis, in which the authors estimate the model with a saturated structural model, and all constructs are freely correlated. The test for overall model fit is acceptable (dULS = 0.4219, P > 0.01). Furthermore, the SRMR is 0.0259, below the recommended threshold of 0.08 (Hu – Bentler 1999), indicating a good model fit.
The indicator, construct reliability, and convergent validity are presented in Table 2. Regarding reflective constructs, indicator loadings are higher than 0.7, indicating good reliability. Additionally, Dijkstra-Henseler's rho (ρA) and Cronbach's alpha (α) of all reflective variables are also greater than 0.7, indicating good constructs' reliability (Hair et al. 2017; Henseler 2021). To establish convergent validity, the average variance extracted (AVE) for each construct should be larger than 0.5 (Hair et al. 2017). Furthermore, the variance inflation factors (VIF) are also below the threshold of 5. The results show that the model achieves convergent validity.
Results for the assessment of reflective measurement and composite models
| Construct | Items | Types | Loadings/ Weights | Cronbach's Alpha | rho A | AVE | VIF |
| Review quality | Reflective | 0.8966 | 0.899 | 0.6845 | |||
| RQ1 | 0.7969 | 2.3254 | |||||
| RQ2 | 0.8861 | 2.5733 | |||||
| RQ3 | 0.7723 | 2.742 | |||||
| RQ4 | 0.8494 | 2.7827 | |||||
| Affective risk | Reflective | 0.8826 | 0.8834 | 0.6532 | |||
| AF1 | 0.8248 | 2.2906 | |||||
| AF2 | 0.8191 | 2.1781 | |||||
| AF3 | 0.7677 | 2.0855 | |||||
| AF4 | 0.8199 | 2.5521 | |||||
| Fraud risk | Reflective | 0.889 | 0.8922 | 0.6176 | |||
| FR 1 | 0.8115 | 2.478 | |||||
| FR 2 | 0.7064 | 1.796 | |||||
| FR 3 | 0.8304 | 2.3552 | |||||
| FR 4 | 0.7468 | 2.241 | |||||
| FR 5 | 0.8266 | 2.2365 | |||||
| Delivery risk | Reflective | 0.8716 | 0.8771 | 0.6959 | |||
| DR 1 | 0.8903 | 2.2643 | |||||
| DR 2 | 0.7591 | 2.1737 | |||||
| DR 3 | 0.8478 | 2.7531 | |||||
| Financial risk | Reflective | 0.8652 | 0.8666 | 0.6169 | |||
| FR 1 | 0.7776 | 2.2855 | |||||
| FR 2 | 0.7445 | 1.8052 | |||||
| FR 3 | 0.8232 | 2.0961 | |||||
| FR 4 | 0.7944 | 2.2315 | |||||
| P&T Loss risk | Reflective | 0.8401 | 0.8455 | 0.6401 | |||
| P&TR 1 | 0.8148 | 2.2941 | |||||
| P&TR 2 | 0.7329 | 1.7667 | |||||
| P&TR 3 | 0.8481 | 2.0981 | |||||
| Product risk | Reflective | 0.863 | 0.866 | 0.6141 | |||
| ProR 1 | 0.7753 | 2.1015 | |||||
| ProR 2 | 0.7394 | 1.7613 | |||||
| ProR 3 | 0.8441 | 2.4014 | |||||
| ProR 4 | 0.7719 | 2.0372 | |||||
| Privacy risk | Reflective | 0.8353 | 0.843 | 0.6322 | |||
| PryR 1 | 0.8695 | 2.0855 | |||||
| PryR 2 | 0.7227 | 1.8232 | |||||
| PryR 3 | 0.7864 | 1.9728 | |||||
| Information risk | Reflective | 0.8478 | 0.8547 | 0.74 | |||
| IR 1 | 0.9046 | 2.1809 | |||||
| IR 2 | 0.8134 | 2.1809 | |||||
| Purchase intention | Reflective | 0.8815 | 0.8831 | 0.7134 | |||
| PI1 | 0.8674 | 2.5317 | |||||
| PI2 | 0.861 | 2.5279 | |||||
| PI3 | 0.8041 | 2.3396 | |||||
| Perceived risk | Composite | NA | NA | NA | |||
| Fraud risk | 0.1533 | 3.1622 | |||||
| Delivery risk | 0.1603 | 2.2255 | |||||
| Financial risk | 0.1735 | 3.116 | |||||
| P&T loss risk | 0.1664 | 2.1107 | |||||
| Product risk | 0.1621 | 2.4522 | |||||
| Privacy risk | 0.1712 | 2.5206 | |||||
| Information risk | 0.1549 | 2.1795 |
Source: authors.
Studies show that the Fornell-Larcker criteria is not suitable for discriminant validity assessment. These criteria exhibit unsatisfactory performance, especially in cases when the indicator loadings on a construct exhibit little variation (e.g., all indicator loadings fall within the range of 0.65–0.85) (Voorhees et al. 2016). Consequently, this study assesses discriminant validity based on the heterotrait-monotrait ratio of correlation (HTMT) (Henseler et al. 2015). Table 3 describes the HTMT result. Accordingly, all values are statistically significantly smaller than 1 using the 95% percentile bootstrap confidence intervals. Therefore, the study concludes that discriminant validity is ensured.
Heterotrait-monotrait ratio of correlations (HTMT)
| Construct | Fraud risk | Delivery risk | Financial risk | P&T risk | Product risk | Privacy risk | Information risk | PI | AR | RQ |
| Fraud risk | ||||||||||
| Delivery risk | 0.7564 | |||||||||
| Financial risk | 0.8278 | 0.7516 | ||||||||
| P&T risk | 0.7757 | 0.6102 | 0.7103 | |||||||
| Product risk | 0.7503 | 0.7451 | 0.8044 | 0.6418 | ||||||
| Privacy risk | 0.8164 | 0.6846 | 0.7707 | 0.7618 | 0.7205 | |||||
| Information risk | 0.6516 | 0.6415 | 0.7955 | 0.6476 | 0.7216 | 0.7173 | ||||
| PI | 0.7659 | 0.683 | 0.7535 | 0.686 | 0.7177 | 0.7142 | 0.6977 | |||
| AR | 0.6493 | 0.5964 | 0.6441 | 0.6025 | 0.6394 | 0.6079 | 0.6001 | 0.7713 | ||
| RQ | 0.5465 | 0.5006 | 0.4951 | 0.4633 | 0.4236 | 0.4595 | 0.4513 | 0.4692 | 0.431 |
Source: authors.
The higher-order composite (i.e. perceived risk) was assessed via three criteria, including nomological validity, reliability, and weights (Henseler 2021; Van et al. 2017). The nomological validity is achieved by testing the overall model fit of the model with a higher-order construct. Accordingly, the structural model with the higher-order construct also achieves a good fit in the saturated and estimated models (dULS = 0.0822, P > 0.01, dULS = 0.112, P > 0.01, respectively). The SRMR values of the two models are 0.0219 and 0.0256, respectively, below the recommended threshold of 0.05, indicating a good model fit. The second-order composite (i.e. perceived risk) is formed by the factor scores of the first-order constructs. Therefore, its reliability has to be adjusted. In the model, the reliability of the second-order composite was 0.9511. The weights of the second-order composite are also computed as presented in Table 2. All weights are significantly different from 0 at a 1% significance level. The study concludes that the second-order construct (perceived risk) is appropriately modelled by a composite. Consequently, H1 is supported.
4.2 Structural model analysis
The structural model is evaluated via the constructs' explanatory power (R2) and effect size (f2). Accordingly, the values of 0.316, 0.542 and 0.787 for the R2 are considered high in behavioural science (Fig. 2). The R2 quantifies the proportion of variance of a dependent construct that its predictors explain (Henseler 2021).
Model containing the second-order composite
Source: authors.
Citation: Society and Economy 46, 2; 10.1556/204.2024.00003
Additionally, the path coefficient estimates should be significant based on the percentile bootstrap confidence interval, and their sign should be consistent with the associated hypotheses. Regarding significant path coefficients, an effect size (f2) above 0.02, 0.15, and 0.35 indicates a small, medium, and large effect size, respectively (Cohen 1988).
Table 4 presents the results of the structural model. The results show negative and significant effects of perceived risk and affective risk on purchase intention (β = −0.5733, P < 0.01; β = −0.3643, P < 0.01, respectively). Therefore, H2 and H4 are supported. The results show that perceived risk can be a predictor of affective risk (negative emotions) (β = 0.6921, P < 0.01). H3 is also supported. The study reveals that review quality negatively impacts consumers' perceived risk (β = −0.5622, P < 0.01). Therefore, H5a is also supported. However, the results also show insignificant effects of review quality on affective risk and purchase intention. H5b and H5c are not supported.
Results of the structural model
| Hypothesis | Direct/Indirect Effect | T-value | P-value (2-sided) | 95% Confidence Interval (Percentile Bootstrap) | Effect size (f2) | Supported | |
| H2 | Perceived risk - > Purchase intention | −0.5733 | −6.8680 | 0.0000 | [−0.7841, −0.3607] | 0.5340 | YES |
| H3 | Perceived risk - > Affective risk | 0.6921 | 12.2155 | 0.0000 | [0.5435, 0.8499] | 0.6741 | YES |
| H4 | Affective risk - > Purchase intention | −0.3643 | −4.8118 | 0.0000 | [−0.5539, −0.1552] | 0.2610 | YES |
| Mediating | PR->AR->PI | −0.2522 | −4.6830 | 0.0000 | [−0.3875, −0.1176] | NA | YES |
| Mediating | RQ->PR->PI | 0.3223 | 5.5196 | 0.0000 | [0.193, 0.5001] | NA | YES |
| Mediating | RQ->PR->AR | −0.3891 | −7.9283 | 0.0000 | [−0.5321, −0.2746] | NA | YES |
| Mediating | RQ->AR->PI | 0.0154 | 0.6543 | 0.5131 | [−0.0549, 0.0858] | NA | NO |
| Mediating | RQ->PR->AR->PI | 0.1418 | 4.4488 | 0.0000 | [0.0651, 0.2348] | NA | YES |
| H5a | Review quality - > Perceived risk | −0.5622 | −11.7998 | 0.0000 | [−0.6757, −0.4423] | 0.4404 | YES |
| H5b | Review quality - > Affective risk | −0.0423 | −0.6612 | 0.5086 | [−0.2069, 0.1316] | 0.0038 | NO |
| H5c | Review quality - > Purchase intention | −0.0088 | −0.1828 | 0.8550 | [−0.1385, 0.1202] | 0.0000 | NO |
Source: authors.
The study confirms that affective risk plays a significant role in mediating the relationship between perceived risk and purchase intention (β = −0.2252, P < 0.01).
5 Discussion
The study demonstrated the validity and reliability of modelling perceived risk as a second-order composite. It found that the concept of perceived risk in e-commerce can be understood based on multiple dimensions. The research is consistent with Glover and Benbasat (2010) who modelled perceived risk as a second-order composite and revealed that online consumers' concerns on different aspects are effective predictors of general perceived risk in e-commerce.
The research showed that perceived risk has a significant negative effect on purchase intention in cross-border e-commerce. The more risk consumers perceive on an e-commerce platform, the less purchase intention they have. This significant effect of perceived risk on purchase intention suggested in this study is consistent with He et al. (2022) and Sadiq et al. (2022). These scholars also found that risk perception is the factor that prevents online consumers from purchasing from online websites.
Similarly, the findings also reported the significant negative effect of affective risk on purchase intention in cross-border e-commerce. It indicates that if e-consumers feel worried, fearful, and distressed throughout the online purchasing process, they are more likely to reject their initial intention to make a purchase. Our findings align with the research by Sha (2017; 2018), who also demonstrates that affective risk can lower consumers' purchase intentions. However, our findings show that the effect size of affective risk on purchase intention is weaker than the effect size of perceived risk on purchase intention. This is different from the previous research by Sha (2017; 2018). That scholar concluded that online consumers gave more importance to their negative feelings to reduce purchase intention compared to their cognitive evaluations. According to Shiv and Fedorikhin (1999), the influence of affect or cognition on consumer behaviour depends on the allocation of processing resources to a decision-making task. This could be the case in our study, as our respondents were invited to accomplish a decision-making task. They were required to browse the specific e-commerce website and make purchases. This can make the effect of cognitive evaluation stronger than the effect of affective evaluation on decision-making.
The research also reveals the significant positive effect of perceived risk on affective risk. It implies that if e-consumers perceive high levels of risk on an e-commerce website, they are more likely to experience negative emotions that can influence their final decision-making. This suggests that consumers' perceptions of risk can predict their negative feelings (such as worry, distress, threat, and fear) when engaging in cross-border e-commerce transactions. The findings also align with the results of previous studies by Sha (2017; 2018), who revealed and confirmed the effect of perceived risk and affective risk in online shopping.
Furthermore, this study suggests that review quality also plays an important role in shaping consumers' perceptions, emotions, and behaviours in cross-border e-commerce. Accordingly, positive and high-quality reviews from other consumers lower consumers' perceived risk and affective risk (negative feelings), which in turn increases purchase intention. The results are consistent with studies by Liao and Huang (2021), Ventre and Kolbe (2020) and Zhang et al. (2014), who revealed the role of review quality in consumers' behaviours. Accordingly, negative reviews were found to affect consumers' perceived risk (Liao – Huang 2021). Whereas, positive reviews enable consumers' purchase intentions in e-commerce (Ventre – Kolbe 2020; Zhang et al. 2014). However, relying on the SOR paradigm, our study reveals the sequential role of the organism (i.e., cognition and emotion) as the mediator of the relationship between review quality and purchase intention. This implies that consumers' mechanisms of receiving and processing environmental stimuli often operate sequentially. A situational stimulus evokes cognitive and affective evaluations, which results in a consumer's intention.
6 Conclusion
The premise of this study is rooted in the recognition that the current marketing literature often explains risk perception as a cognitive process rather than an affective process. Furthermore, limited studies have applied the SOR model to explain consumers' risk perception and purchase intention in cross-border e-commerce. Therefore, this study was conducted to address the two research questions. The study concludes that two aspects of risk, including perceived risk and affective risk, lower consumers' purchase intentions in cross-border e-commerce. Additionally, perceived risk should be assessed as a higher-order composite with multiple sub-dimensions. Furthermore, the results found that review quality as a situational stimulus can contribute to mitigating perceived risk and affective risk while increasing consumers' purchase intention in cross-border e-commerce.
6.1 Theoretical implications
This study contributes in several ways to the existing literature on marketing and online consumer behaviour. First, this is one of a few studies examining perceived risk as a second-order composite in the context of e-commerce. Regarding the aspect of conceptualisation and measurement, this study makes a unique contribution by suggesting and proving the measurement structure of perceived risk as a higher-order variable. It emphasises the nuanced nature of perceived risk and makes a significant methodological contribution to risk perception in cross-border e-commerce.
Second, following the SOR paradigm, the study develops a rational framework to identify types of variables and understand mechanisms of patterns in research on risk perception and consumer behaviour in cross-border e-commerce. Accordingly, two patterns are revealed, including O–R and S–O–R.
Regarding the O–R mechanism, the study shows that perceived risk and affective risk correspond with the cognitive state and affective state of the organism (consumers). The study highlights that the online decision journey is twofold and influenced by the presence of emotions. In other words, the reaction of the consumer in cross-border e-commerce may follow the hierarchy of “perceived risk, affective risk, and purchase intention”. Additionally, the study also differentiates two aspects of risk. It identifies this affective risk as “immediate visceral reactions” (such as anxiety, fear and worry) that arise in response to risks in e-commerce. These immediate negative emotions differ from anticipated negative emotions, such as disappointment or regret, which are experienced in the future when making risky decisions.
Regarding the S–O–R mechanism, the study reveals the sequential reaction of the organism (consumer) when encountering an environmental stimulus. Consequently, cognitive state (perceived risk) and affective state (affective risk) serve as mediators of the relationship between review quality and purchase intention. The study indicates that the use of the SOR model is fruitful in explaining consumers' shopping behaviours.
By incorporating the presence of online reviews, this study also captures the overall virtual experience that a website brings about. The study emphasises the role of situational stimuli compared to other types of stimuli, such as website, product, seller, and marketing stimuli. It proposes that the quality of online reviews needs to be incorporated as an immediate stimulus to understand shopping behaviour and risk perception in cross-border e-commerce.
6.2 Practical implications and suggestions
Practically, our study also presents significance for practitioners. First, the study offers meaningful suggestions for e-businesses and e-retailers when dealing with Hungarian e-consumers. Although Hungarian consumers are familiar with online transactions, they are still worried about the perceived risk they may deal with when purchasing on cross-border platforms. Therefore, a compulsory requirement for e-vendors to furnish comprehensive information (e.g., name of the registered store, telephone number, address, email, product information, product origin, product return, guarantee policy, etc.) should be important to improve the transparency of the vendors and the platform. Furthermore, presenting multi-dimensions of perceived risk might be useful in developing related IT tools. For instance, implementing features that enable users to register as guests on the website or providing the option for anonymous credit or debit card purchases might potentially reduce consumers' perception of risk. Integrating the delivery tracking feature on the website itself may also facilitate consumers.
E-commerce enterprises must also give priority to the establishment of a robust and reliable review system. Authoritative and ample evaluations greatly reduce customers' perceived risks while making purchases. Certain organisations may disregard the significance of their review systems, leading to inadequate or untrustworthy consumer feedback and a deficiency in a robust and dependable social interaction ecosystem. Businesses may adopt a precise and meticulously crafted evaluation template to tackle this issue, offering explicit guidelines for customers to submit insightful input. Incorporating comment and sharing capabilities may foster online customers' social presence, hence augmenting the quality and trustworthiness of reviews.
6.3 Limitation and future research
The paper acknowledges several limitations that can guide future research. First, the explanatory power of the review quality on perceived risk only obtains 31%. This implies that there are other factors stimulating consumers' cognition. Therefore, identifying other stimuli may be a meaningful direction. Second, the study used non-probability sampling techniques to collect the sample. The reason is that the authors were not allowed to access the list of the Hungarian population. This approach may impact the generalisation of the whole population due to the high likelihood of sampling bias. Therefore, future studies may cooperate with the Hungarian census organisation to conduct the study using probability sampling techniques. In other ways, future studies may also utilise a sample of students. Third, the current study designed a cross-section study that offers a snapshot of the impact of the situational stimulus on consumers' cognition, emotion and behaviours at a point in time. Therefore, the causality of the relationships needs to be considered. It will be a direction for further research to conduct a longitudinal study to investigate the change in consumers' perceptions and behaviours.
Acknowledgement
Research for this paper was funded under project no. TKP2021-NKTA-19, implemented with support provided from the National Research, Development and Innovation Fund of Hungary, financed under the TKP2021-NKTA funding scheme.
References
Abdellaoui, M. – Bleichrodt, H. – Paraschiv, C. (2007): Loss Aversion under Prospect Theory: A Parameter-free Measurement. Management Science 53(10): 1659–1674. https://www.jstor.org/stable/20122321.
Afthanorhan, A. – Awang, Z. – Aimran, A. (2020): Five Common Mistakes for Using Partial Least Squares Path Modeling (PLS-PM) in Management Research. Contemporary Management Research 16: 255–278. https://doi.org/10.7903/cmr.20247.
Alimamy, S. – Gnoth, J. (2022): I Want it My Way! The Effect of Perceptions of Personalization through Augmented Reality and Online Shopping on Customer Intentions to Co-create Value. Computers in Human Behavior 128: 107105. https://doi.org/10.1016/j.chb.2021.107105.
Alrawad, M. – Lutfi, A. – Alyatama, S. – Al Khattab, A. – Alsoboa, S. S. – Almaiah, M. A. – Ramadan, M. H. (2023): Assessing Customers Perception of Online Shopping Risks: A Structural Equation Modeling–Based Multigroup Analysis. Journal of Retailing and Consumer Services 71: 103188. https://doi.org/10.1016/j.jretconser.2022.103188.
Ariffin, S. K. – Mohan, T. – Goh, Y. N. (2018): Influence of Consumers’ Perceived Risk on Consumers’ Online Purchase Intention. Journal of Research in Interactive Marketing 12(3): 309–327. https://doi.org/10.1108/JRIM-11-2017-0100.
Chafidon, M. A. A. Z. – Margono, M. – Sunaryo, S. (2022): Social Media Marketing on Purchase Intention through Mediated Variables of Perceived Value and Perceived Risk. Interdisciplinary Social Studies 1(9): 1182–1195. https://doi.org/10.55324/iss.v1i9.228.
Chang, K. C. – Hsu, C. L. – Chen, M. C. – Kuo, N. T. (2019): How a Branded Website Creates Customer Purchase Intentions. Total Quality Management & Business Excellence 30(3–4): 422–446. https://doi.org/10.1080/14783363.2017.1308819.
Chen, Y. – Wang, Q. – Xie, J. (2011): Online Social Interactions: A Natural Experiment on Word of Mouth versus Observational Learning. Journal of Marketing Research 48(2): 238–254. https://doi.org/10.1509/jmkr.48.2.238.
Chiu, C. – Wang, E. T. G. – Fang, Y. – Huang, H. (2014): Understanding Customers’ Repeat Purchase Intentions in B2C E-Commerce: The Roles of Utilitarian Value, Hedonic Value and Perceived Risk. Information Systems Journal 24(1): 85–114. https://doi.org/10.1111/j.1365-2575.2012.00407.x.
CMI (2023): European E-Commerce Report 2023. https://www.upu.int/UPU/media/www-UpuIntUniversalPostalUnionAboutUpuBodiesConsultativeCommittee/2023EuropeanEcommerceReportEn.pdf, accessed 09/02/2024.
Cohen, J. (1988): Statistical Power Analysis for the Behavioral Sciences. L. Erlbaum Associates, Hillsdale, N.J.
Daragmeh, A. – Lentner, C. – Sági, J. (2021): FinTech Payments in the Era of COVID-19: Factors Influencing Behavioral Intentions of ‘Generation X’ in Hungary to Use Mobile Payment. Journal of Behavioral and Experimental Finance 32: 100574. https://doi.org/10.1016/j.jbef.2021.100574.
Éthier, J. – Hadaya, P. – Talbot, J. – Cadieux, J. (2006): B2C Web Site Quality and Emotions during Online Shopping Episodes: An Empirical Study. Information & Management 43(5): 627–639. https://doi.org/10.1016/j.im.2006.03.004.
Etikan, I. (2019): A Basic Approach in Sampling Methodology and Sample Size Calculation. Med Life Clinics 1(2): 1006.
Finucane, M. L. – Alhakami, A. – Slovic, P. – Johnson, S. M. (2000): The Affect Heuristic in Judgments of Risks and Benefits. Journal of Behavioral Decision Making 13(1): 1–17. https://doi.org/10.1002/(SICI)1099-0771(200001/03)13:1<1::AID-BDM333>3.0.CO;2-S.
Gáti, M. – Simay, A. E. (2019): How Consumers Can Understand the Role and Relevance of Social Mobile E-Commerce. In Csordás, T. – Varga, Á. (eds). Budapest: Budapesti Corvinus Egyetem, pp. 9–17.
Glover, S. – Benbasat, I. (2010): A Comprehensive Model of Perceived Risk of E-Commerce Transactions. International Journal of Electronic Commerce 15(2): 47–78.
Guo, Y. – Bao, Y. – Stuart, B. J. – Le, N. K. (2018): To Sell or Not to Sell: Exploring Sellers’ Trust and Risk of Chargeback Fraud in Cross-Border Electronic Commerce. Information Systems Journal 28(2): 359–383. https://doi.org/10.1111/isj.12144.
Hair, J. F. – Hult, G. T. M. – Ringle, C. M. – Sarstedt, M. (2017): A Primer on Partial Least Squares Structural Equation Modeling (PLS-SEM). Thousand Oaks: Sage.
He, Y. – Li, W. – Xue, J. (2022): What and How Driving Consumer Engagement and Purchase Intention in Officer Live Streaming? A Two-Factor Theory Perspective. Electronic Commerce Research and Applications 56: 101223. https://doi.org/10.1016/j.elerap.2022.101223.
Henseler, J. (2021): Composite-Based Structural Equation Modeling: Analyzing Latent and Emergent Variables. New York: The Guilford Press.
Henseler, J. – Ringle, C. M. – Sarstedt, M. (2015): A New Criterion for Assessing Discriminant Validity in Variance-Based Structural Equation Modeling. Journal of the Academy of Marketing Science 43(1): 115–135. https://doi.org/10.1007/s11747-014-0403-8.
Hsiao, Y. H. – Chen, M. C. – Liao, W. C. (2017): Logistics Service Design for Cross-Border E-Commerce Using Kansei Engineering with Text-Mining-Based Online Content Analysis. Telematics and Informatics 34(4): 284–302. https://doi.org/10.1016/j.tele.2016.08.002.
Hu, L. – Bentler, P. M. (1999): Cutoff Criteria for Fit Indexes in Covariance Structure Analysis: Conventional Criteria versus New Alternatives. Structural Equation Modeling: A Multidisciplinary Journal 6(1): 1–55. https://doi.org/10.1080/10705519909540118.
Jacoby, J. – Kaplan, L. B. (1972): The Components of Perceived Risk. ACR Special Volumes SV-02.
Johnson, R. E. – Rosen, C. C. – Chang, C. H. (2011): To Aggregate or Not to Aggregate: Steps for Developing and Validating Higher-Order Multidimensional Constructs. Journal of Business and Psychology 26(3): 241–248. https://doi.org/10.1007/s10869-011-9238-1.
Kim, J. – Lennon, S. (2013): Effects of Reputation and Website Quality on Online Consumers’ Emotion, Perceived Risk and Purchase Intention: Based on the Stimulus-Organism-Response Model. Journal of Research in Interactive Marketing 7. https://doi.org/10.1108/17505931311316734.
Kimiagari, S. – Malafe, A. – Neda, S. (2021): The Role of Cognitive and Affective Responses in the Relationship between Internal and External Stimuli on Online Impulse Buying Behavior. Journal of Retailing and Consumer Services 61: 102567. https://doi.org/10.1016/j.jretconser.2021.102567.
Kollmann, T. – Suckow, C. (2012): Influencing User Attitudes by Managing Online Brand Communication in E-Shops. Marketing: Zeitschrift Für Forschung Und Praxis 34(1): 40–54.
Lee, J. Y. – Bell, D. R. (2013): Neighborhood Social Capital and Social Learning for Experience Attributes of Products. Marketing Science 32(6): 960–976. https://doi.org/10.1287/mksc.2013.0796.
Lho, L. H. – Quan, W. – Yu, J. – Han, H. (2022): The Sharing Economy in the Hospitality Sector: The Role of Social Interaction, Social Presence, and Reciprocity in Eliciting Satisfaction and Continuance Behavior. Humanities and Social Sciences Communications 9(1): 1–12. https://doi.org/10.1057/s41599-022-01379-y.
Li, H. – Sarathy, R. – Xu, H. (2011): The Role of Affect and Cognition on Online Consumers’ Decision to Disclose Personal Information to Unfamiliar Online Vendors. Decision Support Systems 51(3): 434–445. https://doi.org/10.1016/j.dss.2011.01.017.
Liao, H. L. – Huang, Z. (2021): The Effects of Negative Online Reviews on Consumer Perception, Attitude and Purchase Intention: Experimental Investigation of the Amount, Quality, and Presentation Order of eWOM. ACM Transactions on Asian Low-Resource Language Information Processing 20(3): 1–21.
Loewenstein, G. – Weber, E. – Hsee, C. – Welch, N. (2001): Risk as Feelings. Psychological Bulletin 127: 267–86. https://doi.org/10.1037/0033-2909.127.2.267.
Luo, S. – Gu, B. – Wang, X. – Zhou, Z. (2018): Online Compulsive Buying Behavior: the Mediating Role of Self-Control and Negative Emotions. Proceedings of the 2018 International Conference on Internet and E-Business (ICIEB’18). ACM Press, Singapore, pp. 65–69. https://doi.org/10.1145/3230348.3230397.
Malak, F. – Ferreira, J. B. – Pessoa de Queiroz Falcão, R. – Giovannini, C. J. (2021): Seller Reputation within the B2C E-Marketplace and Impacts on Purchase Intention. Latin American Business Review 22(3): 287–307. https://doi.org/10.1080/10978-526.2021.1893182.
Mehrabian, A. – Russell, J. A. (1980): An Approach to Environmental Psychology. The MIT Press.
Meskaran, F. – Ismail, Z. – Shanmugam, B. (2013): Online Purchase Intention: Effects of Trust and Security Perception. Australian Journal of Basic and Applied Sciences 7(6): 307–315.
Mohseni, S. – Jayashree, S. – Rezaei, S. – Kasim, A. – Okumus, F. (2018): Attracting Tourists to Travel Companies’ Websites: the Structural Relationship between Website Brand, Personal Value, Shopping Experience, Perceived Risk and Purchase Intention. Current Issues in Tourism 21(6): 616–645. https://doi.org/10.1080/13683500.2016.1200539.
Naiyi, Y. (2004): Dimensions of Consumer’s Perceived Risk in Online Shopping. Journal of Electronic Science and Technology 2(3): 177–182.
Nilashi, M. – Abumalloh, R. A. – Samad, S. – Alrizq, M. – Alyami, S. – Abosaq, H. – Alghamdi, A. (2022): Factors Impacting Customer Purchase Intention of Smart Home Security Systems: Social Data Analysis Using Machine Learning Techniques. Technology in Society 71: 102118. https://doi.org/10.1016/j.techsoc.2022.102118.
Park, D. H. – Lee, J. – Han, I. (2007): The Effect of On-Line Consumer Reviews on Consumer Purchasing Intention: The Moderating Role of Involvement. International Journal of Electronic Commerce 11(4): 125–148. https://www.jstor.org/stable/27751234.
Pavlou, P. A. – Liang, H. – Xue, Y. (2007): Understanding and Mitigating Uncertainty in Online Exchange Relationships: A Principal-Agent Perspective. MIS Quarterly 31(1): 105–136. https://doi.org/10.2307/25148783.
Pentz, C. D. – du Preez, R. – Swiegers, L. (2020): To Bu(Y) or Not to Bu(Y): Perceived Risk Barriers to Online Shopping Among South African Generation Y Consumers. Cogent Business & Management 7(1): 1827813. https://doi.org/10.1080/23311975.2020.1827813.
Petrova, D. – Cokely, E. T. – Sobkow, A. – Traczyk, J. – Garrido, D. – Garcia-Retamero, R. (2023): Measuring Feelings about Choices and Risks: The Berlin Emotional Responses to Risk Instrument (BERRI). Risk Analysis 43(4): 724–746. https://doi.org/10.1111/risa.13946.
Pham, V. T. – Ha, N. M. – Hiep, B. D. – Huy, N. Q. – Mai, T. T. N. – Phuong, P. M. (2022): The Experience of Using E-Commerce Platforms Affects the Online Purchase Intention of Customers in the FMCG (Fast Moving Consumer Goods) Sector in Hanoi City. International Journal of Business and Social Science Research 3(4): 16–24. https://doi.org/10.47742/ijbssr.v3n4p3.
Pillai, S. G. – Kim, W. G. – Haldorai, K. – Kim, H. S. (2022): Online Food Delivery Services and Consumers’ Purchase Intention: Integration of Theory of Planned Behavior, Theory of Perceived Risk, and the Elaboration Likelihood Model. International Journal of Hospitality Management 105. https://doi.org/10.1016/j.ijhm.2022.103275.
Pratesi, F. – Hu, L. – Rialti, R. – Zollo, L. – Faraoni, M. (2021): Cultural Dimensions in Online Purchase Behavior: Evidence from a Cross-Cultural Study. Italian Journal of Marketing 2021(3): 227–247. https://doi.org/10.1007/s43039-021-00022-z.
Qalati, S. A. – Vela, E. G. – Li, W. – Dakhan, S. A. – Hong Thuy, T. T. – Merani, S.H. (2021): Effects of Perceived Service Quality, Website Quality, and Reputation on Purchase Intention: The Mediating and Moderating Roles of Trust and Perceived Risk in Online Shopping. Cogent Business & Management 8(1): 1869363. https://doi.org/10.1080/23311975.2020.1869363.
Rosillo-Díaz, E. – Blanco-Encomienda, F. J. – Crespo-Almendros, E. (2019): A Cross-Cultural Analysis of Perceived Product Quality, Perceived Risk and Purchase Intention in E-Commerce Platforms. Journal of Enterprise Information Management 33(1): 139–160. https://doi.org/10.1108/JEIM-06-2019-0150.
Sadiq, M. – Dogra, N. – Adil, M. – Bharti, K. (2022): Predicting Online Travel Purchase Behavior: The Role of Trust and Perceived Risk. Journal of Quality Assurance in Hospitality & Tourism 23(3): 796–822. https://doi.org/10.1080/1528-008X.2021.1913693.
Sha, W. (2017): Examining the Construct Validities and Influence of Affective Risk in B2C E-Commerce. Issues in Information Systems 18(4): 46-56. https://doi.org/10.48009/4_iis_2017_46-56.
Sha, W. (2018): Development of Instrument for Affective Risk in Business-To-Consumer E-Commerce. Issues in Information Systems 19(3): 11-21. https://doi.org/10.48009/3_iis_2018_11-21.
Shiv, B. – Fedorikhin, A. (1999): Heart and Mind in Conflict: The Interplay of Affect and Cognition in Consumer Decision Making. Journal of Consumer Research 26(3): 278–292. https://doi.org/10.1086/209563.
Slovic, P. – Finucane, M. L. – Peters, E. – MacGregor, D. G. (2004): Risk as Analysis and Risk as Feelings: Some Thoughts about Affect, Reason, Risk, and Rationality. Risk Analysis 24(2): 311–322. https://doi.org/10.1111/j.0272-4332.2004.00433.x.
Slovic, P. – Peters, E. (2006): Risk Perception and Affect. Current Directions in Psychological Science 15(6): 322–325.
Soper, D.S. (2023): A-Priori Sample Size Calculator for Structural Equation Models. Danielsoper.com. https://www.danielsoper.com/statcalc/cal-culator.aspx?id=89, accessed 27/11/2023.
Statista. (2023a): Global Cross-Border E-Commerce Share 2022. Statista. https://www.statista.com/statistics/867991/cross-border-e-commerce-share-world/, accessed 24/01/2024.
Statista. (2023b): Online Shopping Rate in Hungary 2022. Statista. https://www.statista.com/statistics/1095820/online-shopping-rate-in-hungary/, accessed 20/07/2023.
Tompkins, M. K. – Bjälkebring, P. – Peters, E. (2018): Emotional Aspects of Risk Perceptions. In: Raue, M. – Lermer, E. – Streicher, B. (eds). Psychological Perspectives on Risk and Risk Analysis: Theory, Models, and Applications. Cham: Springer, pp. 109–130. https://doi.org/10.1007/978-3-319-92478-6_5.
Van, R. A. C. R. – Henseler, J. – Kemény, I. – Sasovova, Z. (2017): Estimating Hierarchical Constructs Using Consistent Partial Least Squares: The Case of Second-Order Composites of Common Factors. Industrial Management & Data Systems 117(3): 459–477. https://doi.org/10.1108/IMDS-07-2016-0286.
Ventre, I. – Kolbe, D. (2020): The Impact of Perceived Usefulness of Online Reviews, Trust and Perceived Risk on Online Purchase Intention in Emerging Markets: A Mexican Perspective. Journal of International Consumer Marketing 32(4): 287–299. https://doi.org/10.1080/08961530.2020.1712293.
Verhagen, T. – Dolen, W. (2011): The Influence of Online Store Beliefs on Consumer Online Impulse Buying: A Model and Empirical Application. Information & Management 48(8): 320–327. https://doi.org/10.1016/j.im.2011.08.001.
Voorhees, C. M. – Brady, M. K. – Calantone, R. – Ramirez, E. (2016): Discriminant Validity Testing in Marketing: an Analysis, Causes for Concern, and Proposed Remedies. Journal of the Academy of Marketing Science 44(1): 119–134. https://doi.org/10.1007/s11747-015-0455-4.
Wang, E. S. T. (2010): Internet Usage Purposes and Gender Differences in the Effects of Perceived Utilitarian and Hedonic Value. Cyberpsychology, Behavior and Social Networking 13(2): 179–183. https://doi.org/10.1089/cyber.2009.0200.
Wang, J. C. – Chang, C. H. (2013): How Online Social Ties and Product-Related Risks Influence Purchase Intentions: A Facebook Experiment. Electronic Commerce Research and Applications 12(5): 337–346. https://doi.org/10.1016/j.elerap.2013.03.003.
Wang, Y. – Yu, C. (2017): Social Interaction-Based Consumer Decision-Making Model in Social Commerce: The Role of Word of Mouth and Observational Learning. International Journal of Information Management 37(3): 179–189. https://doi.org/10.1016/j.ijinfomgt.2015.11.005.
Wetzels – Odekerken, S. – Van O. (2009): Using PLS Path Modeling for Assessing Hierarchical Construct Models: Guidelines and Empirical Illustration. MIS Quarterly 33(1): 177. https://doi.org/10.2307/20650284.
Xu, H. – Zhang, K. Z. K. – Zhao, S. J. (2020): A Dual Systems Model of Online Impulse Buying. Industrial Management & Data Systems 120(5): 845–861. https://doi.org/10.1108/IMDS-04-2019-0214.
Yang, J. – Sarathy, R – Lee, J. K. (2016): The Effect of Product Review Balance and Volume on Online Shoppers’ Risk Perception and Purchase Intention. Decision Support Systems 89: 66–76. https://doi.org/10.1016/j.dss.2016.06.009.
Zhang, K. – Hu, B. – Zhao, S. (2014): How Online Social Interactions Affect Consumers' Impulse Purchase on Group Shopping Websites. PACIS 2014 Proceedings.
Zhang, K. Z. K. – Lee, M. K. O. – Zhao, S. J. (2010): Understanding the Informational Social Influence of Online Review Platforms. ICIS 2010 Proceedings 71.
Zhang, H. – Lu, Y. – Shi, X. – Tang, Z. – Zhao, Z. (2012): Mood and Social Presence on Consumer Purchase Behaviour in C2C E-Commerce in Chinese Culture. Electronic Markets 22(3): 143–154. https://doi.org/10.1007/s12525-012-0097-z.
Zhang, L. – Tan, W. – Xu, Y. – Tan, G. (2012): Dimensions of Perceived Risk and Their Influence on Consumers’ Purchasing Behavior in the Overall Process of B2C. In: Zhang, L. – Zhang, C. (eds): Engineering Education and Management. The 2011 International Conference on Engineering Education and Management (ICEEM2011). Berlin: Springer, pp. 1–10. https://doi.org/10.1007/978-3-642-24823-8_1.
Zhang, K. Z. K. – Zhao, S. J. – Cheung, C. M. K. – Lee, M. K. O. (2014): Examining the Influence of Online Reviews on Consumers’ Decision-Making: A Heuristic–Systematic Model. Decision Support Systems 67: 78–89. https://doi.org/10.1016/j.dss.2014.08.005.
Zhu, L. – Li, H. – Wang, F. K. – He, W. – Tian, Z. (2020): How Online Reviews Affect Purchase Intention: a New Model Based on the Stimulus-Organism-Response (S-O-R) Framework. Aslib Journal of Information Management 72(4): 463–488. https://doi.org/10.1108/AJIM-11-2019-0308.
Appendix 1. The sample website made by the authors
Appendix 2. Variables and items
1. Perceived risk with 7 dimensions adapted from Naiyi (2004)
1.1. Fraud risk
Information about the seller on this website may not true.
It may be difficult to get support on this website when product fails.
I may not find the place where to settle disputes on this website.
This website may disappear after a short time.
This website may fail to keep the promise of post-services.
1.2. Delivery risk
The delivered product may be lost.
The product may be delivered a wrong place.
The product may be damaged during the delivering.
1.3 Financial risk
Traditional stores may offer more discounts than this website.
This website offers discount prices but the total cost may not lower.
Online payment on this website will charge extra fees.
Delivering to the home will charge relatively higher fees.
1.4. Process & Time loss risk
The process of purchasing on this website is complex and inconvenient.
Accessing this website will take too much time.
Information transformation is too slow during purchasing.
1.5. Product risk
The quality of the product may not be accepted.
The product performance may not be consistent with the expectation.
The product may be false and the quality will be poor.
It is difficult to return when the product is not satisfied.
1.6. Privacy risk
My personal address, telephone number may be misused by others.
My e-mail address may be misused by others.
The account number of my credit or debit card can be misused by others.
1.7. Information risk
The information about online suppliers on this website is not sufficient.
The information about product to be purchased on this website is not sufficient.
2. Affective risk with 4 items adapted from Petrova et al. (2023) and Sha (2018)
AR1. I feel tense while purchasing on this website.
AR2. I am worried when ordering on this website.
AR3. I am afraid of finishing the order placement.
AR4. I feel uneasy about this website.
3. Purchase intention with 3 items adapted from Zhang et al. (2012)
PI1. I am likely to purchase the products(s) on this website.
PI2. I am likely to recommend this website to my friends.
PI3. I am likely to make another purchase from this website next time.
4. Review quality with 4 items adapted from Park et al. (2007)
RQ1. Reviews about products and the seller on this website are understandable.
RQ2. Reviews about products and the seller on this website are objective.
RQ3. Reviews about products and the seller on the website are accurate.
RQ4. Reviews about products and the seller on the website are credible.
