Evidence brief

Does it matter who we socialize with?

March 3, 2023
Kiffer Card, Cindy Yu, Adam Frost, Jocelle Refol, Pete Bombaci

Background

Social connection is a key factor shaping human health and happiness (Kawachi & Berkman, 2001; O’Callaghan et al., 2021; Holt-Lunstad et al., 2015). As such, individuals need to prioritize their social lives in order to achieve an optimal state of wellbeing (Holt-Lunstad et al., 2017). However, given that social connection requires considerable investments of time and energy (leikas & Ilmarinen, 2016), it may be helpful for people to understand what types of social interaction offer the greatest return on their investments. For instance, if different types of relationships contribute differently to health and wellbeing, individuals can prioritize relationships that meet their own individual needs and circumstances (Hudson et al., 2020). For instance, one might choose to invest more energy talking with strangers and building a large network of acquaintances. Alternatively, they might instead prioritize develop deeply meaningful friendships. In any case, knowing where to invest one’s social energy is important.

Purpose

The purpose of this brief is to answer the question, “Does it matter who we socialize with?” In exploring this question, we recognize that individuals have many different kinds of relationships and most people fulfill multiple roles depending on who they are interacting with. In other words, we are simultaneously – to eachother – friends, family members, colleagues, neighbours, acquaintances, and strangers. Understanding the various contributions that these roles and relationships play in shaping our individual and collective wellbeing therefore provides important information for individuals to act on as we pursue our social lives.

Evidence from Existing Studies

The literature exploring social needs and the relationships that fulfill these needs dates back decades. Seminal work by Weiss (1974) identifies six fundamental social needs grouped into (1) assistance needs ([a] guidance/information, [b] reliable alliance/tangible assistance) and (2) non-assistance needs ([c] nurturance/opportunity to care for others, [d] reassurance of worth/recognition, [e] social integration/common interests, [f] attachment/closeness). Researchers have found that people rely on different types of relationships to fulfill these various needs (Lempers & Lempers, 1992) and in his later work Weiss (1998) identifies attachments (e.g., emotionally close relationships) and affiliations (e.g., goal-directed relationships) as two broad archetypes for the relationships that help us meet these needs. When needs from attachments and affiliations are not met, Weiss (1973) acknowledges that we experience emotional loneliness (deficiencies in intimate attachments) and/or social loneliness (deficiencies in social affiliations). Of course, many other taxonomies for pit relationships exist: The literature identifies strong and weak ties (Granovetter, 2022), kin, kith, and strangers (Douglas et al., 2008), and various gradations of friendship (Litwak, 1985; Hall, 2018; Hill & Dunbar, 2003). These frameworks share a common understanding of relationships: that they are diverse and that this diversity helps fulfill different needs we may have. Based on this literature, it is well reasoned that a diverse social network of differing types of relationships is needed to fulfill our social needs (Wellman & Worley, 1990; Fiori et al, 2006). Empirically, this is what we observe when examining the composition of people’s social networks (Dunbar & Spoors, 1995). For example, Collins et al. (2022) reported that relational diversity (i.e., interacting with a mix of family members, friends, acquaintances, strangers, and others) is robustly associated with wellbeing above and beyond the time that individuals spend socializing. Other studies, including Miritello et al., (2013), Tamarit et al. (2022), and Takano & Fukuda (2017), have conducted research examining emotional closeness with network members and demonstrated that the limitations of time and resources encourage humans to focus more on our closest social relationships but that, nevertheless, people tend to optimize their social lives such that they interact with individuals across the spectrum of emotional closeness. Sutcliffe et al., (2012) shows that we spend approximately 40% of our time with the most inner circle of approximately 5 friends, 20% with our next closes 10 or so friends, and the remaining time is spent across the 135 or so people in our social networks.

However, not all researchers agree that many different types of relationships are needed. Notably, Cantor (1979) suggests that individuals really only need formal, primary attachments that meet all of their needs and that secondary, informal relationships are only compensatory stop-gaps that serve to pick up where there is slack. Under this hierarchical-compensatory model, anybody can step in to fulfill the needs not being met by one’s formal attachments. For example, Fuller-Iglesias et al. (2013) highlights the nuanced ways in which quality friendships can mitigate harms from having low quality relationships with family. Kaufman et al. (2022) also finds that the presence of high quality friendships can mitigate harmful effects from a low quality primary relationship. However, Kaufman’s study also finds that people with happy primary relationships are satisfied with life regardless of their friendship quality. These studies suggest that our primary bonds may be especially important in meeting our social and emotional needs.

Case Study: Relationship between Burnout and Sources of Social Support
The Multidimensional Scale of Perceived Social Support, developed by Zimet et al. (1988), underscores the diverse support functions that are fulfilled by family, friends, and significant others. Card et al. (2022) examines how various social factors contribute as protective factors against burnout. This study finds that in multivariable modelling, social support from family (but not from friends or significant others) independently predicts lower Burnout scores. In fact, dominance analyses reveal that support from family was among the strongest protective factors in the multivariable model. Similar findings have also emerged from other studies demonstrating the uniquely important role that family support plays in protecting against poor mental health (Koutsimani et al., 2021; Possel et al., 2018).

While the debate over how diverse a social network should be may not yet be settled empirically, Hudson et al. (2020) recently explored the benefits of spending time with various kind of people. They reported that in terms of situational benefits, friends provided much more fun that families. However, in controlling for activity characteristics they found that it was not so much who we spent time with, but rather how that time was spent. In other words, it was the particular activities engaged in during social interactions that were most important in explaining differences in how satisfying or beneficial a given social interaction was. However, in looking beyond a specific situation, they found that greater time with an intimate partner was the only factor that contributed to global wellbeing. Taken together, these results suggest that while overall wellbeing may be more contingent on our close relationships, it is also important to have friendships that bring satisfaction and enjoyment in given situations. Sharifian et al. (2020), further supports this assertion by showing that contact with friends, but not family, is associated with greater engagement in cognitive and stimulating activities. This suggests that even if individuals can rely on a primary relationship to meet most of their needs, they do not tend to structure their relationships in this way. We engage in different activities with different people and rely on friends, family, and others to fulfill our various social needs.

In addition to considering the nature of the relationships between individuals in a social network, it is also important to recognize that the particular people and personalities that make up a given network are also relevant. For instance, Fowler & Christakis (2008) showed that happiness is a seemingly infectious attribute. Using data from the Framingham Heart Study they found that happy people tend to cluster together and waves of happiness seem to spread like wildfire. Given these dynamics, it may be more important who you choose to socialize with than whether you describe that person as a friend, neighbour, family member, or coworker. Indeed, the idea of a “chosen family” underscores the reality that some friends may play an even more important role than our biological kin (Fiori et al., 2006; Kim & Feyissa, 2021).

Analyses from the Canadian Social Connection Survey

Using data from the 2022 Canadian Social Connection Survey, we examined how much time participants wanted to spend with various relations as an indicator of their social priorities. After all, it is reasonable to assume that people’s preferences reflect, at least in part, their lived experiences and therefore their social needs. Results from these analyses indicate that, on average, participants want to spend more time socializing with family (Mean = 11.45 hours per week, SD = [15.38]) than with friends (Mean = 7.80 hours per week, SD = [9.42]).

Additionally, we also examined the effect of spending social time with strong ties (i.e., family, friends) versus weak ties (i.e., coworkers and classmates, neighbours, acquaintances, strangers) over the past seven days on participant’s emotional and social loneliness. Results showed that social loneliness was lowest when participants spent 30 - 90% of their social time with strong ties while emotional loneliness was lowest when participants spent 85 – 90% of their social time with strong ties.

To provide greater specificity in our estimates, we conducted cut point analyses, which indicated that participants needed to spend at least 31% (95% CI = 8-53%) of their social time with strong ties to avoid social loneliness and at least 67% (95% CI = 54-86%) of their social time with strong ties to avoid emotional loneliness. This suggests we should spend probably at least two-thirds of our time with close ties. However, our data does not support the idea that 100% of our social time should be spent with only strong ties. Some time with weak ties seems to be beneficial, especially for social loneliness. In fact, cut point analyses indicate that participants should spend at least 33% (95% CI = 27%-46%) of their social time with weak ties to avoid social loneliness and at least 4% (95% CI = 1%-7%) to avoid emotional loneliness. After all, weak ties can provide us with many, low-risk opportunities for social engagement. As well, our weak ties provide a pool from which we can build future strong ties.

In ascertaining the amount of time that should be spent with various strong ties, we recognize that individual family situations likely play an important moderating role. Nevertheless, we used multivariable regression models (controlling for age, gender, ethnicity, sexual orientation, disability status, and income) to explore the relative contribution of time spent with friends and family to emotional and social loneliness. Standardizing the effects allowed us to examine the relative contribution of time with family and friends on each outcome. These analyses revealed that both time with family (sβ = -0.075, SE = 0.030, p = 0.012) and friends (sβ = -0.065, SE = 0.030, p = 0.029) were associated with lower emotional loneliness (even controlling for the effects of each). However, time with friends (sβ = -0.168, SE = 0.033, p < 0.001), but not time with family (sβ = -0.042, SE = 0.032, p = 0.191), was associated with lower social loneliness. In the model for social loneliness, the effect of time with friends was nearly four times larger than the effect of time with family. The differences reflected here highlight the relatively strong contribution that time spent with strong ties has on emotional versus social loneliness – supporting the likelihood that different types of relationships fulfill different social needs.

Discussion

The weight of existing evidence suggests that we need diverse types of relationships – including those with friends, families, and so called “weak ties” (e.g., coworkers, acquaintances, Neighbours, and strangers). The need for these diverse relationships arises from the multifaceted social needs that humans experience. When these needs are not met, we experience different types of loneliness (Cacioppo & Cacioppo, 2012; DiTommaso & Spinner, 1997). While it remains unclear whether or not our needs can be met solely by our primary relationships – as hypothesized by Cantor (1979) – it is evident that this is not how people typically operate to meet their needs. It is difficult to say exactly what proportion of our time should be spent with family versus friends versus others, however it is certain that a considerable amount of time with family and friends – likely upwards of two thirds of your social time – is needed to avert emotional loneliness. Social loneliness, on the other hand, requires individuals to have connections beyond family and friends. As such, based on the current evidence it is clear that individuals must be aware of their social needs, where they experience deficiencies, and thereafter prioritize the relationships that are likely to meet these needs.

Conclusion

Based on the available evidence and our analyses of the Canadian Social Connection Survey, we recommend policies and programs that support individuals in creating diverse social networks that emphasize strong emotional attachments with friends and family, but also include meaningful affiliations with neighbours, friends, acquaintances, and even strangers.