Retirement migrants are individuals who permanently or temporarily relocate in pursuit of a “good life” (9, 10). Typically, these migrants are over the age of 50 or 60, depending on the retirement standards of each country, and have exited the labor market or begun receiving pensions (2, 11). They occupy a space between tourists and permanent residents, aiming for a happy life while placing significant emphasis on the tourism experience (2). Their duration of stay at the destination is flexible, reflecting diverse mobility patterns that range from a few months each year to several years (9). As a distinctive form of tourism mobility within a unique demographic, retirement migrants have garnered significant attention from tourism scholars (12, 13).

The phenomenon of retirement migrants first emerged in North America during the 1950s (14). Since then, retirement migration has evolved into a global spatiotemporal phenomenon (15, 16). Unlike retirees from Europe and North America who often choose to retire abroad (9, 16), Chinese retirement migrants primarily relocate to domestic destination cities such as Zhuhai (2) and Sanya (3, 126). With the acceleration of population aging in China, the number of retirement migrants is rapidly increasing, leading to various social issues in tourist destinations. Enhancing the social integration of retirement migrants can effectively mitigate social conflicts (2).

Social integration denotes the extent of an individual’s connections within broad social relationships and their active participation in social life (17). Social integration is not a static or unidimensional concept; rather, it is dynamic, progressive, multidimensional, and interactive (18). Scholars have proposed various measurement indicators from both macro and micro levels, encompassing subjective and objective perspectives (19–21). For instance, Yang (18) categorizes social integration into four dimensions: economic integration, cultural acceptance, behavioral adaptation, and identity recognition. The integration process for lifestyle migrants encompasses three primary dimensions: emotional attachment, cultural adaptation, and social participation (22).

The social integration of retirement migrants in new living environments represents a relatively new area of research (23). Studies specifically addressing the social integration of retirement migrants generally emphasize interpersonal interactions, lifestyle habits, cultural acceptance, and psychological distance (2). Over time, as these migrants engage in more extensive and meaningful social interactions, they experience a continuously deepening and dynamically evolving process of social integration (24).

This study defines the social integration of retirement migrants as the process of adapting to local life and mutual adaptation among retirement migrants, including three dimensions: lifestyle habits, interpersonal interactions, and activity participation. The factors influencing the social integration of older migrants can be categorized into four levels: individual factors, family factors, social support factors, and policy assurance factors (25). In the digital age, the advent of digital technologies, exemplified by smartphones and social media, has profoundly transformed modes of communication and interaction (57). Consequently, the relationship between digital literacy and the social integration of older adults warrants further investigation.

Digital literacy is defined as the capability to access and utilize information via digital devices (26, 27). This concept extends beyond basic technical skills to include the ability to evaluate, utilize, and produce digital content (28). Initially, literacy pertained to familiarity with computers, which evolved into what is known as computer literacy (29–31). Subsequently, with the proliferation of the internet, information and communication technologies, online platforms, and digital media, the term “computer literacy” further transformed into “digital literacy” (29). Associated terms such as ICT literacy, network literacy, information literacy, and electronic literacy have been discussed in the scholarly literature as interchangeable with digital literacy (32). The concept of digital literacy is increasingly prevalent in the 21st century (33). For instance, Reddy et al. (34) introduced the digital literacy framework for the South Pacific region.

In the contemporary era of mobile internet, the use of smartphones and social media has become emblematic of digital engagement (57). Smartphones, characterized by their portability, immediacy, and convenience, have become an indispensable element of contemporary daily life (35, 36). They fulfill a multitude of needs including information access, instant communication, social interaction, and entertainment (37, 38). Social media platforms offer tourists the opportunity to receive and share travel experiences, as well as disseminate information about destinations and amenities (39). Additionally, various hospitality and tourism enterprises utilize social media to enhance their operations and customer engagement (40).

Empowerment refers to the initiatives undertaken by individuals or groups to control their destiny by enhancing their capabilities (i.e., internal power) and influencing their environment (i.e., external response) (41, 42). Empowerment transforms individuals into proactive agents who act in alignment with their values and interests (43). In the digital age, devoid of the constraints of time and place, anyone with internet access can more readily access vast amounts of information and disseminate information to millions at unprecedented speeds (44, 45). Consequently, consumer power has evolved from being based on individual needs and information to being rooted in networks and crowds (46).

Empowerment research within the tourism sector can be broadly categorized into three primary themes: the empowerment of destination residents, the empowerment of tourism practitioners, and the empowerment of tourists (47). The empowerment of destination residents is deemed crucial for achieving sustainable tourism development (42, 48, 49). Tourism practitioners can leverage digital technologies to analyze tourist preferences via the internet, respond promptly to customer needs, and protect consumer rights (50). Tourists, on the other hand, can not only access travel information but also generate travel content and reviews, thereby enhancing their own capabilities (47).

Smartphone addiction is defined as the excessive use of smartphones despite their negative impacts, taking precedence over other daily activities (51). Earlier research predominantly focused on mobile phone addiction and internet addiction (52, 53). With the shift from mobile phones to smartphones, the term “smartphone addiction” has become more prevalent (54). Smartphone addiction is also referred to as smartphone dependency, overuse, or problematic smartphone use (8).

Characterized by an uncontrollable pattern of excessive smartphone use, smartphone addiction can lead directly or indirectly to a range of issues, such as mental health problems and interpersonal difficulties (55). Due to its convenience and easy internet connectivity, smartphone addiction can be more hazardous than other forms of addiction, making it a significant societal issue (51, 56).

Previous studies on smartphone addiction have primarily focused on children, adolescents, and college students, with relatively few studies on adults, especially older adults (51, 57). The reduction in social roles and weakening of offline relationships among older adults may increase their smartphone usage, potentially leading to smartphone addiction (51).

Tourist empowerment markedly influences the lifestyle habits of retirement migrants by fostering autonomy and enhancing their ability to manage their lives and make informed decisions (94, 95). This empowerment leads to improved self-reliance and health consciousness, enabling retirement migrants to adopt healthier routines and make better lifestyle choices, ultimately benefiting their overall wellbeing (129).

In the realm of interpersonal interactions, tourist empowerment significantly bolsters the social and emotional support networks of retirement migrants. By improving their ability to communicate and maintain connections with friends, family, or like-minded individuals, these migrants experience enriched social interactions (96, 97).

Finally, tourist empowerment positively influences the participation of retirement migrants in various activities. Increased confidence and reduced anxiety about engaging in new activities lead to higher levels of participation, reflecting positive interpersonal interactions and serving as a direct indicator of social integration (98). Enhanced activity engagement strengthens social bonds and contributes significantly to the migrants’ sense of community and overall social integration.

Therefore, we propose the following hypotheses:

More specifically:

Smartphone addiction affects the lifestyle habits of retirement migrants. Smartphone addiction adversely affects both physical and mental health (99). For example, People frequently use smartphones while eating, during breaks, and even place them next to their pillows while sleeping, which can lead to problematic lifestyle habits (7, 100).

Smartphone addiction affects the interpersonal interactions activity participation of retirement migrants. The internet facilitates individuals’ absorption in the virtual world at the expense of engaging with the physical world (96). The lack of meaningful virtual interactions can hinder the establishment of genuine, deep relationships, thus impeding social integration (52). For instance, excessive reliance on smartphone interactions instead of engaging directly with family and friends detracts from more direct and authentic physical and social experiences, leading to isolation from the real world (101, 102). Furthermore, smartphone addiction can result in reduced participation in offline activities and the degradation of social skills (103).

In summary, we propose the following hypotheses:

More specifically:

This study focused on two distinct regions in China: Hangzhou in Zhejiang Province, and Hainan Province (specifically Chengmai County and Ledong Li Autonomous County). The selection of these locations ensures sample representativeness and diversity, capturing both urban and resort-like environments that reflect different retirement migration settings. Hangzhou, famously described as part of “heaven on earth” alongside Suzhou, is not only an internationally renowned tourist destination but also a popular choice for retirement migrants. Hainan, with its significant retiree population, has become a focal point for retirement migration research, attracting considerable academic attention (12).

Participants were recruited from retirement communities and senior living facilities in both locations, focusing on areas with high concentrations of retirement migrants. The study employed a convenience sampling method to recruit individuals aged 50 and above who were in good health. The age threshold of 50 was chosen based on two key considerations: age provides a straightforward and objective criterion, making it easier to define compared to more subjective standards; and individuals aged 50 and above are more likely to exhibit common signs of aging, such as gray hair, physical weakness, wrinkles, and declining functional abilities, making them more representative of the retirement migrant population (104).

While convenience sampling was used due to resource constraints and accessibility issues, steps were taken to minimize potential bias. These included recruiting from various retirement communities and facilities to ensure diversity in socioeconomic status and lifestyle preferences, and employing quota sampling to maintain a balanced representation of age groups and genders. These measures aimed to enhance the sample’s representativeness despite the limitations of the convenience sampling method.

The measurement scales utilized in this study were primarily adapted from established scales in previous research. As the original scales were developed in English, a rigorous translation and back-translation procedure was implemented to ensure cultural and linguistic equivalence. Two bilingual experts independently translated the scales into Chinese, with any discrepancies resolved through collaborative discussion. Subsequently, a third expert back translated the Chinese version to English to verify accuracy and ensure conceptual equivalence across languages. Two bilingual experts independently translated the scales, with discrepancies resolved through collaborative discussion. A third expert then back translated the Chinese version to English to verify accuracy and ensure conceptual equivalence across languages.

The final questionnaire comprises two main sections. The first section gathers demographic information, including gender, age, education level, and income. The second section consists of a 21-item measurement scale assessing four key constructs: digital literacy, tourist empowerment, smartphone addiction, and social integration. These constructs are measured using a 5-point Likert scale, with responses ranging from “1” (strongly disagree) to “5” (strongly agree) for digital literacy, tourist empowerment, and social integration. Smartphone addiction is assessed on a frequency scale from “1” (never) to “5” (always). All scales have been linguistically adapted to suit the research context.

Independent variable. Digital literacy: recognizing the limitations of previous instruments that often measured only one or two components of the DigComp framework (105), this study employs a four-item scale specifically designed to assess digital literacy among older adults (106, 107). A sample item is: “I can effectively find the information I want on the internet through my smartphone.”

Mediating Variables. (a) Tourist empowerment. Measured using three items adapted from the Tourist Consumer Empowerment Scale (108). An example item is: “I can use various means to obtain a better experience in retirement tourism.” (b) Smartphone addiction. Assessed with four items adapted from the Mobile Phone Dependency Scale (52, 109). A representative item is: “People around me have complained about my smartphone use.”

Dependent variable. Social integration: drawing on established Social Integration Scales (2, 18), this construct is measured across three dimensions - “lifestyle habits, ““interpersonal interactions, “and “activity participation” - using a total of 10 items.

Data collection for this study was conducted from January 15, 2024, to January 30, 2024. To ensure consistency and accuracy throughout the process, two research assistants were assigned to manage data collection at each survey location. Prior to initiating the survey, all participants were provided with a detailed explanation of the study’s purpose, potential risks and benefits, and their rights as research subjects. The questionnaires were administered face-to-face, allowing participants to ask questions and receive clarification from the research assistants throughout the completion process. To protect participants’ privacy, all responses were collected anonymously, with no personally identifiable information gathered. A reminder mechanism was also employed during the data collection period to improve response rates.

A total of 500 questionnaires were distributed, of which 425 were returned. After data cleaning—removing responses with low completion rates, internal contradictions, or missing key information—369 valid responses were included in the final analysis, resulting in an effective response rate of 73.8%. All data were securely stored, with access restricted to the research team to maintain confidentiality and data integrity.

This study primarily employs two statistical analysis software packages: IBM SPSS Statistics 23.0 and AMOS 23.0, to analyze the questionnaire survey data. Firstly, descriptive analysis of the questionnaire data is conducted using SPSS, and the common method bias of the data is tested. Secondly, structural equation modeling (SEM) is performed using AMOS software to test the research hypotheses through path analysis. Lastly, the bootstrap method is utilized to examine the mediating effects, specifically verifying the mediating roles of tourist empowerment and smartphone addiction.

A total of 369 respondents participated in this survey. The demographic characteristics of the respondents are shown in Table 1. There were 224 females and 144 males, with more female respondents than male respondents. Regarding the age distribution of the respondents, the survey mainly targeted older adults population aged 50 and above, with a significant concentration in two age groups: 61–70 years old and 71–80 years old, accounting for 32.5 and 31.4%, respectively. In terms of educational attainment, the majority of respondents held a bachelor’s degree, with 169 individuals, accounting for 45.8%, followed by those with a high school or vocational diploma, totaling 126 individuals, accounting for 34.1%. The monthly income of the respondents was primarily concentrated below 7,000 CNY, with 149 individuals earning <5,000 Yuan, accounting for 40.4%, and 128 individuals earning between 5,000 and 7,000 Yuan, accounting for 34.7%.

Data normality was assessed by examining skewness and kurtosis values, as indicated in Table 2. The skewness values ranged from −1.304 to 0.635, while the kurtosis values ranged from −1.258 to 2.459. Following the thresholds suggested by Kline (110) of skewness <3 and kurtosis <8, it was determined that the data exhibited normal distribution.

Since all items in this study were completed by retirement migrants, there is a possibility of a common method bias issue. To examine the presence of common method bias in this study, the Harman single-factor test method was employed to assess the data homogeneity. The first common factor accounts for 32.36% of the total load when not rotated, which falls below the standard threshold of 40% (130). This suggests that there is no significant common method bias in this study, allowing for further data analysis to proceed.

A confirmatory factor analysis (CFA) employing MLM estimation was conducted using AMOS 23.0 to evaluate the construct validity and reliability of the proposed model. Table 3 outlines the variables and their corresponding items. The Cronbach’s alpha (α) coefficients for the six latent variables all surpass the recommended threshold of 0.7suggested by Churchill (111), except for Lifestyle habits. Additionally, the composite reliability (CR) scores for each latent variable exceed the suggested level of 0.70, affirming the internal consistency reliability of the measurement items. Convergent validity of the measures is supported by average variance extracted (AVE) values surpassing the recommended threshold of 0.50 for each construct, except for Lifestyle habits. Lastly, the measurement model exhibits a favorable fit to the data: χ²/df = 1.861, GFI = 0.929, CFI = 0.970, NFI = 0.937, TLI = 0.961, RMSEA = 0.048. These metrics collectively indicate that the model is well-suited to the observed data (112).

Discriminant validity was evaluated by contrasting the intercorrelations of the variables with the square root of the average variance, as proposed by Petrick (113). Table 4 displays the findings. The shared variance between variable pairs was found to be lower than the corresponding Average Variance Extracted (AVE) values, except for Lifestyle habits, consistent with the criteria outlined by Fornell and Larcker (114), indicating satisfactory discriminant validity.

AMOS 23.0 was utilized to conduct Structural Equation Modeling (SEM) to estimate the main effects model, with its parameters utilized to test all the hypotheses. Additionally, the main effects hypotheses were assessed through SEM. The results of SEM demonstrated a satisfactory overall fit: χ²/df = 1.917, TLI = 0.959; CFI = 0.968; NFI = 0.935; GFI = 0.927; IFI = 0.968; RMSEA = 0.050.

As demonstrated in Table 5, the SEM results also indicated that elders’ digital literacy has a significant positive effect on their social integration, Lifestyle habits (β = 0.198, p < 0.05), Social interaction (β = 0.194, p < 0.05), and activity participation (β = 0.269, p < 0.05), thus, H1a, H1b and H1c were supported. At the same time, the SEM results indicated that elders’ digital literacy has a significant positive effect on Smartphone addiction (β = 0.394, p < 0.05) and Tourist empowerment (β = 0.245, p < 0.05), H2 and H3 were supported. Additionally, Tourist empowerment has a significant positive effect on their Lifestyle habits (β = 0.533, p < 0.05), Social interaction (β = 0.364, p < 0.05) and activity participation (β = 0.558, p < 0.05), thus, H4a, H4b and H4c were accepted. Smartphone addiction has a significant negative effect on Lifestyle habits (β = −0.183, p < 0.05), and has no significant effect on social interaction (β = −0.081, p > 0.05), and activity participation (β = −0.034, p > 0.05), thus, H5a was accepted, but H5b, H5c were rejected.

We further investigated the mediating effects of Tourist Empowerment (TE) and Smartphone addiction (SA) using the methodology advocated by Zhao et al. (115). Results are presented in Tables 6. In the bootstrapping analysis, a path is deemed significant and corroborated if the bootstrap confidence interval excludes zero. First of all, tourist empowerment (β = 0.065, p < 0.05, CI = [0.031, 0.114]) and smartphone addiction (β = −0.025, p < 0.05, CI = [−0.058, −0.004]) play a significant role in the relationship between digital literacy and Lifestyle habits, given the direct effect is significant, thus, tourist empowerment and Smartphone addiction are partial mediator. Secondly, tourist empowerment (β = 0.073, p < 0.05, CI = [0.031, 0.128]) play a significant role in the relationship between digital literacy and social interaction given the direct effect is significant, thus, tourist empowerment plays a partial mediator role in the relationship between digital literacy and social interaction. However, the mediating effect of Smartphone addiction is not significant. Finally, tourist empowerment (β = 0.102, p < 0.05, CI = [0.045, 0.171]) play a significant role in the relationship between digital literacy and activity participation, given the direct effect is significant, thus, tourist empowerment plays a partial mediator role in the relationship between digital literacy and activity participation. However, the mediating effect of Smartphone addiction is not significant.

This research seeks to explore the dual effects and mediating pathways through which digital literacy influences social integration among retirement migrants. Based on the results and discussion, the research conclusion can be summarized as follows: digital literacy demonstrates a double-edged sword effect on the social integration of retirement migrants. Specifically, digital literacy facilitates social integration through tourist empowerment while impeding it through smartphone addiction. These findings underscore the paradoxical “empowerment-alienation” dynamic inherent in technology-enabled aging processes.

The investigation into the relationship between digital literacy and social integration among retirement migrants is a relatively novel research domain, and this study presents several limitations that highlight areas for future inquiry.

Firstly, the sample exclusively comprised Chinese retirement migrants, which may restrict the generalizability of the findings to other cultural contexts. Future research should consider extending the sample to include individuals from diverse cultural backgrounds to assess the cross-cultural applicability of the results.

Secondly, the utilization of cross-sectional data in this study constrains the capacity to infer causal relationships between variables. Implementing longitudinal data or experimental designs, in future research would furnish more compelling evidence to elucidate the causal mechanisms between variables and offer a deeper understanding of the evolution of digital literacy and social integration over time.

Thirdly, other mediating factors may influence the relationship between digital literacy and social integration. Future research should explore alternative transmission mechanisms from various theoretical perspectives. Additionally, examining the role of moderating variables could provide a more nuanced understanding of this relationship.