Relational event models (REMs) for the analysis of social interaction were first introduced 15 years ago. Since then, a number of important substantive and methodological contributions have produced their progressive refinement and hence facilitated their increased adoption in studies of social and other networks. Today REMs represent a well-established class of statistical models for relational processes. This special issue of Network Science demonstrates the standing and recognition that REMs have achieved within the network analysis and networks science communities. We wrote this brief introductory editorial essay with four main objectives in mind: (i) positioning relational event data and models in the larger context of contemporary network science and social network research; (ii) reviewing some of the most important recent developments; (iii) presenting the innovative studies collected in this special issue as evidence of the empirical value of REMs, and (iv) identifying open questions and future research directions.

The Shiji (史記 Records of the Grand Historian) is of great value for Chinese history before 90 BCE. Many online databases provide character-based search of the Shiji. We go beyond simple search by creating an word-based open-access database of the Basic Annals (本纪) of the Shiji that allows the exploration of relationships between persons and the relationships between persons and named places.

While it is increasingly recognized that shame is a pernicious component of the experience of poverty, the stigma generally associated with social assistance provision is less marked with respect to China's Minimum Living Security System, also known as dibao. This enigma is explored and illuminated drawing on two streams of indigenous Chinese scholarship and qualitative fieldwork in eight villages in Shanxi province. Economic and political changes prioritizing economic growth and individual wealth have increased the shame associated with poverty, manifest as loss of face, low mian (status) and lack of lian (integrity). However, this shame does not transfer to dibao because the scheme has been transformed locally into a universal age supplement that partially fulfils the demands of filial piety and which is seen to reflect and contribute to guanxi (social influence).

Coupled with the social practice of female hypergamy, the male surplus within the never-married population means that today's Chinese marriage market is extremely tight in particular for men from a rural background and the least privileged socio-economic categories. Drawing on quantitative data from a survey conducted in 2014–2015, this article sheds light on the situation of single men who are past prime marriage age in three rural districts of Shaanxi particularly affected by this phenomenon. It compares single men's characteristics to those of their married counterparts and offers insights into the heterogeneity of single men with the aim of challenging some commonly accepted assumptions about bachelorhood in rural China. Results suggest a strong internalization of the various characteristics, centred on being able to offer social mobility to a potential wife, that a man is expected to have to be attractive to women in a context where women have more choice in mate selection. We conclude that mate selection is highly marked by class, social norms, social interactions, health, generation and age, and requires the mobilization of certain amounts of individual, social and economic resources. Unwanted bachelorhood would thus be better understood using an intersectional approach rather than mainly in numeric terms.

This study examined which of the identified health, psychological, social and environmental mediators could most effectively explain the socio-economic status (SES)-based differences in participation in exercise among elderly Japanese. The candidates for mediators were composed based on the socio-ecological model. A representative sample of people 65 years and older living in two areas with different residential SES in Tokyo, Japan produced 739 effective participants. The intensity of exercise was evaluated based on whether the participants exercise for 30 minutes or longer twice a week, or for 20 minutes or longer three times a week. SES was evaluated by education and income. Mediators were assessed through four dimensions: (a) health, (b) psychological status, (c) social relations and (d) environmental context. As a result, SES's indirect effect through the mediators was evaluated using a multiple mediator model. The influence of both education and income on exercise was mediated by self-efficacy for exercise and social support for exercise. Self-efficacy for exercise had the strongest effect, while social support for exercise explained more of the income differences affecting participation in exercise than it did regarding educational differences. Self-efficacy for exercise may have the strongest effect as a mediator, which would explain the differences in participation in exercise among elderly Japanese based on education and income.

Loneliness is often associated with old age, but many studies have shown that the relationship is not straightforward. This paper seeks a better understanding of the impact of social isolation on feelings of loneliness among older people, by building on the theoretical and actual distinction between social and emotional loneliness. Social loneliness refers to a lack of feelings of social integration; emotional loneliness emerges in the absence of an attachment figure. This paper focuses on social loneliness and has two aims, first to disentangle the direct and intermediate effects of both the number and the quality of social relationships on social loneliness in old age, and second to detect the groups at risk of social loneliness by identifying which personal features correspond with which relational deficits and therefore indirectly increase the risk on social loneliness. Data are analysed for a sample of 1,414 respondents aged 55 or more years drawn from the Panel Study of Belgian Households conducted in 2000. The results confirm that improved understanding is gained by decomposing the interrelation between age and other background features, on the one hand, and the social relational features, on the other, as indirect and direct predictors of social loneliness. Generally, this approach promotes a correct identification of the groups at risk of social loneliness in old age.

Using social capital and social exchange theories, this investigation examined ethnic variation in patterns of social exchange in two heterogeneous racial groups, Blacks and Whites in the United States, and the effects of education and income on these patterns. The sample was 1,043 people aged 65–86 years from four ethnic groups (US-born European-Americans, immigrant Russians/Ukrainians, US-born African-Americans, and immigrant English-speaking Caribbeans) who had provided details of their instrumental and advice exchanges with kin and non-kin. Hierarchical multinomial logistic regressions were used to predict patterns of social exchange, variations by ethnicity, income and education, and the interactions. Ethnic differences in patterns of social exchange were found, but almost all were qualified by interactions. Those with income showed within-group heterogeneity: African-Americans and Russians/Ukrainians with higher income were more likely to engage in reciprocal instrumental kin exchange, whereas among English-speaking Caribbeans and European-Americans such exchanges were not associated with income. Unlike among European-Americans and English-speaking Caribbeans, Russians/Ukrainians with higher income and education were more likely to engage in reciprocal non-kin exchange. The findings suggest that ethnic variation in social exchange reflects both aspects of ethnic group membership and the relational context, as well as the enactment of reciprocity values in varying resource contexts.

Background: Research with nondepressed adults suggests that social networks and social support are stable over the life course until very late age. This may not hold true for older adults with depression. We examined baseline status and trajectories of social networks and social support at the group and individual levels over five years.

Methods: The sample consisted of 339 initially depressed adults aged 59 or older (M = 69 years) enrolled in a naturalistic study of depression. Measures of social ties, including social network size, frequency of interaction, instrumental support, and subjective support, were administered at baseline and yearly for five years.

Results: Latent growth curve models were estimated for each aspect of social ties. On average, social network size and frequency of interaction were low at baseline and remained stable over time, whereas subjective and instrumental support were high at baseline yet increased over time. There was significant variation in the direction and rate of change over time, which was not predicted by demographic or clinical factors.

Conclusions: Because increasing social networks may be ineffective and may not be possible for a portion of people who already receive maximal support, interventions to increase social support may only work for a portion of older depressed adults.

This paper reports a study of the associations between social relations, language and cognition among people aged 85 or more years (hereafter the ‘oldest old’). Although the links between cognition and both social relations and language ability are well-established, less is known about the relationship between social relations and language skills, especially among the oldest old. With a sample of 20 adults aged 85–93 years living independently in two retirement communities (a seniors' apartment complex and independent apartments in a continuum-of-care setting), we used the approach of the Rochester Interaction Record to assess the frequency, purpose and quality of their social interactions over one week. It was hypothesised that aspects of social relations, as well as the type of residential setting, would associate with both cognitive skills (measured using the Composite Cognistat) and language skills (measured using the Boston Naming Test). It was found that participants who had a low proportion of interactions with family members, or a high proportion with friends, as well as those with diverse relationships, performed better on the cognitive and language tasks. Furthermore, those from the setting with more programmed activities performed better on the language task. The conclusion critically examines the findings about the influence of family relationships and the importance of residential setting.