China's rapid economic growth has led to significant internal migration, with increasing numbers of elderly parents ('trailing parents') migrating to urban areas to live with their adult children. These older migrants face challenges in integrating into urban society due to factors such as the Hukou system (household registration system) restricting access to social benefits, age-related decline in technological proficiency and social skills, cultural differences, and strained intergenerational relationships. Social media has emerged as a crucial tool for older adults, but its impact on the social integration of trailing parents remains under-researched. This study addresses this gap by examining the relationship between social media engagement and social integration among trailing parents in China, exploring the mediating roles of perceived social support and psychological resilience. The researchers hypothesize that social media engagement will positively influence social integration, primarily through perceived social support, with a potential mediating role for psychological resilience. The study draws on Media System Dependency (MSD) theory and the Stress-buffering hypothesis to frame its investigation.

The study reviews literature on Media System Dependency (MSD) theory, which posits that individuals' ability to achieve goals depends on media information resources. The researchers adapt the theory to the context of social media use among older migrants, considering the three dependency relations: understanding, orientation, and play. The Stress-buffering hypothesis is introduced to examine the role of perceived social support as a protective factor against stress. Existing research highlights the positive effects of social media on social integration among migrants, but the underlying psychological mechanisms, particularly the role of resilience, remain under-explored. The literature review justifies the inclusion of perceived social support and psychological resilience as mediating variables, linking social media engagement to social integration.

This study employed a quantitative research design using a cross-sectional online survey. The survey was conducted in four major Chinese cities (Beijing, Xi'an, Nanjing, and Shenzhen) from June 20 to July 15, 2023, targeting elderly migrants aged 55 and above who had lived in the urban area for more than six months and used social media for more than two months. A non-probabilistic convenience sampling method was employed using the Tencent Questionnaire platform. After screening for eligibility, 1001 usable questionnaires were collected. The survey included sections on demographic information, social media use frequency, and measurement scales for social media dependency (understanding, orientation, play), intensity of social media use, social media engagement (information production, retrieval, social activities), perceived social support (significant others, family, friends), psychological resilience, and social integration. Partial Least Squares Structural Equation Modeling (PLS-SEM) was used to analyze the data. The researchers addressed common method bias through Harman's single-factor test and variance inflation factor (VIF) analysis. Higher-order constructs (HOCs) were used for social media engagement and perceived social support, employing a two-stage approach to assess their reliability and validity. Construct reliability and validity were assessed using composite reliability (CR), average variance extracted (AVE), and Heterotrait-Monotrait Ratio (HTMT). Bootstrapping with 5000 subsamples was used to test the hypotheses of direct and indirect relationships. The study examined direct and mediating effects to test the hypotheses.

The study found that understanding and orientation dependency relations significantly and positively affected social media engagement among trailing parents (H1, H2 supported). Play dependency had a smaller, but still significant, positive effect (H3 supported). Intensity of social media use did not significantly predict social integration (H4 supported, H5 not supported). Social media engagement significantly and positively predicted perceived social support (H6 supported) and social integration (H5 supported, contrary to H4). Perceived social support significantly and positively predicted social integration (H7 supported). Social media engagement significantly and positively predicted psychological resilience (H9 supported), and psychological resilience significantly and positively predicted social integration (H10 supported). The mediating effect of perceived social support on the relationship between social media engagement and social integration was significant (H8 supported, 47% variance accounted for). However, the mediating effect of psychological resilience was not sufficiently strong to be considered significant (H11 not supported, 17.4% variance accounted for). The sequential mediating effect of perceived social support and psychological resilience was significant and strong (H13 supported, 39% variance accounted for).

The findings suggest that active engagement with social media, driven by the needs for understanding and orientation, is beneficial for social integration among elderly migrants. The results support the Stress-buffering hypothesis, demonstrating that perceived social support, enhanced by social media use, plays a crucial role in facilitating social integration. Although psychological resilience is positively associated with social integration, its mediating role was less pronounced than that of perceived social support. This might be because the study participants engaged more in passive information consumption than active social interaction, limiting the potential of social media to enhance psychological resilience. The sequential mediation model highlights the importance of considering both internal and external resources in understanding the pathway from social media engagement to social integration. The study emphasizes the importance of quality over quantity in social media use for successful social integration.

This study provides valuable insights into the role of social media in the social integration of elderly migrants in China, highlighting the importance of perceived social support and the complex interplay between internal and external resources. The findings have implications for the design of more user-friendly social media apps for older adults, community initiatives to strengthen social support networks, and family involvement in assisting elderly migrants with social media use. Future research should employ longitudinal studies and probability sampling to strengthen the causal inferences and generalizability of the findings. Further investigation into different types of social media engagement and the various dimensions of social integration would also be beneficial.

The cross-sectional design limits the ability to establish causal relationships. The non-probabilistic convenience sampling may affect the generalizability of the findings. Treating social media engagement and social integration as broad constructs might have masked more nuanced relationships. The study might have overlooked additional drivers of social media engagement beyond the three dependency relations. Lastly, while the sequential mediation model is supported, exploration of additional mediators might further refine our understanding of the psychological processes involved.