The study, conducted by Tuo Zhang, Peng Li, Akmar Hayati Ahmad Ghazali, Nur Atirah Kamaruzaman, and Shougui Pang, set out to answer a question that has puzzled sociologists and communication scholars for years: through what relational pathways do modern forms of connection — digital media use and cross-border ties — become associated with generalized trust? Rather than assuming that more connectivity equals more trust, the researchers distinguished between two distinct relational routes. The first is an expansion-oriented pathway, captured by occupational network heterogeneity, meaning the diversity of occupational positions a person can access through their contacts, from university professors and lawyers to farmers and hairdressers. The second is an embeddedness-oriented pathway, captured by neighborhood social embeddedness, which reflects repeated face-to-face interaction, familiarity, and mutual support within identifiable local relationships.

To test these ideas, the team analyzed responses from 5,307 participants in the 2023 Chinese General Social Survey household module, a nationally representative sample. Generalized social trust was measured with a single survey item asking whether most people can generally be trusted, rated on a four-point scale. Digital media use intensity was constructed from three indicators — frequency of internet use, whether the internet was the primary information source, and how often the internet was used during leisure time — combined into a single composite score using principal component analysis. Cross-border relational ties were identified by whether respondents had relatives, friends, or acquaintances living abroad. Occupational network heterogeneity was counted as the number of occupational categories, out of ten, in which respondents personally knew someone. Neighborhood social embeddedness combined how often people interacted with neighbors, how many neighbors they greeted, and how many neighbors they could turn to for help.

The results were striking in their asymmetry. Digital media use intensity showed a consistent negative association with generalized social trust across multiple statistical specifications — baseline ordinary least squares regression, ordered logit and probit models, multinomial logistic regression, survey-weighted estimates, and multiple imputation analyses for missing data. In other words, the more intensely people engaged with digital media, the less they tended to trust strangers, and this pattern held up under nearly every robustness check the researchers ran. By contrast, cross-border relational ties — having connections abroad — showed no significant direct association with generalized trust at all. Having overseas contacts, it seems, neither builds nor erodes trust in unfamiliar others in any measurable way.

Both digital media use intensity and cross-border ties were strongly and positively associated with occupational network heterogeneity. People who used digital media more intensively reported knowing individuals across a wider range of occupational positions, and those with cross-border connections reported even broader occupational networks — cross-border ties increased the expected occupational network range by approximately 35 percent, according to negative binomial count models. But here is the critical twist: occupational network heterogeneity itself showed no significant association with generalized social trust. Knowing a professor, a nurse, a police officer, and a farmer does not, by itself, make someone more likely to believe that most people can be trusted. The expansion of relational boundaries, in this data, opened doors to diverse contacts without generating the deeper experiential foundations that trust in strangers seems to require.

Neighborhood social embeddedness told a very different story. It was positively and consistently associated with generalized social trust across every model specification the researchers tested — baseline models, alternative ordinal specifications, weighted estimates, multiple imputation, and both rural and urban subsamples. The association persisted even after the researchers removed the single item most directly referring to receiving help from neighbors, a targeted sensitivity analysis designed to rule out conceptual overlap between the two constructs. This suggests that the repeated, familiar, mutually supportive interactions of neighborhood life — not merely the availability of help — provide an experiential foundation from which people generalize expectations of trustworthiness to broader social others.

Perhaps the most theoretically intriguing finding concerns the indirect pathways. Digital media use intensity was positively associated with neighborhood social embeddedness after adjusting for demographic and socioeconomic covariates, and this produced a small but statistically significant positive indirect association linking digital media use to generalized trust through neighborhood ties. However, the researchers were careful to flag an important caveat: this positive adjusted association actually reverses the negative zero-order correlation between digital media use and neighborhood embeddedness observed in the raw data. Supplementary diagnostic analyses showed that the sign reversal emerged only after accounting for age and education, a pattern consistent with statistical suppression rather than a straightforward positive relationship. The indirect pathway, the authors emphasize, should be understood as a small, covariate-adjusted, model-dependent channel — not as evidence that heavier digital media use generally strengthens neighborhood life.

The coexistence of a negative direct association between digital media use and generalized trust alongside a small positive indirect association through neighborhood embeddedness creates what statisticians call an inconsistent indirect association structure. Digital media engagement, in this framework, is neither uniformly trust-eroding nor trust-building; it participates in multiple relational processes that pull in different directions. The authors suggest that intensive digital media use may coexist with informational fragmentation, uneven information credibility, and weakly contextualized social encounters — conditions that do not necessarily foster confidence in unfamiliar others. They caution, however, that their data do not directly measure media credibility, information disorder, or platform governance, so these interpretations remain theoretically informed rather than empirically tested mechanisms.

The urban–rural comparison added further nuance. The negative association between digital media use intensity and generalized trust was significant among rural residents but statistically insignificant among urban residents, with a significant interaction term between digital media use and urban residence. Yet the positive association between neighborhood social embeddedness and generalized trust remained stable across both rural and urban groups, reinforcing the argument that locally embedded relational experiences provide a more consistent foundation for generalized trust than relational expansion, regardless of residential context.

The broader implication is a challenge to connectivity-centered thinking about social capital. Expanding relational boundaries — whether through digital platforms or transnational ties — creates opportunities for diverse contacts but does not by itself produce the repeated interaction, reciprocity, and social predictability that appear to underpin generalized trust. Instead, the study points toward the enduring value of locally embedded relationships: the neighbors we greet, the familiar faces on the street, the people we know we can knock on when something goes wrong. In an era when digital transformation and social mobility continue to reshape interpersonal life in China and beyond, the findings suggest that trust in strangers is not a simple byproduct of being connected to more people. It grows, if it grows at all, from the accumulated experience of reliable, reciprocal, face-to-face relationships — a foundation that no amount of online connectivity can fully replace.

Subject of Research: The relationship between digital media use, cross-border relational ties, and generalized social trust in contemporary China

Article Title: Digital media use intensity, cross-border relational ties and generalized social trust: occupational network heterogeneity and neighborhood social embeddedness in contemporary China

Article References: Zhang, T., Li, P., Ghazali, A. H. A., Kamaruzaman, N. A., & Pang, S. (2026). Digital media use intensity, cross-border relational ties and generalized social trust: occupational network heterogeneity and neighborhood social embeddedness in contemporary China. Frontiers in Sociology, 11, Article 1911185. https://doi.org/10.3389/fsoc.2026.1911185

Image Credits: AI Generated

DOI: 10.3389/fsoc.2026.1911185

Keywords: generalized social trust, digital media use, cross-border relational ties, neighborhood social embeddedness, occupational network heterogeneity, social capital, China General Social Survey, trust in strangers, relational pathways, urban-rural differences, Digital, media