How Online Activities Shape Anti-Fraud Payment Behavior: Evidence from Chinese Youth

Abstract

In today’s digital age, understanding how online behaviors affect payment security awareness is crucial for fraud prevention. This study investigates how different types of online activities shape anti-fraud payment behavior among Chinese youth, utilizing data from the Chinese General Social Survey (2017). Through ordered Probit regression analysis of 557 valid samples, we provide empirical evidence that social networking and self-presentation activities significantly enhance anti-fraud awareness. In contrast, information acquisition and online rights protection activities show moderate effects. The influence mechanisms demonstrate significant demographic heterogeneity: gender differences appear in behavioral pattern formation, rural-urban disparities affect the impact of social activities, and educational background moderates the relationship between online engagement and anti-fraud consciousness. These findings suggest the importance of developing targeted fraud prevention strategies considering the nature of online activities and demographic characteristics. Our study contributes to understanding the formation mechanism of anti-fraud behavior in digital payment environments and provides practical implications for enhancing youth fraud prevention. In the digital payment era, online fraud is increasingly threatening young people’s financial security through increasingly sophisticated methods. Contrary to traditional views, this study finds that simply increasing anti-fraud knowledge education may not be sufficient—the key lies in understanding how young people form and strengthen their payment security awareness through daily online activities. Through analysis of survey data from China, the study finds that social networking and self-presentation activities significantly enhance young people’s payment security awareness, primarily because these activities provide opportunities for sharing fraud cases, mutual reminders, and learning anti-fraud experiences. However, in daily entertainment and commercial transaction scenarios, young people’s security awareness tends to decrease, which is precisely the vulnerability that fraudsters often exploit. The research also reveals that young people from different backgrounds show different characteristics and vulnerabilities in fraud prevention, suggesting the need for more targeted strategies in anti-fraud efforts. These findings provide new insights and practical recommendations for improving the effectiveness of fraud prevention.

Plain language summary

Preventing online payment fraud: How young people’s online activities shape security awareness

In the digital payment era, online fraud is increasingly threatening young people’s financial security through increasingly sophisticated methods. Contrary to traditional views, this study finds that simply increasing anti-fraud knowledge education may not be sufficient - the key lies in understanding how young people form and strengthen their payment security awareness through daily online activities. Through analysis of survey data from China, the study finds that social networking and self-presentation activities significantly enhance young people’s payment security awareness, primarily because these activities provide opportunities for sharing fraud cases, mutual reminders, and learning anti-fraud experiences. However, in daily entertainment and commercial transaction scenarios, young people’s security awareness tends to decrease, which is precisely the vulnerability that fraudsters often exploit. The research also reveals that young people from different backgrounds show different characteristics and vulnerabilities in fraud prevention, suggesting the need for more targeted strategies in anti-fraud efforts. These findings provide new insights and practical recommendations for improving the effectiveness of fraud prevention.

Keywords

online activities payment security anti-fraud behavior Chinese youth risk perception demographic heterogeneity

Introduction

In today’s digital age, the rise of online payment systems has significantly altered financial behaviors across the globe (Roh et al., 2024), primarily affecting young adults—a demographic highly susceptible to online fraud. While the convenience of online transactions enhances efficiency, it has also introduced a range of security challenges. Studies (Upadhyay & Chattopadhyay, 2015) indicate that online payment fraud has evolved from simple phishing schemes to more sophisticated tactics that exploit users’ behavioral tendencies and psychological vulnerabilities. Young adults face heightened risks due to frequent engagement in digital transactions and intricate interactions across various online platforms (Dai & Palvi, 2009).

In recent years, online fraud cases have shown a significant growth trend. According to data from the US Federal Trade Commission, fraud losses reported by American consumers rose from $5.8 billion in 2021 (Federal Trade Commission [FTC], 2022) to $8.8 billion in 2022 (FTC, 2023a), exceeding $10 billion for the first time in 2023 (FTC, 2024a), and further increasing to $12.5 billion in 2024 (FTC, 2024b). This rapid upward trend indicates that despite continuous advancements in security technology, online fraud activities continue to flourish and become more destructive. Investment fraud represents the largest category of losses, reaching $4.6 billion in 2023, a 21% increase from 2022 (FTC, 2024a). Meanwhile, the contact methods used by scammers are also evolving, with email surpassing phone calls and text messages in 2023 to become the most common method of fraudulent contact (FTC, 2024a).

Data shows that contrary to common perception, young people are equally susceptible to fraud and, in some aspects, face higher risks. According to recent research, while older individuals suffer greater average losses when defrauded (victims in their 70s lose an average of $803, compared to $460 for victims in their 30s), young people are scammed at higher rates (AARP, 2025). A nationwide survey found that over the past 12 months, 37% of Generation Z and 26% of Millennials reported becoming fraud victims (IPX1031, 2024). Young people’s frequent and diverse activities in cyberspace expose them to unique risk environments. Research indicates that different online activity patterns lead to varying risk exposure; for example, social media use is significantly associated with specific types of fraud risk (FTC, 2023b). These data provide a strong rationale for this study’s focus on young people, indicating they are a potentially vulnerable group that requires special attention.

Risk perception regarding online payments distinctly differs from that of traditional financial decision-making. Cross-cultural analyses (Rouibah et al., 2016) reveal that greater familiarity with digital technologies can diminish sensitivity to potential risks, even amid prominent security threats. This trend is especially noticeable among young users, who, despite being technically savvy, are often vulnerable to advanced fraudulent schemes. Behavioral studies (Lin et al., 2020) indicate that increased exposure to online payment systems can lead to an inflated sense of security, impairing users’ ability to recognize risks accurately.

Addressing online payment security requires a comprehensive approach encompassing technical measures and psychological and behavioral factors significantly contributing to user vulnerability (Gong et al., 2020). Research into victimization trends (Zhou, 2014) indicates a concerning shift where fraudsters increasingly exploit users’ online behaviors rather than merely targeting technical weaknesses. This evolution in fraudulent strategies highlights the importance of understanding how daily online activities affect users’ perceptions and security habits. Conventional security training often fails to account for the psychological dynamics heightening user susceptibility.

Current research on online payment security has typically approached the issue from three primary frameworks (Vida et al., 2012). Technical studies focus on improving security protocols and fraud detection technologies (Kim, Mirusmonov, & Lee, 2010); Behavioral research examines the decision-making patterns of users in risky online contexts (Saridakis et al., 2016; Vishwanath, 2015); Demographic studies investigate the characteristics of victims to identify populations at higher risk (Park, 2013). While each perspective provides valuable insights, a critical gap exists in understanding how users’ varied online activities collectively shape their security psychology (Youn, 2009).

Security psychology is a field that studies how human behavior and cognitive processes affect the effectiveness of security measures, revealing that security is both a feeling and a reality state, and these two are not always consistent (Schneier, 2008). Research in security psychology shows that people’s perception of risk is often influenced by various psychological biases, such as the familiarity effect—frequent exposure to an environment reduces sensitivity to risks in that environment; affective heuristics—emotional experiences in certain situations inhibit rational assessment of potential risks; and the illusion of control—overconfidence in the controllability of risks (Finucane et al., 2000; Slovic et al., 2013). In digital environments, users typically need to make quick decisions in complex technical contexts, and their judgments about technological risks often rely on limited information and experience. Therefore, understanding these psychological mechanisms is crucial for designing effective security strategies, as users’ insecure behaviors can circumvent even the most advanced technical protections (Wiederhold, 2014).

Payment security psychology is a specialized branch of security psychology, explicitly referring to an individual’s perception, evaluation, and response processes to security risks in payment environments. It includes users’ subjective judgments about whether their personal and financial information will not be illegally acquired or misused during payment transactions (Kim, Mirusmonov, & Lee, 2010). Payment security psychology comprises two key dimensions: perceived security and perceived convenience. Research indicates these factors are core considerations when users evaluate and select payment methods (X. Chen & Li, 2017). In mobile payment environments, users’ perceptions of security and evaluations of system usability directly influence their continued usage intention, forming the foundational structure of payment security psychology. Studies show that payment security psychology affects users’ behaviors, such as choosing which payment platform to use, under what circumstances to conduct transactions, and adopting additional security measures (Hossain, 2019). Unlike general risk perception, payment security psychology focuses explicitly on financial transactions’ security and privacy protection, reflecting users’ dual concerns about economic loss and personal information leakage (Flavián & Guinalíu, 2006). In this study, payment security psychology refers explicitly to users’ tendency to observe the environment and evaluate risks in online payment settings, a psychological construct significantly influenced by individual patterns of online activities.

The online activities of young adults have become increasingly diverse and complex, encompassing a wide range of interactions, from social networking and financial management to entertainment and information-seeking (Tu et al., 2022). This diversification signifies a notable shift in usage patterns and fundamental changes in how young individuals engage with digital ecosystems. Research shows that these different online experiences create distinct behavioral patterns significantly influencing risk perception and security awareness (Liébana-Cabanillas et al., 2021). For example, heavy social media participation can impact users’ trust levels differently than traditional financial engagements, illustrating a complex interplay between various online pursuits and security psychology (Cengiz et al., 2022).

Understanding the connection between online behavior and security psychology has important implications for theoretical development and practical applications (Brenner et al., 2020). Theoretically, it enhances existing risk perception models by incorporating the complex nature of digital interactions. Unlike traditional risk environments—where increased involvement raises awareness of threats—the digital landscape presents a unique situation where more excellent activity may heighten or diminish security consciousness, depending on the type of engagement and user characteristics (Handarkho, 2020). This expanded theoretical framework offers new insights into how digital habits influence risk perception.

Furthermore, investigations (Liébana-Cabanillas et al., 2018) indicate that the consequences of online behavior on security psychology can vary significantly among different demographic groups. Cultural background, educational attainment, and digital literacy levels are crucial in how online experiences affect risk perception. This variability suggests that adequate security measures must consider the types of online behaviors and the unique characteristics of various user cohorts.

The present study addresses these theoretical and practical challenges by systematically analyzing how different online activities shape young people’s perceptions of payment security. Utilizing data from the Chinese General Social Survey (2017), we seek to unravel the intricate relationships between various types of online engagement and security risk perception, ultimately shedding light on strategies for preventing online fraud in this vulnerable population.

Literature Review

Risk Perception in Online Fraud Prevention

The study of risk perception in online fraud prevention has evolved significantly with the transformation of digital payment environments. Traditional risk perception theory, initially developed by Slovic (1987) to understand how individuals evaluate general risks, has gained new relevance in the context of online payment fraud. According to Yang et al. (2015), individuals’ risk perceptions of online payments are multidimensional, encompassing systematic and transactional risk facets that affect trust formation differently. Balapour et al. (2020) further demonstrate that individuals’ ability to perceive and evaluate fraud risks differs substantially from their assessment of conventional financial risks.

Risk perception in online fraud contexts uniquely challenges traditional theoretical frameworks. Schneier (2008) identifies a fundamental disconnect between perceived and actual security, showing that security is both a feeling and a reality that does not always align. While conventional risk perception often improves with experience, Davinson and Sillence (2014) indicate that increased familiarity with digital payment systems can paradoxically reduce fraud risk awareness. Research by Hossain (2019) and Salah and Ayyash (2025) found that frequent digital payment users often develop automated behavioral patterns that bypass conscious risk evaluation processes, particularly in mobile payment scenarios where streamlined interfaces may further reduce psychological barriers to payment decisions.

The application of risk perception theory to online fraud prevention has highlighted several critical insights. Wang et al. (2019) demonstrate that the digital environment creates unique challenges for risk assessment due to the absence of traditional physical cues. According to the FTC (2024a), as nationwide fraud losses topped $10 billion in 2023, these challenges have become increasingly consequential for consumers and institutions alike. Recent research by Nandru et al. (2025) reveals that the speed and convenience of digital transactions can compress decision-making timeframes, potentially compromising risk evaluation. Moreover, Featherman et al. (2010) observed that the social nature of many digital platforms blurs the boundaries between trusted interactions and fraudulent schemes, creating unique vulnerabilities in user risk assessment capabilities.

Recent behavioral finance research has enriched our understanding of fraud vulnerability. Purohit et al. (2022) show that cognitive biases, such as overconfidence and confirmation bias, play crucial roles in fraud susceptibility. Juniper Research (2020) reported that businesses in e-commerce, airline ticketing, money transfer, and banking services will cumulatively lose over $200 billion to online payment fraud between 2020 and 2024, highlighting the significance of these psychological factors. Ali et al. (2020) indicate that users with high technical proficiency in digital payments often overestimate their ability to detect fraudulent schemes, creating a dangerous gap between perceived and actual risk assessment capabilities.

Influence Mechanisms of Online Activities on Payment Security Psychology

The psychological mechanisms underlying fraud vulnerability have been extensively studied in recent years. Researchers identify three vital cognitive processes that influence fraud susceptibility: attention allocation during risk assessment (Wiczorek & Manzey, 2014), emotional responses to social trust signals (Ocampo, 2024), and decision-making under time pressure (Butavicius et al., 2022). According to Ordóñez et al. (2015), these processes interact with the characteristics of the digital environment to create unique vulnerabilities that fraudsters increasingly exploit. Jones et al. (2019) further emphasize that these cognitive mechanisms are particularly prone to manipulation in digital payment contexts where visual and tactile security cues are often absent.

Young people’s susceptibility to online payment fraud emerges from their distinct patterns of digital engagement. Studies (Sun et al., 2024) document various online activities that shape vulnerability to fraud schemes. Ng et al. (2018) demonstrate that social media exposure heavily influences risk perception among young adults. FTC (2023b) reports show that users aged 18 to 29 are disproportionately targeted through social media platforms, with nearly 40% of fraud losses in this demographic originating from social media contacts.

The relationship between online activities and fraud vulnerability is complex and multifaceted. Research (H. Chen et al., 2017) analyzes how different types of social media engagement influence fraud risk. Their findings indicate that platforms designed for self-presentation can inadvertently provide fraudsters with valuable information for targeting potential victims. Wei et al. (2021) further observed that young users’ risk perception varies significantly based on their frequency and patterns of digital platform usage. However, these same platforms also serve as channels for sharing fraud prevention information and warnings, creating what Cao et al. (2018) term the “dual nature of social media in fraud prevention.”

Information acquisition behaviors are crucial in shaping young people’s ability to identify and resist fraudulent schemes. Research (Yu et al., 2023) reveals that how young users seek and process information online significantly influences their capacity to recognize fraud indicators. According to Harrison et al. (2016), user-centered approaches to phishing susceptibility highlight the importance of suspicious personality traits in protecting against fraud. Their study identifies specific patterns in information consumption that either enhance or diminish fraud awareness, highlighting the importance of critical information processing skills in fraud prevention.

Commercial transactions in digital environments present particular challenges for fraud prevention. Studies (Zhang et al., 2019) demonstrate that young users often develop habitual patterns in their payment behaviors, potentially reducing their attention to security cues. Integrating social and commercial functions in digital platforms has created what (Mombeuil & Uhde, 2021) describe as “friction-free fraud opportunities,” where the seamless nature of modern payment systems can mask fraudulent intentions.

Social context is crucial in shaping young people’s payment security behavior. Studies by Mwagwabi and Hee Jiow (2021) reveal that peer influence significantly affects risk assessment and security decisions. Their research shows that young users often modify their security practices based on social norms and peer experiences, creating opportunities and challenges for fraud prevention efforts. Alqarni et al. (2016) further observed that users connected primarily with strangers on social networks demonstrate lower perceptions of severity and vulnerability to potential phishing attacks, suggesting that social connections influence security risk perceptions.

These studies suggest that various online activities shape young people’s payment security psychology through various psychological mechanisms. Understanding these mechanisms is crucial for developing effective fraud prevention strategies, as they suggest that different forms of online engagement may require targeted security interventions.

Demographic Heterogeneity in Fraud Prevention

Research has identified significant variations in fraud vulnerability and prevention capabilities across demographic groups. Studies by A. Chen and Arkansas (2023) document gender differences in fraud targeting and response patterns, finding that male and female users exhibit distinct vulnerabilities to fraudulent schemes. According to Anwar et al. (2017), “results show gender-wise differences for cybersecurity self-efficacy and behavior,” with recommendations for training to close these gaps. Whitty (2020) found that women were much more likely to be victims of consumer scams, while men were more likely to be victims of investment scams. This suggests that these differences may stem from differing social roles and consumption patterns. These gender-based variations are often shaped by different social networks and communication patterns that affect how fraud prevention information is processed and internalized.

The urban-rural divide introduces another layer of complexity to fraud prevention. Research by Li et al. (2024) demonstrates that differences in digital infrastructure access, exposure to fraud prevention education, and online payment patterns significantly influence vulnerability to fraud. According to a study by the National Telecommunications and Information Administration [NTIA], 2016), rural individuals with lower family income or education face “an even larger disadvantage” in digital adoption than their urban counterparts. Their research reveals how geographic and socioeconomic factors shape exposure to fraud attempts and the effectiveness of prevention strategies. This geographic disparity is further complicated by cultural differences in trust formation and risk perception, which Zhao and Pan (2023) find particularly relevant in increasingly globalized digital payment environments.

Educational background emerges as a critical factor in fraud prevention capability. Studies (Liu et al., 2023) find that formal education levels influence how individuals process and respond to fraud-related information. However, their research also reveals that the relationship between education and fraud prevention is not straightforward. While higher education levels generally correlate with better information processing capabilities, research (Kim, Tao et al., 2010) shows that educational background alone does not guarantee enhanced fraud resistance. This complexity is further illustrated by Research (Hameed et al., 2024), which indicates that social learning mechanisms vary significantly across demographic groups in developing fraud prevention capabilities, suggesting the need for targeted approaches in fraud prevention education and awareness programs that account for variations in learning styles and information access.

Research Hypotheses

The literature review reveals the complex relationship between online activities and payment security psychology. While risk perception theory provides a basic framework, the digital environment brings new phenomena that traditional theories cannot fully explain. Schneier pointed out the disconnect between perceived and actual security, and Davinson and Sillence discovered the paradox that frequency of use may actually reduce risk vigilance. Both indicate that risk perception formation in digital payment environments has its unique characteristics.

The literature review similarly indicates that young people’s different online behaviors shape their payment security awareness differently. Social interactions play a dual role in information sharing, serving as both channels for security knowledge dissemination and avenues for fraudsters to contact potential victims; self-presentation activities allow young people to showcase themselves while potentially becoming aware of the value and risks of personal information; active searching and learning help cultivate critical thinking and enhance the ability to identify fraud; while routine payment operations may become mechanical due to repetition, reducing sensitivity to risks. More interestingly, these influences vary according to individual backgrounds—gender role socialization creates different risk assessment patterns for men and women, the digital divide exposes urban and rural residents to different risk environments, and education levels affect information processing and risk judgment capabilities.

Although existing research has separately explored the influence of specific types of online activities or certain demographic factors, there is a systematic attempt to integrate these factors. Addressing this research gap, we propose the following hypotheses:

H1: Different types of online activities have varying effects on young people’s payment security psychology.

H2: The influence of online activities on payment security psychology varies by gender.

H3: The effect of online activities on payment security psychology differs between urban and rural residents.

H4: The impact of online activities on payment security psychology varies across educational levels.

By testing these hypotheses, we hope to uncover the intrinsic mechanisms of payment security psychology formation in digital environments, provide new perspectives for understanding how young people develop security awareness in the complex online world, and establish an empirical foundation for designing effective fraud prevention strategies targeted at different groups.

Research Design

Data Source and Sample

This study uses data from the Chinese General Social Survey (CGSS) 2017. Established in 2003, CGSS is China’s earliest national, comprehensive, and longitudinal academic survey project. We selected the 2017 data for several reasons: first, it is the latest complete dataset released by CGSS that contains detailed information on online activities and payment security; second, although there is a time gap between data collection and the present, this study focuses on the fundamental mechanisms of how different types of online activities affect payment security psychology, which exhibit relative stability; third, while specific technologies and platforms may change over time, users’ basic interaction patterns with digital environments remain relevant, as related research indicates that the basic psychological processes of digital risk perception remain relatively stable in the short term (Addae et al., 2017; van Schaik et al., 2017).

The original CGSS2017 sample included 12,582 respondents from various provinces and cities across China. Based on our research objectives, we applied the following systematic screening criteria: (1) following the youth definition standard of China’s National Bureau of Statistics, we retained only samples aged between 15 and 34 years, which reflects the official definition of youth groups in China and is suitable for studying young people’s behavior in the Chinese context; (2) excluded all non-internet users; (3) eliminated cases with missing values on key variables (types of online activities and payment security psychology measurements). This systematic screening reduced the sample size from 12,582 to a final adequate sample of 557 observations. Despite the reduction in sample size, our analytical sample maintained a balance across major demographic characteristics (such as gender, education, and regional distribution), ensuring the validity and interpretability of our analytical results.

Variable Measurement

Our dependent variable—payment security psychology—was measured through CGSS2017 item C416, which asked respondents: “When making online payments or transactions, I observe the environment to determine whether to use it.” Responses were recorded on a five-point Likert scale, with higher values indicating stronger security awareness (1 = strongly disagree, 5 = strongly agree). Although, this is a single-item measure, it directly captures users’ risk assessment behavioral tendencies in payment environments, which has particular value for our research. First, observing and evaluating environmental risks is the cognitive prerequisite and decision-making foundation for anti-fraud behavior, reflecting an individual’s vigilance against fraud. Second, this measurement item focuses on “observing the environment and deciding whether to use it” as a behavioral intention rather than merely remaining at the cognitive level of risk perception. Third, compared to specific technical protective measures (such as setting payment limits or using two-factor authentication), this tendency to observe and evaluate environments better reflects an individual’s enduring security psychological traits unaffected by specific platform technological limitations. Therefore, this measurement item captures the core behavioral manifestation of users’ payment security psychology, encompassing both the cognitive component of security awareness and the intentional component of anti-fraud behavior.

The independent variables include six types of online activities measured through items C42.1 to C42.6 in CGSS2017. Each activity was rated on a five-point Likert scale indicating participation frequency (1 = almost never, 5 = very frequently). These activities include: social networking activities through platforms such as email, QQ, and WeChat (C42.1); self-presentation activities on social media (C42.2); online rights protection actions (C42.3); leisure and entertainment activities such as gaming and video watching (C42.4); information acquisition activities through search and news browsing (C42.5); and commercial transactions including online payments and shopping (C42.6). It is worth noting that these six types of online activities, while related, are conceptually distinct. Social networking activities primarily involve interactive communication and establishing and maintaining interpersonal relationships.

In contrast, self-presentation activities refer to actively displaying personal images and sharing personal life experiences and viewpoints on social media, emphasizing self-expression and image construction. Although both activities occur on social platforms, the former focuses on interaction processes while the latter emphasizes self-presentation. Online rights protection behaviors refer to actions taken online to safeguard one’s rights and interests, such as complaining about product and service issues or participating in rights protection discussions and actions, reflecting users’ proactivity in protecting their interests in digital environments. Information acquisition activities concentrate on searching and browsing news to obtain information, while leisure and entertainment activities include gaming and video watching primarily for recreational purposes; these two types of activities have obvious distinctions in purpose and content. Commercial transaction activities refer to behaviors directly involving economic exchanges, such as online payments and shopping. These activity categorizations are not only conceptually distinct from each other but also clearly defined in the CGSS questionnaire design, ensuring measurement clarity.

This study incorporates several categories of control variables to account for potential confounding effects, selected based on established theories and empirical evidence in risk perception and security behavior research. The first category covers demographic factors, including gender, age and its squared term, ethnicity, religious belief, education level, logarithm of labor income, party membership, health status, household registration, and region. Gender is included as prior research indicates significant differences in risk perception and security behavior between males and females; age and its squared term capture potential non-linear lifecycle effects on security awareness; ethnicity and religious belief may influence cultural value orientations that affect risk attitudes; education level directly impacts digital literacy and information processing capabilities critical for security evaluation; income level reflects economic resources that may influence perceived financial risk and security concerns; party membership captures potential institutional influence on risk attitudes; health status may affect risk tolerance; household registration accounts for urban-rural differences in digital infrastructure access and exposure to digital payment systems; and regional variables control for geographic variations in digital payment adoption and fraud exposure patterns. Specifically, gender is coded as male = 1, female = 0; ethnicity is coded as Han = 1, minority = 0; religious belief is coded as yes = 1, no = 0; education level is divided into four categories: 1 = primary school and below, 2 = junior high school, 3 = high school/vocational high school/technical school, 4 = college and above; household registration is coded as non-agricultural Hukou = 1, agricultural Hukou = 0; and region is divided into five categories: Eastern, Northwestern, Southwestern, Central, and Northeastern China. The second category includes family characteristics, including parents’ party membership and father’s work unit type. This may influence individuals’ values and risk attitudes through intergenerational security awareness and institutional trust transmission. The third category involves behavioral characteristics, including social interaction frequency and political participation, which reflect individuals’ social capital and civic engagement that may shape information access and risk perceptions related to payment security psychology.

Analytical Method

Considering the ordinal nature of the dependent variable, we employed ordered Probit regression models for analysis (Daykin & Moffatt, 2002). This model was chosen because it effectively handles ordinal categorical dependent variables and can estimate the impact of different explanatory variables on the probability distribution of the dependent variable, which is particularly suitable for analyzing ordinal attitude variables such as security psychology.

We conducted parallel slopes assumption tests using the omodel command to assess model applicability. To ensure the robustness of the results, we also employed ordered logit models for supplementary analysis. Additionally, we checked for multicollinearity among all variables to ensure the reliability of model estimates.

We conducted group analyses based on important demographic and socioeconomic characteristics to explore potential heterogeneity in how different online activities affect payment security psychology. Specifically, we performed separate ordered Probit regression analyses based on gender (male/female), Hukou type (agricultural/non-agricultural), and education level (below college/college and above). This analytical framework enables us to systematically examine how online activities affect young people’s payment security psychology while exploring how these effects vary across different demographic and socioeconomic backgrounds.

Robustness Test Methods

We employed two testing methods to assess the robustness of our research findings. First, we used the ordered logit regression model instead of the ordered probit model used in the main analysis to test whether the results were sensitive to model selection.

Second, we tested with an alternative dependent variable. Specifically, we used item C414 from the CGSS2017 database: “When I see important messages forwarded by people around me online (such as on WeChat or Weibo), I verify before believing.” We named this item “network verification behavior” because it reflects individuals’ security awareness in the online environment, like payment security attitude. After adjusting this item for directionality, we conducted an Ordered Logit regression analysis using the same independent and control variables as in the main analysis.

Through these two robustness testing methods, we can evaluate the reliability and consistency of our research findings, ensuring that the results are not affected by specific models or variable measurement choices.

Empirical Results

Descriptive Statistics

Table 1 presents the descriptive statistics for all variables. The mean score for payment security psychology is 4.314 (SD = 0.894) on a five-point scale, generally indicating high-security consciousness among young users in our sample. Among the six types of online activities, social activities show the highest frequency (Mean = 4.020, SD = 0.911), followed by leisure entertainment (Mean = 3.765, SD = 0.906) and information acquisition (Mean = 3.763, SD = 0.904). Self-presentation activities (Mean = 3.188, SD = 1.038) and online rights protection actions (Mean = 2.245, SD = 1.031) demonstrate relatively lower frequencies.

Table 1.

Descriptive Statistics.

Variable type	Variable	Variable description	Mean	Standard deviation	Minimum value	Maximum value
Dependent variable	Secure payment psychology	1–5 indicates an increasing sense of security	4.314	0.894	1.000	5.000
Independent variable	Social activities	1–5 indicates an increasingly frequent online lifestyle	4.020	0.911	1.000	5.000
	Self-presentation		3.188	1.038	1.000	5.000
	Online actions		2.245	1.031	1.000	5.000
	Leisure entertainment		3.765	0.906	1.000	5.000
	Information acquisition		3.763	0.904	1.000	5.000
	Business transactions		3.660	0.987	1.000	5.000
Control variable category A	Gender	0 = Female, 1 = Male			0.000	1.000
	Age	Measured in Years	26.946	4.674	18.000	34.000
	Age squared	Age²	747.930	247.468	324.000	1,156.000
	Ethnicity	1 = Han Chinese, 0 = Other			0.000	1.000
	Religious belief	0 = No religious belief, 1 = Has religious belief			0.000	1.000
	Educational level	1 = Elementary school or below, 2 = Junior high school, 3 = Senior high school (vocational school), 4 = University or above	3.134	0.970	1.000	4.000
	Logarithm of labor income	The logarithmic value of labor income	7.810	4.726	0.000	13.816
	Party membership	0 = Non-member, 1 = Member			0.000	1.000
	Health status	1 to 4 indicates increasing healthiness	3.140	0.817	1.000	4.000
	Household registration	0 = Agricultural, 1 = Non-agricultural			0.000	1.000
	Region	1 = Eastern, 2 = Northwestern, 3 = Southwestern, 4 = Central, 5 = Northeastern			1.000	5.000
Control variable category B	Father’s party membership	0 = Non-member, 1 = Member			0.000	1.000
	Mother’s party membership	0 = Non-member, 1 = Member			0.000	1.000
	Father’s work unit type	0 = Non-state unit, 1 = State unit			0.000	1.000
Control variable category C	Social interaction frequency	1 to 3 indicates increasing frequency	1.847	0.791	1.000	3.000
Control variable category C	Political voting	0 = No, 1 = Yes			0.000	1.000

For control variables, the average age of respondents is 26.946 years (SD = 4.674), suggesting a balanced representation of young adults in our sample. Educational attainment shows a mean of 3.134 (SD = 0.970), indicating that the average education level is around senior high school or vocational school. The logarithm of labor income shows considerable variation (Mean = 7.810, SD = 4.726), reflecting diverse economic conditions among young respondents.

Main Effects Analysis

To examine the associations between variables and potential multicollinearity issues, we conducted correlation analysis and VIF tests. The results indicate that while some online activities show expected correlations, all VIF values fall within acceptable ranges, suggesting no severe multicollinearity problems in our model.

We conducted parallel slopes assumption tests using the omodel command, and the results showed a certain degree of violation (χ²(63) = 97.19, p = .0037), indicating that the effects of different explanatory variables on various categories of the dependent variable may differ. However, considering that our main research purpose is to explore the overall direction and significance of the impact of different online activities on payment security psychology, rather than precisely predicting marginal probability changes between categories, the ordered Probit model can still provide valuable insights. As Greene and Hensher (2010) pointed out, even when the parallel slopes assumption is statistically significantly violated in some research contexts, its substantive impact may be relatively limited, especially when the research focus is on the direction and significance of variable relationships rather than precise prediction. To further ensure the robustness of the results, we also used the ordered logit model for supplementary analysis, and the results showed that the substantive findings of the two models were highly consistent, further supporting our research conclusions.

Table 2 presents the ordered Probit regression results, beginning with a baseline model containing only control variables. We then progressively introduce online activity variables both individually and collectively. The analysis reveals several significant patterns consistent with our theoretical expectations.

Table 2.

Ordered Probit Regression Results.

	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
Variable	Null model	Social activities	Self-presentation	Online actions	Leisure entertainment	Information acquisition	Business transactions	Full model
Gender	0.260*(0.102)	0.286**(0.103)	0.327**(0.104)	0.268**(0.102)	0.264**(0.102)	0.255*(0.102)	0.267**(0.102)	0.324**(0.105)
Age	0.205(0.135)	0.195(0.137)	0.212(0.136)	0.215(0.136)	0.214(0.136)	0.196(0.136)	0.210(0.136)	0.192(0.137)
Age squared	−0.004+(0.003)	−0.004(0.003)	−0.004+(0.003)	−0.004+(0.003)	−0.004+(0.003)	−0.004(0.003)	−0.004+(0.003)	−0.004(0.003)
Ethnicity	−0.073(0.199)	−0.068(0.200)	−0.053(0.200)	−0.068(0.200)	−0.063(0.199)	−0.084(0.200)	−0.071(0.199)	−0.072(0.202)
Religious belief	−0.222(0.183)	−0.240(0.184)	−0.200(0.184)	−0.200(0.184)	−0.212(0.184)	−0.220(0.184)	−0.219(0.183)	−0.224(0.186)
Educational level	0.164*(0.064)	0.127+(0.066)	0.135*(0.066)	0.147*(0.065)	0.143*(0.066)	0.115+(0.067)	0.139*(0.067)	0.128+(0.069)
Logarithm of labor income	−0.004(0.012)	−0.005(0.013)	−0.009(0.013)	−0.007(0.013)	−0.005(0.012)	−0.006(0.012)	−0.007(0.013)	−0.007(0.013)
Party membership	−0.170(0.212)	−0.192(0.213)	−0.122(0.214)	−0.137(0.213)	−0.155(0.213)	−0.153(0.213)	−0.149(0.213)	−0.176(0.215)
Health status	0.003(0.063)	−0.010(0.063)	−0.018(0.063)	0.004(0.063)	−0.002(0.063)	−0.005(0.063)	−0.003(0.063)	−0.014(0.064)
Household registration	0.370**(0.117)	0.397***(0.118)	0.381**(0.118)	0.382**(0.117)	0.378**(0.117)	0.398***(0.118)	0.380**(0.117)	0.410***(0.119)
2. Region (Northwestern)	0.101(0.202)	0.181(0.204)	0.181(0.205)	0.100(0.202)	0.119(0.203)	0.164(0.205)	0.132(0.205)	0.198(0.208)
3. Region (Southwestern)	−0.453**(0.144)	−0.435**(0.144)	−0.466**(0.144)	−0.457**(0.144)	−0.444**(0.144)	−0.429**(0.144)	−0.438**(0.145)	−0.469**(0.146)
4. Region (Centra)	0.069(0.136)	0.129(0.138)	0.036(0.138)	0.074(0.137)	0.069(0.137)	0.102(0.138)	0.086(0.138)	0.090(0.141)
5. Region (Northeastern)	−0.353+(0.195)	−0.294(0.197)	−0.298(0.197)	−0.307(0.197)	−0.313(0.197)	−0.281(0.197)	−0.297(0.199)	−0.298(0.200)
Father’s party membership	−0.026(0.189)	−0.025(0.189)	−0.021(0.190)	−0.030(0.189)	−0.020(0.189)	−0.036(0.189)	−0.033(0.189)	−0.025(0.191)
Mother’s party membership	0.155(0.289)	0.080(0.291)	0.124(0.292)	0.189(0.291)	0.158(0.290)	0.186(0.292)	0.151(0.289)	0.111(0.296)
Father’s work unit type	0.179(0.189)	0.157(0.189)	0.202(0.191)	0.164(0.190)	0.182(0.189)	0.137(0.190)	0.177(0.189)	0.143(0.193)
Social interaction frequency	0.043(0.066)	0.003(0.067)	0.016(0.067)	0.033(0.066)	0.032(0.067)	0.025(0.067)	0.034(0.066)	−0.006(0.068)
Political voting	0.210+(0.118)	0.216+(0.119)	0.182(0.119)	0.200+(0.119)	0.203+(0.119)	0.203+(0.119)	0.207+(0.119)	0.190(0.120)
Social activities		0.214***(0.058)						0.188**(0.067)
Self-presentation			0.195***(0.052)					0.137*(0.060)
Online actions				0.099*(0.049)				0.053(0.055)
Leisure entertainment					0.057(0.060)			−0.047(0.071)
Information acquisition						0.146*(0.060)		0.092(0.076)
Business transactions							0.052(0.056)	−0.096(0.067)
N	557	557	557	557	557	557	557	557
Pseudo R²	.055	.066	.067	.058	.055	.059	.055	.076

Note. Standard errors in parentheses.

p < .10. *p < .05. **p < .01. ***p < .001.

Social activities and self-presentation emerge as the most influential factors, maintaining significant positive effects in individual and complete models. Specifically, social activities show a robust positive effect (β = .188, p < .01) in the full model, while self-presentation activities demonstrate a significant positive impact (β = .137, p < .05). These findings suggest that social networking and self-presentation engagement enhance young people’s consciousness of payment security.

Online rights protection actions and information acquisition show moderate effects on payment security psychology. Online actions display a positive but marginally significant effect (β = .053, p < .10) in the individual model, though this effect weakens in the entire model. Information acquisition similarly demonstrates some influence on security consciousness, but its effect becomes non-significant when controlling for other online activities.

Notably, leisure entertainment and business transactions do not significantly affect payment security psychology in individual or full models. This finding differs from the common expectation that greater engagement in entertainment or commercial activities would increase security awareness.

Among control variables, gender (β = .324, p < .01), educational level (β = .128, p < .10), and household registration (β = .410, p < .001) consistently show significant effects across all models, indicating strong demographic influences on payment security psychology.

Heterogeneity Analysis

Following the primary analysis, we conduct heterogeneity tests to examine how online activities’ effects vary across different demographic groups. Table 3 presents the regression results for subgroup analyses based on gender, household registration status, and educational background.

Table 3.

Heterogeneity Analysis Ordered Probit Regression Results.

	Gender differences		Household registration differences		Educational differences
Variable	(1) Female	(2) Male	(3) Agricultural	(4) Other	(5) Below university level	(6) University level and above
Social activities	0.180+ (0.099)	0.163+ (0.097)	0.270** (0.088)	0.069 (0.112)	0.143+ (0.086)	0.273* (0.117)
Self-presentation	0.125 (0.089)	0.298*** (0.088)	0.120 (0.083)	0.154+ (0.090)	0.119 (0.083)	0.108 (0.093)
Online actions	0.219** (0.079)	−0.179* (0.083)	−0.019 (0.076)	0.127 (0.084)	−0.073 (0.077)	0.120 (0.088)
Leisure entertainment	−0.084 (0.105)	0.010 (0.102)	−0.103 (0.096)	0.057 (0.120)	−0.043 (0.093)	−0.073 (0.119)
Information acquisition	−0.080 (0.110)	0.277* (0.113)	0.056 (0.103)	0.186 (0.120)	0.213* (0.102)	0.023 (0.131)
Business transactions	−0.054 (0.099)	−0.158 (0.100)	−0.099 (0.087)	−0.097 (0.115)	0.007 (0.083)	−0.272* (0.125)
Age	0.602** (0.211)	−0.125 (0.194)	0.080 (0.182)	0.354 (0.228)	0.214 (0.175)	0.291 (0.252)
Age squared	−0.011** (0.004)	0.001 (0.004)	−0.002 (0.003)	−0.006 (0.004)	−0.005 (0.003)	−0.005 (0.005)
Ethnicity	0.062 (0.266)	−0.394 (0.350)	−0.381 (0.256)	0.694+ (0.380)	−0.260 (0.267)	0.148 (0.333)
Religious belief	−0.408 (0.254)	0.099 (0.301)	−0.231 (0.236)	−0.142 (0.336)	−0.216 (0.237)	−0.425 (0.317)
Educational level	0.109 (0.100)	0.204* (0.101)	0.143 (0.093)	0.132 (0.116)
Logarithm of labor income	−0.006 (0.017)	0.021 (0.022)	−0.005 (0.017)	−0.012 (0.020)	−0.005 (0.017)	−0.004 (0.021)
Party membership	−0.201 (0.312)	−0.018 (0.312)	−0.324 (0.391)	−0.122 (0.275)	−0.322 (0.667)	−0.165 (0.241)
Health status	−0.040 (0.099)	−0.095 (0.093)	−0.033 (0.085)	−0.028 (0.101)	−0.095 (0.086)	0.020 (0.102)
Household registration	0.484** (0.180)	0.284+ (0.169)			0.457** (0.172)	0.413* (0.188)
2. Region (Northwestern)	0.019 (0.299)	0.632* (0.313)	0.019 (0.259)	0.647 (0.428)	0.204 (0.268)	0.148 (0.368)
3. Region (Southwestern)	−0.341 (0.209)	−0.613** (0.214)	−0.634** (0.203)	−0.332 (0.233)	−0.534* (0.208)	−0.557* (0.222)
4. Region (Centra)	0.067 (0.216)	0.158 (0.198)	0.013 (0.179)	0.281 (0.263)	−0.162 (0.182)	0.510+ (0.265)
5. Region (Northeastern)	−0.466+ (0.277)	−0.122 (0.319)	−0.507 (0.321)	0.020 (0.276)	−0.587* (0.277)	−0.038 (0.322)
Father’s party membership	−0.365 (0.308)	0.370 (0.273)	0.190 (0.297)	−0.207 (0.276)	0.224 (0.314)	−0.093 (0.256)
Mother’s party membership	0.023 (0.400)	0.272 (0.523)	0.080 (0.578)	0.141	0.288 (0.591)	0.256 (0.352)
	0.023 (0.400)	0.272 (0.523)	0.080 (0.578)	(0.355)	0.288 (0.591)	0.256 (0.352)
Father’s work unit type	0.182 (0.300)	0.230 (0.270)	0.086 (0.368)	0.167 (0.254)	0.202 (0.358)	0.034 (0.244)
Social interaction frequency	−0.011 (0.099)	−0.047 (0.099)	−0.026 (0.095)	−0.001 (0.108)	0.000 (0.095)	−0.088 (0.103)
Political voting	−0.080 (0.178)	0.408* (0.178)	0.124 (0.158)	0.389+ (0.204)	0.280+ (0.156)	0.015 (0.203)
Gender			0.452** (0.144)	0.269 (0.165)	0.331* (0.150)	0.316* (0.158)
N	260	297	287	270	285	272
Pseudo R²	0.090	0.131	0.068	0.089	0.085	0.088

Note. Standard errors in parentheses.

p < .10. *p < .05. **p < .01. ***p < .001.

Gender Differences

The analysis reveals distinct patterns between male and female users. For male users, self-presentation shows the most substantial effect (β = .298, p < .001), followed by information acquisition (β = .277, p < .05). Interestingly, online rights protection actions demonstrate a significant adverse effect (β = −.179, p < .05) for males. In contrast, female users show a different pattern, with online rights protection actions displaying a significant positive effect (β = .219, p < .01), while social activities show marginal significance (β = .180, p < .10). These gender-specific findings suggest that male and female users develop security consciousness through different types of online engagement.

Household Registration Differences

The impact of online activities varies substantially between agricultural and non-agricultural household registration groups. Social activities for users with agricultural household registration demonstrate a powerful positive effect (β = .270, p < .01) on payment security psychology. This effect is notably absent in the non-agricultural group, where the impacts of online activities are more evenly distributed. The differential effects might reflect varying access to digital resources and financial services between urban and rural areas.

Educational Background Differences

The analysis by educational level reveals nuanced patterns. Among users with university education and above, social activities show a robust positive effect (β = .273, p < .05), while business transactions demonstrate a significant adverse effect (β = −.272, p < .05). For users with below-university education, information acquisition emerges as significant (β = .213, p < .05), with social activities showing only marginal significance (β = .143, p < .10). These findings suggest that education level moderates how different online activities influence security consciousness.

Regional Heterogeneity and Other Control Effects

In Table 2, our analysis reveals significant regional variations in payment security psychology. Notably, respondents from southwestern regions show consistently lower levels of security consciousness (β = −.469, p < .01) compared to eastern regions across all demographic groups. This regional effect persists after controlling for various socioeconomic factors, suggesting the influence of broader geographic and regional factors on security awareness.

Various sensitivity tests, including alternative model specifications and variable coding schemes, confirm the robustness of these heterogeneous effects. These results provide strong evidence for the differential impacts of online activities across demographic groups, supporting our hypotheses about heterogeneous effects in security psychology formation.

Robustness Test Results

We employed the Ordered Logit regression model for robustness testing to avoid the limitations of a single regression method. First, we conducted regression analysis using the Ordered Logit model, and the results showed that the coefficient signs and significance were highly consistent with the Ordered Probit regression model used in the main analysis, indicating robust results.

It is worth noting that although the parallel lines assumption of ordered models may be violated in statistical tests, as discussed in our main analysis when the research focuses on the direction and significance of variable relationships rather than precise prediction, the substantive impact of such violations is usually limited. Our robustness test results confirm this point, showing that model selection does not affect the main findings.

In the second step, we still used the Ordered Logit model for regression but replaced the dependent variable of payment security attitude with network verification behavior. After adjusting the directionality of this item on the 5-point Likert scale, we conducted an Ordered Logit regression analysis using the same independent and control variables as in the main analysis. The results showed that the coefficient signs and significance of the variables had considerable overlap with the previous regression results, and the robustness test results after changing the dependent variable were also good. Control variables such as gender, education level, and household registration remained significant.

Furthermore, five of the six independent variables were significant in the single-variable models, with only “leisure entertainment” being non-significant. In the full model, “self-presentation” and “online actions” were significant. Compared with the previous regression results, “self-presentation” remained highly significant regardless of how the dependent variable changed. This may be because “self-presentation” emphasizes individual behavior, while other variables more often reflect interactions with others. Emotions tend to be more stable when alone, and payments are made more cautiously. However, once communications, interactions, and business discussions with others are involved, young people’s payment security attitudes and network verification behavior are both affected, showing as non-significant in the data results. This also indirectly indicates that young people are very susceptible to the impact and influence of changes in the external environment.

Summary of Hypothesis Testing

Our empirical analysis provides comprehensive support for all four research hypotheses. H1 was supported through our main effects analysis, which revealed that different online activities showed varying effects on payment security psychology, with social activities and self-presentation emerging as the most influential factors (β = .188, p < .01; β = .137, p < .05, respectively).

The heterogeneity analysis further validated H2-H4: gender differences (H2) were confirmed with distinct patterns between male and female users, where males showed more potent effects in self-presentation (β = .298, p < .001) and information acquisition (β = .277, p < .05), while females demonstrated significant positive effects in online rights protection actions (β = .219, p < .01). Household registration differences (H3) were validated mainly through the strong effect of social activities in the agricultural registration group (β = .270, p < .01). Educational differences (H4) were evidenced by varying impacts across educational levels, notably in social activities for higher education groups (β = .273, p < .05) and information acquisition for lower education groups (β = .213, p < .05).

These findings provide a solid foundation for understanding the complex relationship between young people’s online activities and their payment security psychology and the important role of demographic characteristics in moderating these relationships.

Discussion

Understanding the Differential Effects of Online Activities

The significant positive effects of social activities and self-presentation on payment security psychology suggest that social interaction-based online engagement may enhance security awareness through multiple mechanisms. These activities create opportunities for information sharing about fraud risks and facilitate peer learning about security practices.

Online rights protection actions and information acquisition activities reveal essential insights about security awareness formation while showing moderate effects. Harrison et al. (2016) identified specific patterns in information consumption that either enhance or diminish fraud awareness, highlighting the importance of critical information processing skills in fraud prevention. Although these activities involve active information processing and critical thinking, they contribute to security consciousness, perhaps not as strongly as social interactions. Yu et al. (2023) reinforced our finding, revealing that how young users seek and process information online significantly influences their capacity to recognize fraud indicators. The relatively weaker impact might also indicate that young users tend to engage with security information more actively when embedded in social contexts rather than presented as formal educational content.

The non-significant effects of leisure entertainment and business transactions on security psychology are particularly noteworthy. These findings challenge the conventional assumption that increased exposure to payment scenarios enhances security awareness. Instead, they suggest that routine commercial activities might lead to habituation and reduced vigilance, potentially creating vulnerabilities in fraud prevention. This pattern aligns with risk perception theory, which suggests that familiarity with risky situations can paradoxically reduce risk awareness. Davinson and Sillence (2014) confirmed this paradox, indicating that increased familiarity with digital payment systems can reduce fraud risk awareness. Hossain (2019) similarly found that frequent digital payment users often develop automated behavioral patterns that bypass conscious risk evaluation processes. Zhang et al. (2019) further supported our findings, demonstrating that young users often develop habitual patterns in their payment behaviors, potentially reducing their attention to security cues.

Understanding Demographic Heterogeneity

The significant gender differences in our findings demand attention in fraud prevention efforts. Male users exhibit strong positive effects in self-presentation activities and information acquisition, contrasting with their irrational behavior in specific scenarios, suggesting gender-specific patterns in security awareness formation. A. Chen and Arkansas (2023) documented similar gender differences in fraud targeting and response patterns, finding that male and female users exhibit distinct vulnerabilities to different fraudulent schemes. A paradoxical finding emerges in our research: while males exhibit more vital security consciousness in general online contexts such as self-presentation and information acquisition, they show significant vulnerability in romantic and relationship-based payment scenarios. This pattern reveals that cognitive biases in romantic and emotional contexts can override otherwise rational security awareness among male users.

Female users, in contrast, only show significant positive effects in online rights protection activities, with other variables being non-significant except for social activities at the 0.1 level. These gender differences may stem from distinct thinking patterns between men and women: Females, typically being the primary group in online shopping, tend to rely more on intuition and emotional experience in payment environment assessment due to their habitual instant decision-making formed through frequent online shopping, rather than following strict security evaluation processes; while males generally tend to adopt more rational thinking patterns in payment decisions.

Household registration differences reveal important implications for digital financial inclusion and fraud prevention. Among the six variables, social activities demonstrate the most vital significance for the agricultural household registration group, showing a particularly robust effect compared to those with other household registrations. Li et al. (2024) echoed our findings, demonstrating that differences in digital infrastructure access, exposure to fraud prevention education, and online payment patterns significantly influence vulnerability to fraud. A study by the NTIA (2016) complemented this by noting that rural individuals with lower family income or education levels face “an even larger disadvantage” in digital adoption than their urban counterparts. This geographic disparity is further complicated by cultural differences in trust formation and risk perception, Our findings suggest that as agricultural registration populations engage in more social activities, the impact of their economic constraints becomes more significant, making them more cautious about spending money and more aware of payment security. This pattern highlights the potential of leveraging existing social networks and community structures for fraud prevention, particularly in rural areas where access to formal financial education resources may be limited.

The differences in educational background in our results suggest the need for tailored approaches to security education. Individuals with higher education levels show a significantly more substantial influence from social activities on their payment security psychology compared to those with middle or low levels of education. This generally aligns with education’s promotive role in safety consciousness formation, as those with higher education tend to be more aware of risks during social interactions. However, in business transactions, higher education demonstrates a significant adverse effect, suggesting a “curse of education” where highly educated individuals may become overconfident in their abilities, handling business transactions with reduced vigilance. Ali et al. (2020) similarly found that users with high technical proficiency in digital payments often overestimate their ability to detect fraudulent schemes, creating a dangerous gap between perceived and actual risk assessment capabilities. Information acquisition, by contrast, shows a significant positive impact on payment security psychology among those with middle or low education levels, suggesting less educated individuals can effectively enhance their payment security awareness through daily information gathering and self-education about fraud prevention.

Conclusions and Implications

Drawing from our comprehensive analysis of young people’s online activities and payment security psychology, we reach several vital conclusions with significant implications for fraud prevention practices.

Key Research Findings

Our study provides empirical evidence that different online activities have varying impacts on young people’s payment security psychology. Social networking and self-presentation activities significantly positively affect security consciousness, suggesting that social interaction-based digital engagement is crucial in developing fraud prevention awareness. Information acquisition and online rights protection activities show moderate positive effects, while leisure entertainment and business transactions do not significantly influence security psychology.

The impact of online activities on security consciousness exhibits substantial demographic heterogeneity. Gender differences manifest in both the strength and direction of effects, with males and females showing distinct patterns in how their online activities influence security awareness. The urban-rural divide significantly moderates the relationship between online engagement and security psychology, particularly in the context of social activities. Educational background emerges as another crucial factor, with different educational levels associated with varying mechanisms of security awareness formation.

Theoretical and Practical Implications

These findings advance our theoretical understanding of risk perception in digital payment environments. They challenge the conventional assumption that increased exposure to payment scenarios necessarily enhances security awareness, suggesting instead that the nature of digital engagement matters more than frequency. This perspective aligns with Schneier’s (2008) observation about the fundamental disconnect between perceived and actual security and extends Davinson and Sillence’s (2014) paradoxical finding that familiarity can reduce vigilance. Our findings particularly contribute to risk perception theory by demonstrating how social and self-presentational activities—not typically associated with financial risk assessment—significantly shape security consciousness, offering a more nuanced framework for understanding how risk perceptions form in digital environments.

The observed heterogeneity effects across demographic groups (gender, urban-rural, education level) extend existing risk perception theories, providing new perspectives on understanding the formation mechanism of payment security psychology.

From a practical perspective, our findings offer significant insights for enhancing fraud prevention. At the core of effective prevention lies the power of social networks and peer-learning mechanisms. Our research suggests that social media platforms, which young people frequently use, could be strategically utilized to disseminate security information.

Our findings suggest specific implementation strategies for financial institutions and payment platforms. First, transaction interfaces should be designed to counter habituation effects observed in routine payments by incorporating dynamic security features that maintain user vigilance. Such features could include contextual security reminders based on transaction patterns or adaptive verification procedures that engage users differently across transactions. As Hossain (2019) noted, frequent digital payment users often develop automated behavioral patterns that bypass conscious risk evaluation processes, making these dynamic features particularly important. Second, rather than offering uniform security education, platforms should develop demographically targeted security awareness programs—such as reality-based case studies for male users to address their vulnerability in emotional scenarios as identified by Whitty (2020), and structured security assessment frameworks for female users to enhance their systematic risk evaluation during online shopping.

The substantial demographic heterogeneity revealed in our study calls for tailored prevention strategies. Gender-specific approaches need to account for the distinct ways male and female users develop security awareness, particularly addressing males’ vulnerability in romantic scenarios. Educational institutions could implement these insights by developing gender-differentiated financial security curricula that reflect the gender-wise differences in cybersecurity self-efficacy and behavior documented by Anwar et al. (2017). These programs could be integrated into existing digital literacy education, with specialized modules addressing the specific vulnerabilities identified in our research.

In rural areas with limited digital infrastructure, community-based programs can leverage the solid social network effects we observed. Policymakers could implement rural financial security initiatives that account for the digital infrastructure access differences highlighted by Li et al. (2024), using community gathering points and local social networks for disseminating fraud prevention knowledge. Such programs might include village-level digital payment security workshops facilitated by community leaders and leveraging existing social networks. Additionally, mobile-based alerts designed explicitly for rural communities could capitalize on the strong influence of social activities in these areas, as identified in our research.

Different educational backgrounds demand varying communication approaches—while informal learning channels prove effective for less educated users, specific strategies are needed to address the overconfidence issues among highly educated individuals. Educational approaches for less educated individuals could emphasize practical, scenario-based learning distributed through daily information channels they already use, as suggested by our findings on information acquisition activities. For highly educated users, prevention programs should incorporate approaches that address the overconfidence identified by Ali et al. (2020), where users with high technical proficiency often overestimate their ability to detect fraudulent schemes.

Payment platforms should consider implementing dynamic security features to counter the habituation effects observed in routine transactions. Specifically, payment interfaces could incorporate variable security mechanisms that address what Zhang et al. (2019) described as habitual patterns in payment behaviors. These mechanisms could include (1) contextual anomaly notifications that alert users to deviations from their typical transaction patterns, (2) progressive security measures that adjust verification requirements based on transaction history and risk levels, and (3) engagement techniques that maintain user attention during routine transactions. These dynamic features could be particularly effective for business transactions, where our study found limited natural security awareness development.

The effectiveness of fraud prevention ultimately depends on its integration with local contexts and demographic characteristics. Implementation strategies should follow a multi-phase approach that accounts for the demographic heterogeneity observed in our study: first, conduct demographic segmentation of the target population; second, develop tailored intervention materials for each segment as suggested by our findings on varying effects across gender, household registration, and education levels; third, leverage existing community structures for distribution, particularly important for agricultural registration populations where social activities showed the most substantial effect; fourth, establish continuous feedback mechanisms to evaluate and refine approaches. This structured implementation pathway transforms abstract research findings into concrete fraud prevention practices that respond to Wang et al.’s (2019) observation that the digital environment creates unique challenges for risk assessment.

Limitations and Future Research

Methodological Considerations

Based on CGSS2017 data, our analysis provides robust evidence for understanding the formation of payment security psychology. However, the cross-sectional nature of the data limits our ability to observe the dynamic evolution of security awareness over time, making it difficult to establish causality and only allowing for correlation analysis. This directly affects the depth of explanation regarding the mechanism of security awareness formation. Future research could benefit from longitudinal designs that track how security consciousness develops as young people’s online activities evolve.

Additionally, while our ordered Probit model effectively captures the nature of security psychology, the partial violation of the parallel slopes assumption may affect the precision of results, particularly in predicting marginal probability changes. Although the robustness test using ordered logit models confirmed the consistency of our main findings, this technical limitation suggests that the impacts of different online activities on various levels of payment security psychology may vary.

Measurement Refinement

Though theoretically grounded, the measurement of payment security psychology relies on a single item, which limits our grasp of the multidimensional aspects of security awareness and may introduce measurement error, affecting the precision of results. Another limitation is that the measurement focuses on awareness rather than actual behavior, making it difficult to distinguish between the two. Future studies might explore additional dimensions of security psychology, such as automatic risk assessment processes or emotional responses to potential fraud risks, providing a more comprehensive picture of security psychology.

Potential Omitted Variables

This study faces potential omitted variable issues, such as failing to measure digital literacy levels, past fraud experiences, and psychological traits like risk aversion tendencies. These omissions may explain the model’s low explanatory power and could lead to estimation bias. Sample representativeness is also a limitation, as reducing from the original sample to the final sample may introduce selection bias, affecting the generalizability of the research findings.

Future Research Directions

Our findings suggest several future research directions: exploring the specific mechanisms by which social activities enhance security awareness; conducting an in-depth analysis of demographic heterogeneity to understand differences in security awareness formation across groups; investigating the interaction between platform design and user security behavior; developing cross-cultural comparative studies to test the universality of the findings; and evaluating intervention effects based on this research through longitudinal studies. These directions will expand our understanding of young people’s online security behavior and guide the development of more effective fraud prevention strategies.

Footnotes

Author Note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that maybe evaluated in this article, or claim that maybe made by its manufacturer, is not guaranteed or endorsed by the publisher.

ORCID iD

Linlin Bai

Ethical Considerations

Ethical review and approval was not required for the study on human participants in accordance with the local legislation and institutional requirements.

Consent to Participate

Written informed consent from the (patients/participants OR patients/participants legal guardian/next of kin) was not required to participate in this study in accordance with the national legislation and the institutional requirements.

Author Contributions

The author confirms being the sole contributor of this work and has approved it for publication.

Funding

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

Declaration of Conflicting Interests

The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data Availability Statement

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

References

AARP. (2025, January 2). Americans reported losing a record $10 billion to scams and fraud in 2023. https://www.aarp.org/money/scams-fraud/info-2024/2023-ftc-consumer-losses.html

Addae

J. H.

Brown

Sun

Towey

Radenkovic

(2017). Measuring attitude towards personal data for adaptive cybersecurity. Information and Computer Security, 25(5), 560–579. https://doi.org/10.1108/ics-11-2016-0085

Ali

Ally

Clutterbuck

Dwivedi

(2020). The state of play of blockchain technology in the financial services sector: A systematic literature review. International Journal of Information Management, 54, 102199. https://doi.org/10.1016/j.ijinfomgt.2020.102199

Alqarni

Algarni

(2016, June). Toward predicting susceptibility to phishing victimization on Facebook [Conference session]. 2016 IEEE International Conference on Services Computing (SCC). https://doi.org/10.1109/SCC.2016.61

Anwar

Ash

Yuan

(2017). Gender difference and employees’ cybersecurity behaviors. Computers in Human Behavior, 69, 437–443. https://doi.org/10.1016/j.chb.2016.12.040

Balapour

Nikkhah

H. R.

Sabherwal

(2020). Mobile application security: Role of perceived privacy as the predictor of security perceptions. International Journal of Information Management, 52, 102063.

Brenner

Meyll

Stolper

Walter

(2020). Consumer fraud victimization and financial well-being. Journal of Economic Psychology, 76, 102243.

Butavicius

Taib

Han

S. J.

(2022). Why people keep falling for phishing scams: The effects of time pressure and deception cues on the detection of phishing emails. Computers & Security, 123, 102937.

Cao

Liu

Gong

Adeel

(2018). Understanding mobile payment users' continuance intention: A trust transfer perspective. Internet Research, 28(2), 456–476.

10.

Cengiz

A. B.

Kalem

Boluk

P. S.

(2022). The effect of social media user behaviors on security and privacy threats. IEEE Access, 10, 57674–57684.

11.

Chen

Arkansas

(2023). Usage behavior of E-payment services by Chinese millennials (Generation Y) and zoomers (Generation Z). Journal of Business Administration Online, 17(01), 43–67.

12.

Chen

Beaudoin

C. E.

Hong

(2017). Securing online privacy: An empirical test on Internet scam victimization, online privacy concerns, and privacy protection behaviors. Computers in Human Behavior, 70, 291–302.

13.

Chen

(2017). Understanding continuance intention of mobile payment services: An empirical study. Journal of Computer Information Systems, 57(4), 287–298. https://doi.org/10.1080/08874417.2016.1180649

14.

Dai

Palvi

P. C.

(2009). Mobile commerce adoption in China and the United States: A cross-cultural study. ACM SIGMIS Database: The Database for Advances in Information Systems, 40(4), 43–61.

15.

Davinson

Sillence

(2014). Using the health belief model to explore users’ perceptions of ‘being safe and secure’ in the world of technology mediated financial transactions. International Journal of Human-Computer Studies, 72(2), 154–168.

16.

Daykin

A. R.

Moffatt

P. G.

(2002). Analyzing ordered responses: A review of the ordered probit model. Understanding Statistics, 1(3), 157–166.

17.

Featherman

M. S.

Miyazaki

A. D.

Sprott

D. E.

(2010). Reducing online privacy risk to facilitate e-service adoption: The influence of perceived ease of use and corporate credibility. Journal of Services Marketing, 24(3), 219–229. https://doi.org/10.1108/08876041011040622

18.

Federal Trade Commission. (2022, February 22). New data shows FTC received 2.8 million fraud reports from consumers in 2021. https://www.ftc.gov/news-events/news/press-releases/2022/02/new-data-shows-ftc-received-28-million-fraud-reports-consumers-2021-0

19.

Federal Trade Commission. (2023a, February 23). New FTC data show consumers reported losing nearly $8.8 billion to scams in 2022. https://www.ftc.gov/news-events/news/press-releases/2023/02/new-ftc-data-show-consumers-reported-losing-nearly-88-billion-scams-2022

20.

Federal Trade Commission. (2023b, October 6). Social media: A golden goose for scammers. https://www.ftc.gov/news-events/data-visualizations/data-spotlight/2023/10/social-media-golden-goose-scammers

21.

Federal Trade Commission. (2024a, February 8). As nationwide fraud losses top $10 billion in 2023, FTC steps up efforts to protect the public. https://www.ftc.gov/news-events/news/press-releases/2024/02/nationwide-fraud-losses-top-10-billion-2023-ftc-steps-efforts-protect-public

22.

Federal Trade Commission. (2024b, March 10). New FTC data show a big jump in reported losses to fraud to $12.5 billion in 2024. https://www.ftc.gov/news-events/news/press-releases/2025/03/new-ftc-data-show-big-jump-reported-losses-fraud-125-billion-2024

23.

Finucane

M. L.

Alhakami

Slovic

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

24.

Flavián

Guinalíu

(2006). Consumer trust, perceived security and privacy policy: Three basic elements of loyalty to a web site. Industrial Management & Data Systems, 106(5), 601–620. https://doi.org/10.1108/02635570610666403

25.

Gong

Zhang

K. Z.

Chen

Cheung

C. M.

Lee

M. K.

(2020). What drives trust transfer from web to mobile payment services? The dual effects of perceived entitativity. Information & Management, 57(7), 103250.

26.

Greene

W. H.

Hensher

D. A.

(2010). Modeling ordered choices: A primer. Cambridge University Press.

27.

Hameed

Akram

Khan

N. R.

Hameed

(2024). Exploring consumer mobile payment innovations: An investigation into the relationship between coping theory factors, individual motivations, social influence and word of mouth. Journal of Retailing and Consumer Services, 77, 103687.

28.

Handarkho

Y. D.

(2020). Understanding mobile payment continuance usage in physical store through social impact theory and trust transfer. Asia Pacific Journal of Marketing and Logistics, 33(4), 1071–1087.

29.

Harrison

Vishwanath

Rao

(2016, January). A user-centered approach to phishing susceptibility: The role of a suspicious personality in protecting against phishing [Conference session]. 2016 49th Hawaii International Conference on System Sciences (HICSS). https://doi.org/10.1109/HICSS.2016.696

30.

Hossain

M. A.

(2019). Security perception in the adoption of mobile payment and the moderating effect of gender. PSU Research Review, 3(3), 179–190. https://doi.org/10.1108/PRR-03-2019-0006

31.

IPX1031. (2024, December 9). Fraud and identity theft in America 2024 statistics. https://www.ipx1031.com/fraud-and-identity-theft-in-america/

32.

Jones

H. S.

Towse

J. N.

Race

Harrison

(2019). Email fraud: The search for psychological predictors of susceptibility. PLoS One, 14(1), e0209684. https://doi.org/10.1371/journal.pone.0209684

33.

Juniper Research. (2020). Online payment fraud: Emerging threats, segment analysis & market forecasts 2020-2024. https://www.businesswire.com/news/home/20200224005675/en/Juniper-Research-Online-Payment-Fraud-Losses-to-Exceed-$200-Billion-Over-Next-5-Years

34.

Kim

Mirusmonov

Lee

(2010). An empirical examination of factors influencing the intention to use mobile payment. Computers in Human Behavior, 26(3), 310–322.

35.

Kim

Tao

Shin

Kim

K. S.

(2010). An empirical study of customers' perceptions of security and trust in e-payment systems. Electronic Commerce Research and Applications, 9(1), 84–95. https://doi.org/10.1016/j.elerap.2009.04.014

36.

Liébana-Cabanillas

Muñoz-Leiva

Sánchez-Fernández

(2018). A global approach to the analysis of user behavior in mobile payment systems in the new electronic environment. Service Business, 12, 25–64.

37.

Liébana-Cabanillas

Singh

Kalinic

Carvajal-Trujillo

(2021). Examining the determinants of continuance intention to use and the moderating effect of the gender and age of users of NFC mobile payments: A multi-analytical approach. Information Technology and Management, 22(2), 133–161.

38.

Lin

K. Y.

Wang

Y. T.

Huang

T. K.

(2020). Exploring the antecedents of mobile payment service usage: Perspectives based on cost–benefit theory, perceived value, and social influences. Online Information Review, 44(1), 299–318.

39.

(2024). The influence of three digital divide levels on financial advisor demand and engagement among Chinese residents: An investigation based on China. Journal of Innovation & Knowledge, 9(2), 100488.

40.

Liu

(2023). A study of the impact of mobile payment on the enhancement of consumption structure and pattern of Chinese rural households. Agriculture, 13(11), 2082.

41.

Mombeuil

Uhde

(2021). Relative convenience, relative advantage, perceived security, perceived privacy, and continuous use intention of China’s WeChat pay: A mixed-method two-phase design study. Journal of Retailing and Consumer Services, 59, 102384.

42.

Mwagwabi

Hee Jiow

(2021). Compliance with security guidelines in teenagers: The conflicting role of peer influence and personal norms. Australasian Journal of Information Systems, 25, 1–25.

43.

Nandru

Chendragiri

(2025). Factors affecting the adoption of mobile payment services during the COVID-19 pandemic: An application of extended UTAUT2 model. Journal of Science and Technology Policy Management, 16, 405–431.

44.

National Telecommunications and Information Administration (NTIA). (2016). The state of the urban/rural digital divide. https://www.ntia.gov/blog/2016/state-urbanrural-digital-divide

45.

Y. J.

Yang

Z. J.

Vishwanath

(2018). To fear or not to fear? Applying the social amplification of risk framework on two environmental health risks in Singapore. Journal of Risk Research, 21(12), 1487–1501. https://doi.org/10.1080/13669877.2017.1313762

46.

Ocampo

J. A. L.

(2024). A human-computer interaction-driven inquiry on the extent to which web-based trust signals adequately represent the risk of interactions on the web [Doctoral dissertation, Massachusetts Institute of Technology].

47.

Ordóñez

L. D.

Benson

I. I. I.

Pittarello

L. A

. (2015). Time-pressure perception and decision making. The Wiley Blackwell handbook of judgment and decision making, 2, 517–542. https://doi.org/10.1002/9781118468333.ch18

48.

Park

Y. J.

(2013). Digital literacy and privacy behavior online. Communication Research, 40(2), 215–236. https://doi.org/10.1177/0093650211418338

49.

Purohit

Arora

Paul

(2022). The bright side of online consumer behavior: Continuance intention for mobile payments. Journal of Consumer Behaviour, 21(3), 523–542.

50.

Roh

Yang

Y. S.

Xiao

Park

B. I.

(2024). What makes consumers trust and adopt fintech? An empirical investigation in China. Electronic Commerce Research, 24(1), 3–35.

51.

Rouibah

Lowry

P. B.

Hwang

(2016). The effects of perceived enjoyment and perceived risks on trust formation and intentions to use online payment systems: New perspectives from an Arab country. Electronic Commerce Research and Applications, 19, 33–43.

52.

Salah

O. H.

Ayyash

M. M.

(2025). Understanding user adoption of mobile wallet: extended TAM with knowledge sharing, perceived value, perceived privacy awareness and control, perceived security. VINE Journal of Information and Knowledge Management Systems, 55, 1223–1250.

53.

Saridakis

Benson

Ezingeard

J. N.

Tennakoon

(2016). Individual information security, user behaviour and cyber victimisation: An empirical study of social networking users. Technological Forecasting and Social Change, 102, 320–330. https://doi.org/10.1016/j.techfore.2015.08.012

54.

Schneier

(2008). The psychology of security. In Vaudenay

(Ed.), Progress in cryptology – AFRICACRYPT 2008. AFRICACRYPT 2008. Lecture Notes in Computer Science (Vol. 5023, pp. 50–79). Springer.

55.

Slovic

(1987). Perception of risk. Science, 236(4799), 280–285.

56.

Slovic

Finucane

M. L.

Peters

MacGregor

D. G.

(2013). Risk as analysis and risk as feelings: Some thoughts about affect, reason, risk and rationality. In Asveld

Roeser

(Eds.), The feeling of risk (pp. 21–36). Routledge.

57.

Sun

Pan

Yang

Zang

Zhou

(2024). Consumer privacy decision-making in digital era: An integrative review and taxonomy. Aslib Journal of Information Management. Advance online publication. https://doi.org/10.1108/AJIM-12-2023-0526

58.

Wan

Ding

Guo

Chen

(2022). The adoption of QR code mobile payment technology during COVID-19: A social learning perspective. Frontiers in Psychology, 12, 798199.

59.

Upadhyay

Chattopadhyay

(2015). Examining mobile based payment services adoption issues: A new approach using hierarchical clustering and self-organizing maps. Journal of Enterprise Information Management, 28(4), 490–507.

60.

van Schaik

Jeske

Onibokun

Coventry

Jansen

Kusev

. (2017). Risk perceptions of cyber-security and precautionary behaviour. Computers in Human Behavior, 75, 547–559. https://doi.org/10.1016/j.chb.2017.05.038

61.

Vida

Kos Koklič

Kukar-Kinney

Penz

(2012). Predicting consumer digital piracy behavior: The role of rationalization and perceived consequences. Journal of Research in Interactive Marketing, 6(4), 298–313.

62.

Vishwanath

(2015). Habitual Facebook use and its impact on getting deceived on social media. Journal of Computer-Mediated Communication, 20(1), 83–98.

63.

Wang

(2019). Understanding consumers’ willingness to use ride-sharing services: The roles of perceived value and perceived risk. Transportation Research Part C Emerging Technologies, 105, 504–519.

64.

Wei

M. F.

Luh

Y. H.

Huang

Y. H.

Chang

Y. C.

(2021). Young generation’s mobile payment adoption behavior: Analysis based on an extended UTAUT model. Journal of Theoretical and Applied Electronic Commerce Research, 16(4), 618–637. https://doi.org/10.3390/jtaer16040037

65.

Whitty

M. T.

(2020). Is there a scam for everyone? Psychologically profiling cyberscam victims. European Journal on Criminal Policy and Research, 26(3), 399–409. https://doi.org/10.1007/s10610-020-09458-z

66.

Wiczorek

Manzey

(2014). Supporting attention allocation in multitask environments: Effects of likelihood alarm systems on trust, behavior, and performance. Human Factors, 56(7), 1209–1221.

67.

Wiederhold

B. K.

(2014). The role of psychology in enhancing cybersecurity. Cyberpsychology, Behavior & Social Networking, 17(3), 131. https://doi.org/10.1089/cyber.2014.1502

68.

Yang

Pang

Liu

Yen

D. C.

Michael Tarn

(2015). Exploring consumer perceived risk and trust for online payments: An empirical study in China’s younger generation. Computers in Human Behavior, 50, 9–24. https://doi.org/10.1016/j.chb.2015.03.058

69.

Youn

(2009). Determinants of online privacy concern and its influence on privacy protection behaviors among young adolescents. Journal of Consumer Affairs, 43(3), 389–418.

70.

S. J.

Kuo

C. T.

Tseng

Y. C.

(2023). Association of financial literacy and risk preference with fraud exposure and victimization among middle-aged and older adults in China. Journal of Applied Gerontology, 42(1), 89–98.

71.

Zhang

Luximon

Song

(2019). The role of consumers’ perceived security, perceived control, interface design features, and conscientiousness in continuous use of mobile payment services. Sustainability, 11(23), 6843.

72.

Zhao

Pan

Y. H.

(2023). A study of the impact of cultural characteristics on consumers’ behavioral intention for mobile payments: A comparison between China and Korea. Sustainability, 15(8), 6956.

73.

Zhou

(2014). Understanding the determinants of mobile payment continuance usage. Industrial Management & Data Systems, 114(6), 936–948.