Use of Social Network Sites and Instant Messaging Does Not Lead to Increased Offline Social Network Size,or to Emotionally Closer Relationships with Offline Network Members

Abstract

The effect of Internet use on social relationships is still a matter of intense debate. This study examined the relationships between use of social media (instant messaging and social network sites), network size, and emotional closeness in a sample of 117 individuals aged 18 to 63 years old. Time spent using social media was associated with a larger number of online social network “friends.” However, time spent using social media was not associated with larger offline networks, or feeling emotionally closer to offline network members. Further, those that used social media, as compared to non-users of social media, did not have larger offline networks, and were not emotionally closer to offline network members. These results highlight the importance of considering potential time and cognitive constraints on offline social networks when examining the impact of social media use on social relationships.

Introduction

One of the primary reasons people use the Internet is to communicate with others.¹ With e-mail, instant messaging (IM), social network sites (SNS), blogs, and most recently Twitter, the Internet offers a plethora of ways to communicate with a large number of friends and strangers. Given this extensive social use of the Internet, a key question is: Does Internet use have a positive or negative impact on the number and quality of social relationships people can maintain with others?

Very broadly, the research in this area can be split into two camps, which have been described as “cyberpessimists” and “cyberoptimists.”² Cyberpessimists assert that Internet use has a negative effect on social life. Thus Nie³ argued that time is inelastic, and found that time spent on the Internet displaces time spent on socializing, particularly with family and doing face-to-face activities (see also Mesch⁴ and Nie and Hillygys⁵). This is often referred to as the “time displacement” hypothesis.⁶ In contrast, the “cyberoptimists” point to findings showing that the Internet has a positive effect on social life. Communication over the Internet tends to supplement, rather than replace, other forms of communication,^7,8 and can have positive effects on building and maintaining strong and satisfying social relationships (for reviews, see Bargh and McKenna¹ and Valkenburg and Peter⁹).

Both sides in this debate recognize that the impact of the Internet on social life is likely to vary according to the type of Internet use, the type of social relations, and the characteristics of the Internet user.^8,10 For example, there are effects of personality type on the impact of Internet use on social life (e.g., Kraut et al.¹¹ and Valkenburg and Peter¹²). Further, specifically social use of the Internet may be expected to have a different impact on social relations than using the Internet for gathering information.

Thus more recent research has moved away from examining overall Internet use and focused specifically on how the use of social media (such as IM and SNS) impacts on social relationships (e.g., Boyd and Ellison¹³). Interaction via social media such as IM and SNS tends to be with existing friends rather than with strangers.¹³ Some studies have shown that use of these media can be effective at building high-quality friendships and have an overall positive impact on well-being.^12,14–16 Further, if maintaining relationships via SNS is more temporally efficient and cognitively effective than maintaining relationships using other modes of communication, SNS use may allow for weak relationships to be maintained at higher levels of trust.¹⁷ However, other research suggests that use of social media can have a negative impact on social relationships and well-being.^18–20 In particular, the quality of communication over social media, compared with face-to-face communication or talking over the phone, has been highlighted as a potential limiting factor in using these social media to build strong, emotionally intense relationships.^14,18,21,22

The role of time constraints in affecting how Internet use impacts on social relationships has been extensively discussed (e.g., Bargh and McKenna¹ and Nie³). However, there are also likely to be cognitive constraints on the number of relationships that can be maintained at particular levels of emotional intensity.^23,24 Maintaining social relationships over time is a cognitively complex task, and individuals with better social cognitive abilities have larger personal networks.²⁵ Given these constraints on network size, the question becomes: Does use of social media allow these time and cognitive constraints to be overcome, and allow for more relationships and/or stronger relationships with existing network members?^17,26

In this study, we focus specifically on the use of IM and SNS, rather than Internet use in general or use of chatrooms, as use of these social media appears to have most impact on social life.¹² We examine how time spent using these social media affects the number and quality of relationships in an individual's circle of family and friends—their personal network.²⁷

We extend previous work in this area in two key ways. First, we asked participants to list explicitly each member of their personal, offline network. These personal networks consist of a series of subgroupings, arranged in a hierarchically inclusive sequence.^24,28 The successive grouping layers increase in size but decrease in the intensity of the typical relationship.²⁹ Thus we asked participants not just about the small number of people closest to them, but about all three layers of their network—the support group, the sympathy group, and the active network.^24,28 The mean size of these groupings is around 5, 15, and 50–150 respectively,^24,28, although they tend to vary depending on how the layers have been defined. Members of the support and sympathy group tend to be contacted frequently, and provide material and emotional support.³⁰ Members of the active network tend to be contacted much less frequently, and these relationships are much less emotionally intense.²⁶ The study therefore includes both strong relationships (support and sympathy group) and weak relationships (active network), enabling us to identify precisely which network layer social media use has an impact on. In contrast, previous research has not asked for details of specific relationships but tended to use overall measures of social capital, social support, well-being, or closeness to friends (e.g., Valenzuela et al.,² Wellman et al.,⁸ Valkenburg and Peter,¹² Ellison et al., ³¹ and Lee³²).

Second, previous work has simply focused on the number of relationships an individual has (e.g., Kraut et al.³³) or the frequency of communication with network members (e.g., Pew Internet⁷). However, quantity does not necessarily equal quality. Those with larger networks tend to be less emotionally close to each network member, suggesting a trade-off between the quantity and the quality of relationships and the influence of constraints on network size.²⁶ Thus in addition to offline network size, we also examine what impact social media use has on the emotional closeness of relationships with specific offline network members.

In light of the contradictory findings of previous research in this area, we do not make specific hypotheses about the effect of social media use on personal networks, but instead identify two key research questions:

RQ1: What effect does the use of social media have on the size of each offline network layer (support group, sympathy group, active network)?

RQ2: What effect does the use of social media have on the emotional closeness of relationships within each offline network layer?

Methods

Participants

Due to the length of the questionnaire (which typically takes between one and two hours to complete), participants were recruited via the personal networks of students. Eight students (four male and four female) taking part in a compulsory practical course at a large European university were asked to complete a questionnaire, and were asked to hand out 19 surveys to friends, colleagues, and family. Participation was voluntary and anonymous (use of sealed envelopes to return surveys). This procedure was approved by the psychology ethics committee at the University where the study was carried out.

The final sample consisted of 117 participants, 73 women and 44 men. The return rate of completed questionnaires within a month was 73%. The sample was a mixture of university students and other participants not at university (71% aged 25 or younger; 29% aged older than 25). The mean age of the participants was 28 years (SD = 12 years; range 18–63 years), and 86% of the participants did not have a completed university degree.

Questionnaire

After some basic socio-demographic questions (including age, gender, educational attainment, and partner status), participants listed their complete, offline social network. This consisted of all their living relatives (kin, step, and adopted) and their friends and acquaintances. To elicit friends and acquaintances, participants were asked to look through their mobile telephone, e-mail lists, and address books, and to include everyone they personally know and would like to continue to have a personal relationship during the next year, as in previous research.²⁶ Participants were asked how emotionally close they felt to each person on a 1 to 10 scale (1 = “not close at all” and 10 = “very close”), and when they last had face-to-face contact with this person. Subsequently, they completed items on their media usage. Participants reported whether or not they used IM or a SNS (“Do you use an IM program or SNS?” yes/no). If they reported using a SNS, they indicated how many “friends” they had in their online social network (typically listed as number of friends in Facebook, Hyves, or Netlog), and then estimated how many of those online friends they contacted weekly in any way. Participants also reported how many minutes on a typical day they used SNS and IM respectively (e.g., “On an average day, how many minutes do you spend on a social network site?” … minutes).

The support group was defined as those individuals with whom the participant had weekly contact and who scored eight or higher on the emotional support measure. The sympathy group was defined as the individuals with whom the participant had contact within the last month. Weekly and monthly contacts have been shown to correspond to the support group and sympathy group respectively.^29,34,35

Conceptually, the layers of the social network are hierarchically inclusive.³⁴ However, in this study, each layer was analyzed separately so as only to include the “extra” individuals in that layer in each set of analyses.³⁵ Thus for the sympathy-group analysis, the members of the support group were not included, otherwise these offline network members would be included in two sets of analyses. Similarly, the outer layer of the network was calculated by subtracting the support-group and sympathy-group size from the total offline social network size. For the outer layer, we only counted the number of friends and not the number of family members: social media use might lead to adding more friends to the outer layer but not to adding more family, as the total number of relatives is not under an individual's control. For each of the layers (support group, sympathy group, outer layer), we calculated the average emotional closeness to individuals in that layer.

Statistics

After presentation of descriptive statistics, correlations are presented for the three network layers (outer layer, sympathy group, support group) and new media usage. As a first step, we correlated the size of each network layer to social media usage. We subsequently tested whether controlling for potentially confounding variables substantially altered the strength of the correlation (partial correlations). As control variables, we used whether or not the participants had completed a university degree (0 = “no”; 1 = “yes”), their age, and their gender (0 = “male”; 1 = “female”), as these variables tend to affect network size and composition.^26,36 The same procedure was followed to examine if social media usage predicted average emotional closeness at any of the three layers.

We carried out two sets of analyses. First, we examined whether users of social media, compared with participants who did not use social media at all, had larger offline social networks or were closer to offline network members. Second, for only those participants who did use social media, we examined whether the amount of time spent using social media was associated with larger online or offline networks or closer relationships with offline network members.

Correlations are relatively robust to deviations from normality,³⁷ but where necessary variables were transformed to normalize the data as far as possible. Sensitivity power analysis by use of G-Power 3.1.2,³⁸ showed that our sample size allows detecting even weak effect sizes (r = 0.25; Power = 0.8).

Results

Descriptive statistics

Mean support group size was 6.75 (SD = 4.05; range = 0–25; mean friends = 3.51, mean family = 3.24) and mean sympathy group size was 10.56 individuals (SD = 7.79; range = 0–42; mean friends = 4.46, mean family = 6.10). The mean number of friends in the outer layer was 22.16 (SD = 0–84), giving a total mean offline network size (consisting of support group, plus sympathy group, plus active network of friends) of 39.48 (SD = 22.48; range = 3–117).

Of the 117 participants, 102 indicated that they currently used IM or SNS. Of those who did use IM or SNS, the online social network of individuals (“SNS friends”) consisted on average of 180.42 individuals (SD = 117.94; range 2–614). Of this online social network, participants reported that they contacted 13.94 individuals weekly (SD = 18.82, range =0–150). In order to reduce the effect of potential outliers, online social network size and individuals contacted weekly online were log-transformed. This made the distributions of these variables not significantly different from a normal distribution (online social network size: Shapiro–Wilk's W = 0.98, p = 0.353; online weekly social network size: Shapiro–Wilk's W = 0.968, p = 0.074).

Participants who used social media reported that on average they spent 39 minutes on IM (SD = 52; range =1–360 min) and 10 minutes on SNS a day (SD = 11; range =1–60 min). In order to reduce the effect of outliers, we conducted a log transformation of these variables. This transformation made the distribution close to normal for IM (Shapiro–Wilk's W = 0.972; p = 0.122), but the log transformation (time spent on SNS) remained significantly different from a normal distribution (Shapiro–Wilk's W = 0.92; p = 0.0003). However, this log transformation was a strong improvement over the use of raw scores (Shapiro–Wilk's W = 0.877; p < 0.0001).

Is use of instant messaging or social networking sites associated with an increased offline and online social network size?

Participants who used an IM or SNS had slightly larger support groups, sympathy groups, and number of friends in the outer layer than participants who did not use IM or SNS at all (Table 1). However, these differences were not statistically significant, and remained not significant after controlling for age, gender, and having a university degree.

Table 1.

Raw and Partial Correlations Between Use of Social Media (Social Network Sites and Instant Messaging), Size of Network, and Closeness to Network Members. Partial Correlations Are Controlled for Age, Gender, and Having a University Education

	Users vs. non-user of IM or SNS		Min per day on SNS		Min per day on IM
Variable	Raw	Partial	Raw	Partial	Raw	Partial
Size of offline network layer
Support group	0.156	0.042	0.009	−0.017	−0.007	−0.073
Sympathy group	0.061	−0.043	−0.148	−0.123	0.120	0.150
Outer layer	0.135	0.128	−0.115	0.014	0.014	0.072
Online network
Size of online network	—	—	0.214	0.214	0.118	0.117
Number of weekly contacts	—	—	0.334^**	0.289^*	0.236^*	0.147
Mean emotional closeness of offline network
Support group	−0.164	−0.062	0.174	0.187	0.093	0.137
Sympathy group	0.082	−0.082	0.03	−0.051	−0.115	−0.072
Outer layer	−0.101	−0.063	0.104	0.087	−0.030	−0.051

p < 0.05; **p < 0.01.

For participants who did use an IM or SNS, there was no significant correlation between time spent on IM and online social network size, but there was a significant correlation between time spent on IM and the number of individuals contacted online weekly. However, the partial correlation between time spent on IM and number of weekly online contacts (controlling for having a university degree or not, gender, and age) was not statistically significant. There was a significant correlation between time spent on SNS and online network size, as well as number of weekly online contacts. In contrast to the finding for IM, time on SNS remained positively associated with the total number of online contacts and the number of weekly contacts after controlling for age, gender, and having a university education. Thus those who spent more time on SNS reported more online contacts. However, the amount of time spent on IM (−0.07 < r < 0.12; p ≥ 0.26) or SNS (−0.148 < r < 0.009; p ≥ 0.19) did not correlate with size of any offline network layer. Controlling for age, gender, or having a university education did not substantially change the size or significance of these correlations.

The size of the online social network did not correlate with the size of any layer of the offline social network (r ≤ 0.12; p ≥ 0.2). The size of contacts contacted weekly online via SNS also did not correlate with the size of any network layer (r ≤ 12; p ≥ 0.2). The size of these correlations was not substantially altered when taking into account potentially confounding variables such as age, gender, or having a university education.

Is use of instant messaging or social networking sites associated with an increase in average emotional closeness to offline network members?

Participants who used IM or a SNS (yes/no) reported being slightly less close on average with their support group and friends in their outer layer than those who did not. However, these correlations were not statistically significant and reduced in magnitude after controlling for age, gender, and having a university education. There was no association between using IM or a SNS and closeness with the sympathy group and controlling for age, gender, and having a university degree did not alter this result. Finally, the amount of time spent on IM (−0.12 < r < 0.009; p > 0.25) or SNS (0.03 < r < 0.17; p > 0.1) did not correlate with emotional closeness at any network layer. Controlling for age, gender, and having a university education did not alter these results. Given that there is a trade-off between average emotional closeness of a layer and the size of that layer, we checked whether additionally controlling for the size of the layer altered the findings for emotional closeness. This partial correlation was not significant in any of the cases (p > 0.3).

Discussion

In this study, we examined whether using IM and SNS was associated with the number of relationships at each layer of the offline network and the quality of these relationships. While participants that spent more time using SNS had more online “friends” and had a greater number of weekly online contacts, this did not seem to translate into a larger offline network. Thus there was no relationship between time spent using IM or SNS and the size of any of the three layers of the offline network (support group, sympathy group, outer layer). Further, spending more time on IM or SNS did not increase the emotional closeness of relationships in any of these layers. These two results held even after controlling for age, gender, and having a university education. Finally, there was no difference in either offline network size or emotional closeness between those that used social media and those that did not use social media at all. Thus neither the use of IM and SNS nor the intensity of their use was associated with a greater number of offline relationships or the emotional closeness of these relationships.

These results contrast with previous findings, which suggested a positive impact of social media use on social relationships.^12,14 One possible reason for this difference is that the effects of social media use may be age specific, and the enhanced self-disclosure on social media may apply specifically to adolescents who are particularly prone to shyness and self-consciousness.¹² Further, previous research in this area has tended to rely on general measures of social well-being (social support, social capital, overall closeness to friends). Asking participants to list explicitly and rate their closeness to each offline network member may give a more accurate assessment of the composition of offline networks.

A more fundamental reason behind these results may be that social media use does not relax the time and cognitive constraints on offline network size sufficiently to allow for either larger networks or closer relationships with each network member. The findings of Nie³ have come in for sustained criticism (e.g., Bargh and McKenna¹), but his basic point is surely beyond argument: time is inelastic and there is a limited amount of leisure time in any given day. Time is a crucial constraint shaping primate sociality,³⁹ and recent diary studies with adolescents have shown that time spent using a computer does negatively affect time spent interacting with parents, though not time spent with friends.^3,14

There are some limitations to the current research. Due to the length of the questionnaire participants were required to complete, this study used snowball sampling to recruit participants. This method has been successfully used in previous studies,^26,40 and one advantage is that it results in data from a broader range of ages and backgrounds than a typical student sample.⁴¹ It remains to be seen, however, if these findings generalize to the broader population as a whole. Further, this study used a cross-sectional rather than a longitudinal design. A longitudinal design would more clearly allow disentangling if and how increased social media usage influences the size and composition. Finally, in this study we used relatively crude measures for IM and SNS usage, not distinguishing between the various different activities participants may be doing on SNS. Moreover, the amount and SNS was relatively low, and, as such, it is possible that our sample did not contain sufficient “heavy users.” In spite of these limitations, our results suggest that use of SNS and/or IM appear to have a very limited impact on either the size or closeness of offline social networks.

There are two key issues for future work to address. First, exactly how the time and cognitive constraints act to limit offline network size and closeness in relationships is currently unclear.^23,24 A better understanding of these constraints would allow us to make more precise predictions about the effect of current—and potential new forms—of social media on the size and nature of offline networks. It may be that different constraints operate at different layers of the offline network.²⁵ Time diary studies, combined with detailed measures of offline and online social networks, would allow us to understand how much time people devote to socializing, how they divide this social time across these networks, how often they use different modalities of communication, and how these factors impact on network size and closeness to specific network members both offline and online. The cognitive constraints are also incompletely understood, but recent brain-imaging research^42,43 suggests that these cognitive limits may be relatively hardwired, and thus may not be lifted by the use of social media (see also Stiller and Dunbar²⁵).

A second and related issue for future research is exploring how effective different communication modalities are at maintaining social relationships. Social relationships are prone to decay over time, and each relationship needs active maintenance to prevent this decay.^44,45 Even if time on the Internet is spent on social media rather than on non-social activities, these media may be less effective at building and maintaining emotionally intense relationships than other types of communication.^14,18,21,22 Thus one view is that computer-mediated communication (CMC), compared with face-to-face communication, has reduced visual, auditory, and contextual cues, and is less good at signalling affect.^3,21,46,47 However, other researchers argue that precisely because of these reduced cues, CMC actually allows for greater self-disclosure and thus for communication to become unusually intimate and “hyperpersonal,” leading to stronger relationships.^9,48,49 Longitudinal research on how the use of different types of social media impacts on how relationships develop and change over time would address how effective these media are at building and then maintaining social relationships.

Footnotes

Acknowledgments

SR is supported by funding from the EPSRC and ESRC as part of the “Developing Theory for Evolving Socio-Cognitive Systems” (TESS) project. SR and RD are also supported by funding from the SOCIALNETS project, which is part of the Pervasive Adaptation Initiative of the EU-funded 7th Framework Programme. The authors acknowledge and thank the students enrolled in the practical for their help with data collection and processing. The authors also thank two anonymous referees and the editor for helpful suggestions, which helped to improve a previous version of this paper.

Disclosure Statement

No competing financial interests exist.

References

Bargh

, McKenna

KYA

. The Internet and social life. Annual Review of Psychology, 2004; 55:573–90.

Valenzuela

, Park

, Kee

. Is there social capital in a social network site?: Facebook use and college students' life satisfaction, trust and participation. Journal of Computer-Mediated Communication, 2009; 14:875–901.

Nie

. Sociability, interpersonal relations, and the Internet: Reconciling conflicting findings. The American Behavioral Scientist, 2001; 45:420–35.

Mesch

. Family relations and the Internet: Exploring a family boundaries approach. The Journal of Family Communication, 2006; 6:119–38.

Nie

, Hillygus

. The impact of Internet use on sociability: Time-diary findings. IT & Society, 2002; 1:1–20.

Nie

, Hillygus

. Where does Internet time come from? A reconnaissance. IT & Society, 2002; 1:1–20.

The strength of Internet ties. 2006. Pew Internet & American Life Project. www.pewinternet.org/Reports/2006/The-Strength-of-Internet-Ties.aspx. 2010 Jan 6.

Wellman

, Quan-Haase

, Witte

et al. Does the Internet increase, decrease or supplement social capital? Social networks, participation and community commitment. The American Behavioral Scientist, 2001; 45:436–65.

Valkenburg

, Peter

. Social consequences of the Internet for adolescents: A decade of research. Current Directions in Psychological Science, 2009; 18:1–5.

10.

Tyler

. Is the Internet changing social life? It seems the more things change, the more they stay the same. Journal of Social Issues, 2002; 58:195–205.

11.

Kraut

, Kiesler

, Boneva

et al. The Internet paradox revisited. Journal of Social Issues, 2002; 58:49–74.

12.

Valkenburg

, Peter

. Online communication and adolescent well-being: Testing the stimulation versus the displacement hypothesis. Journal of Computer-Mediated Communication, 2007; 12:1169–82.

13.

Boyd

, Ellison

. Social network sites: Definition, history and scholarship. Journal of Computer-Mediated Communication, 2008; 13:210–38.

14.

Boneva

, Quinn

, Kraut

et al. 2006. Teenage communication in the instant messaging era. Kraut

, Brynin

, Kiesler

. Computers, phones, and the Internet: Domesticating information technology. New York: Oxford University Press.

15.

LaRose

, Eastin

, Gregg

. Reformulating the Internet paradox: Social cognitive explanations of Internet use and depression. Journal of Online Behavior, 2001. www.behavior.net/JOB/v1n2/paradox.htm. 2010 Jan 6.

16.

Liu

, Larose

. Does using the Internet make people more satisfied with their lives? The effects of the Internet on college students' school life satisfaction. CyberPsychology & Behavior, 2008; 11:310–20.

17.

Donath

. Signals in social supernets. Journal of Computer-Mediated Communication, 2008; 13:231–51.

18.

Schiffrin

, Edelman

, Falkenstern

et al. The associations among computer-mediated communication, relationships, and well-being. CyberPsychology, Behavior, & Social Networking, 2010, 10.1089/cpb.2009.0173.

19.

van den Eijnden

, Meerkerk

, Vermulst

et al. Online communication, compulsive Internet use, and psychosocial well-being among adolescents: A longitudinal study. Developmental Psychology, 2008; 44:655–65.

20.

Wolak

, Mitchell

, Finkelhor

. Escaping or connecting? Characteristics of youth who form close online relationships. Journal of Adolescence, 2003; 26:105–19.

21.

Cummings

, Butler

, Kraut

. The quality of online social relationships. Communications of the ACM, 2002; 45:103–8.

22.

Gross

, Juvonen

, Gable

. Internet use and well-being in adolescence. Journal of Social Issues, 2002; 58:75–90.

23.

Dunbar

RIM

. Cognitive constraints on the structure and dynamics of social networks. Group Dynamics: Theory, Research & Practice, 2008; 12:7–16.

24.

Roberts

SGB

. Constraints on social networks. Proceedings of the British Academy, 2010; 158:117–38.

25.

Stiller

, Dunbar

RIM

. Perspective-taking and memory capacity predict social network size. Social Networks, 2007; 29:93–104.

26.

Roberts

SGB

, Dunbar

RIM

, Pollet

et al. Exploring variations in active network size: Constraints and ego characteristics. Social Networks, 2009; 31:138–46.

27.

Quan-Haase

. University students' local and distant social ties: Using and integrating modes of communication on campus. Information, Communication & Society, 2008; 10:671–93.

28.

Zhou

, Sornette

, Hill

et al. Discrete hierarchical organization of social group sizes. Proceedings of the Royal Society B, Biological Sciences, 2005; 272:439–44.

29.

Hill

, Dunbar

RIM

. Social network size in humans. Human Nature: An Interdisciplinary Biosocial Perspective, 2003; 14:53–72.

30.

Plickert

, Cote

, Wellman

. It's not who you know, it's how you know them: Who exchanges what with whom? Social Networks, 2007; 29:405–29.

31.

Ellison

, Steinfield

, Lampe

. The benefits of facebook “friends”: Social capital and college students' use of online social network sites. Journal of Computer-Mediated Communication, 2007; 12:1143–68.

32.

Lee

. Online communication and adolescent social ties: Who benefits more from Internet use? Journal of Computer-Mediated Communication, 2009; 14:509–31.

33.

Kraut

, Patterson

, Lundmark

et al.

Internet paradox: A social technology that reduces social involvement and psychological well-being?

American Psychologist, 1998; 53:1017–31.

34.

Dunbar

RIM

, Spoors

. Social networks, support cliques, and kinship. Human Nature: An Interdisciplinary Biosocial Perspective, 1995; 6:273–90.

35.

Roberts

SGB

, Wilson

, Fedurek

et al. Individual differences and personal social network size and structure. Personality & Individual Differences, 2008; 44:954–64.

36.

McPherson

, Smith-Lovin

, Brashears

. Social isolation in America: Changes in core discussion networks over two decades. American Sociological Review, 2006; 71:353–75.

37.

Fowler

. Power and robustness in product-moment correlation. Applied Psychological Measurement, 1987; 11:419–28.

38.

Faul

, Erdfelder

, Lang

et al. G*Power 3: A flexible statistical power analysis program for the social, behavioral, and biomedical sciences. Behavioral Research Methods, 2007; 39:175–91.

39.

Dunbar

RIM

, Korstjens

, Lehmann

. Time as an ecological constraint. Biological Reviews, 2009; 84:413–29.

40.

Pollet

, Kuppens

, Dunbar

RIM

. When nieces and nephews become important: Differences between childless women and mothers in relationships with nieces and nephews. Journal of Cultural & Evolutionary Psychology, 2006; 4:83–93.

41.

Henry

. College sophomores in the laboratory redux: Influences of a narrow data base on social psychology's view of the nature of prejudice. Psychological Inquiry, 2008; 19:49–71.

42.

Lewis

, Rezaie

, Browne

et al. Orbitofrontal volume predicts both understanding of others and social network size. Cerebral Cortex; submitted.

43.

Powell

, Lewis

, Dunbar

et al. Orbital prefrontal cortex volume correlates with social cognitive competence. Neuropsychologia 10.1016/j.neuropsychologia.2010.08.004.

44.

Burt

. Decay functions. Social Networks, 2000; 22:1–28.

45.

Oswald

, Clark

. Best friends forever?: High school best friendships and the transition to college. Personal Relationships, 2003; 10:187–96.

46.

Daft

, Lengel

. Organizational information requirements, media richness and structural design. Management Science, 1986; 32:554–71.

47.

Kock

. Media richness or media naturalness? The evolution of our biological communication apparatus and its influence on our behavior toward e-communication tools. IEEE Transactions on Professional Communication, 2005; 48:117–30.

48.

Walther

. Computer-mediated communication: Impersonal, interpersonal, and hyperpersonal interaction. Communication Research, 1996; 19:3–43.

49.

Tidwell

, Walther

. Computer-mediated communication effects on disclosure, impressions, and interpersonal evaluations: Getting to know one another a bit at a time. Human Communication Research, 2002; 28:317–48.