With social networking sites (SNS) such as Facebook, individuals have immediate access to hundreds of people from different aspects of their lives. On one hand, this may increase the number of people that individuals can interact with directly because communication now requires less effort. On the other hand, individuals may still only interact with a small portion of their networks because humans have limited time and resources. Mayhew and Levinger (1976) proposed that because of time and resource constraints, individuals in larger networks spend, on average, less time with each contact. Thus, while people may have opportunities through Facebook to interact with more people, they may not actually do so. Using logging software, we explored the percentage of their Facebook network that individuals sent direct messages to and whether individuals with larger networks sent direct messages to a smaller percentage of their networks. We found that in line with Mayhew and Levinger’s claim, users messaged a very small percentage of their networks (less than two percent) and that users with larger friend networks sent messages to a smaller percentage of their friend networks. This suggests that while Facebook may enable users to connect with more people than ever before, there are still limits to the number of ties with whom individuals actively interact.
The Internet, and social networking sites (SNS) in particular, enables individuals to connect easily with people that they no longer see regularly such as high school friends, relatives, and former neighbors, while also strengthening relationships with new acquaintances, colleagues, and nearby friends (Manago, et al., 2012; Wellman, 2001). Nevertheless, the number of direct interactions (e.g., direct messages) individuals can have is finite even online because people have limited amounts of time and resources to spend each day. This suggests that users on SNS, such as Facebook, will not have direct interactions with a large percentage of their network. Moreover, users that are connected to a larger number of people may not have more direct interactions with more people than users who are connected to fewer people. Thus, easier access does not necessarily mean more direct interactions.
Network size and interaction time
In 1976, Mayhew and Levinger argued that the size of individuals’ networks affects the amount of time individuals interact with each member of their network because humans have limited time and resources. Individuals in larger networks, such as urban centers, are in contact with more individuals than individuals in smaller networks. Consequently, they may devote less time to each interaction or forgo interactions that are less important.
In support of their model, Mayhew and Levinger (1976) cited Dodd (1957), who found that individuals from larger cities were less likely to respond to leaflets that had been dropped on the city than individuals from smaller towns. They concluded that the larger city’s lower response rates were evidence that individuals from larger networks interacted with many others and as a result, did not have the time or resources to respond to a stranger’s request when they had competing demands. However, individuals’ single response to a leaflet may not reflect their interactions with known others.
More recent studies have examined how off-line network size affects interactions with known others. When asked to report for each member of their networks, the amount of time that had elapsed since they last interacted, women with more extended family members reported longer average times from the last interaction (Roberts and Dunbar, 2011). With more people in their networks to divide their time between, people with larger networks do not see each member of their networks as often as people with smaller networks.
Moreover, when asked to report for each member of their networks, how close they felt to them, individuals with larger networks reported lower emotional closeness on average (Roberts, et al., 2009). As people with larger networks see each member of their networks less frequently and it takes time and effort to maintain emotionally close relationships (Oswald and Clark, 2003), people with larger networks may feel less close to each member of their networks. These studies support Mayhew and Levinger’s model (1976) that the average amount of time people spend with each member of their networks depends on the size of their networks.
With the rise of SNS, the relation between network size and interaction time may no longer exist because people can now communicate with their networks quickly and easily. By directly examining the effect of network size on the percentage of direct interactions in users’ Facebook networks, we can test Mayhew and Levinger’s claim in an online, modern-day context. Computer-mediated networks such as Facebook enable users to gain access to resources such as social support, companionship, and information (e.g., news, health advice) that are exchanged between people (Wellman, et al., 1996), making them similar to off-line communities. For example, college students have reported using Facebook to gain access to resources such as information about upcoming events and social support in the form of companionship and intimacy from friends (Lewis and West, 2009; Pempek, et al., 2009; Wellman, et al., 1996).
Interaction overload on Facebook
The average number of friends people have on Facebook exceeds the average number of social relationships that people can manage off-line (Duggan, et al., 2015; Manago, et al., 2012). People are estimated to have on average 10–20 strong relationships and 125 social relationships (Hill and Dunbar, 2003; Parks, 2007), but the median number of friends on Facebook is 200 among adult users (Duggan, et al., 2015) and 370 for young adults/college students (Manago, et al., 2012).
While having large networks of friends is common on Facebook, the number of close relationships on Facebook does not surpass the number of close relationships found in traditional off-line networks. The average number of active social relationships on Facebook, which was determined by participants’ rating of the social relationship between them and each Facebook friend, was 105 (Arnaboldi, et al., 2013). Likewise, the number of Facebook friends whom people would consider “going to for advice or sympathy in times of great emotional or other distress” is also consistent with the number reported in studies of off-line networks (Dunbar, 2016). Despite the ease of communication online, there are still constraints on interactions.
In further support of Mayhew and Levinger’s hypothesis, Facebook users have different communication expectations for their Facebook friends depending on the closeness of the relationship. This suggests that investing the same amount of time to all of their interactions is not feasible. Although users frequently send private messages to their friends (Reich, 2010), Bryant and Marmo (2012) found that expectations for contact on Facebook varied with emotional closeness. The authors found that for close friends (i.e., strong ties), greater contact (i.e., a private message) was expected as compared to casual friends and acquaintances (i.e., weak ties).
Instead of contacting casual friends and acquaintances directly, users may engage in passive “stalking” interactions, where they browse peers’ profiles and photos to “keep up” with what their peers are doing (Lewis and West, 2009; Reich, 2010; Yang and Brown, 2013). This enables users to feel connected without expending much effort.
While relying on self-reported perceptions of behavior, these findings suggest that because of the large size of Facebook networks, Facebook users may not interact directly with a large percentage of their network. To date, however, research has yet to explore if this is true using objective measures (logged data), rather than fallible self-report (Junco, 2013). One exception is a study that found that on Facebook, people interacted with a few members of their networks through posts on their timelines frequently and with most members infrequently, much like in off-line networks, (Dunbar, et al., 2015).
Furthermore, although Facebook is touted as a way to facilitate and strengthen ties (Burke and Kraut, 2014), the relationship between network size and direct communication frequency is not known. Thus, in this paper, we explore whether having more connections on Facebook (i.e., friends) also increases opportunities to interact directly with more people or whether there is a limit to the number of people with whom individuals can interact.
Data for this study are part of a larger project funded by the U.S. National Science Foundation on the multitasking behaviors of millennials. Undergraduate students in the United States (n = 76) had their computer and phone activities, including Facebook activities, logged for one week. Participation was limited to Android and Windows users because the logging software was not compatible with Apple devices. Participants were paid US$100 and a university Institutional Review Board approved all study procedures.
Fifteen people were excluded from the analytic sample because they did not log time or activities on Facebook during the week or were missing demographic or computer data (e.g., software uninstalled by antivirus software or participant). One of these participants was also excluded because he or she only had four Facebook friends, which is well outside the range found by other studies of college-aged users (e.g., Manago, et al., 2012). The final sample was comprised of 61 participants (57 percent female) between the ages of 18 and 23 years (M= 19.4) and was diverse, with participants identifying predominantly as Asian American (46 percent) and Hispanic (31 percent) (see Table 1). Students in their first year of their undergraduate studies (36 percent), second (28 percent), third (23 percent), and fourth year and beyond (13 percent) were represented.
On average, participants started using the Internet at the age of 10 years (SD=2.5) and their first SNS before the age of 14 years (SD=1.8). They did not obtain their first smartphone however, until on average, the age of 17 years (SD=1.8). Participants’ networks ranged from 92 to 1207 friends (M=462.93, SD=271.05). They messaged between 0 and 23 friends and the percent of their network messaged ranged from 0 percent to 5.26 percent (M=1.74 percent, SD=1.29 percent). On average, participants spent 194 minutes (SD=193.10) on Facebook during the course of the study (one week).
To illustrate the relationship between network size and number of friends messaged, percent of friends messaged, and number of minutes spent on Facebook, participants were divided into quartiles based on network size and descriptives for each quartile are provided in the Appendix (see Table A.1).
Table 1: Sample demographics.
Mean (SD) Percentage (n) Gender Male 43% (26) Female 57% (35) Age 19.39 (1.26) Year in school First 36% (22) Second 28% (17) Third 23% (14) Fourth or beyond 13% (8) Ethnicity Asian American 46% (28) Hispanic 31% (19) White 15% (9) African American 3% (2) Other or did not disclose 5% (3) Age started using Internet 10.1 yrs (2.47) Age started using SNS 13.6 yrs (1.83) Age first Smartphone 17.2 yrs (1.80) Facebook network size 462.93 (271.05) Number of friends messaged 6.84 (4.62) Percent of network messaged 1.74 (1.29) Time spent on Facebook (min.) 193.87 (193.10)
Participants arrived at the lab so that logging software could be installed onto their laptops and cell phones. The software was installed for about seven days and participants were encouraged to continue using their devices as they usually did.
To automatically record participants’ Facebook activities, we developed a Facebook logging application using PHP and the Facebook Application Programming Interface (API). The app captured the size of participants’ networks and the number of friends with whom they interacted directly during the study period.
In this study, we define direct interactions as those that take place through private messages. Posts that participants shared on their friends’ timelines were not captured in order to protect individuals’ who had not consented to participate in the study. Similarly, the content of messages or the identities of participants’ friends were not captured due to privacy concerns.
Nevertheless, private messages may be a more accurate representation of private face-to-face conversations than timeline posts because they are perceived as being more appropriate for intimate disclosures (Bazarova, 2012; Bazarova and Choi, 2014). Timeline posts are visible not only to the recipients, but to their entire network as well. Participants’ Facebook activities were stored using a MySQL database and were retrieved retroactively at the end of study.
Researchers also installed the logging software KidLogger (2013) and the AWARE Framework (2013) onto participants’ laptops and cell phones respectively so that the amount of time participants spent on Facebook could be captured. At the end of the study, participants returned to the lab and these apps were removed from their devices.
Background survey. Participants reported their gender, date of birth, year in their undergraduate program, ethnicity, age at which they started using the Internet, age at which they started using SNS, and age at which they obtained their first smartphone.
The percent of network messaged was obtained by summing the number of unique individuals that participants messaged over the course of the week, dividing by the total number of friends, and multiplying by 100. The average percent of network messaged was then calculated for the entire sample.
We used hierarchical regression to test the association between network size and percent of network messaged. In Model 1, we tested the effect of network size.
In Model 2, we tested the effect of the number of days spent in the study. Due to scheduling, participants were in the study for between 6–10 days (M=7.08, SD=0.73).
In Model 3, we tested the effect of whether or not participants were first-year students. First-year students may message a larger percentage of their network as they try to stay in contact with friends and family back home while also expanding their network at school. Other work with this age group has found that sending messages and commenting were more common among first-year students than more senior students (Yang and Brown, 2013).
In Model 4, we tested the effect of self-identified gender (0=female, 1=male) since research has found that adolescent girls are more likely than boys to establish intimacy through discussion (McNelles and Connolly, 1999).
In Model 5, we tested the effect of ethnicity because previous studies among college students found that Hispanic participants were less likely to use Facebook than other ethnic groups (Hargittai, 2007; Subrahmanyam, et al., 2008). Ethnicity was included in our models as four dichotomous variables (i.e., whether participants identified as African American, Hispanic, White, or Other/Did not disclose). Asian-Americans were the reference group because it was the largest ethnic group represented in the study.
On average, participants messaged a very small percentage of their network (M=1.74, SD=1.29, range: 0–5.26). Participants averaged networks of over 450 friends but messaged, on average, less than seven of them.
As hypothesized, the percent of network messaged decreased as network size increased (see Figure 1).
Figure 1: The association between network size and percent of network contacted.
In Model 1, network size was negatively associated with the percent of network messaged, b=-0.002, t(59)=-2.88, p=0.005 (see Table 2). An increase in the number of friends by one was associated with messaging 0.002 percent less of the network. Network size also explained a significant proportion of variance in percent of network messaged, R2=.12, F(1, 59)=8.31, p=0.01.
When days in the study was added in Model 2, the effect of network size did not change; however, the total variance explained by the model increased from 12 percent to 17 percent, F(2, 58)=6.05, p=0.00. Number of days in the study had a marginally positive effect; participants who were in the study for longer messaged a larger percent of their networks, b=0.39, t(59)=1.86, p=0.07.
Adding a covariate in Model 3 for whether participants were first year students, reduced the coefficient of network size to from -0.002 to -0.001, t(57)=-2.63, p=0.011. First year students also messaged on average, a smaller percentage of their network than non-first year students, b=0.53, t(57)=-1.68, p=0.10. This marginal effect may be explained by the fact that first-year students had larger networks (M=520.32, SD=266.79) than non-first-year students (M=430.56, SD=274.60), although this difference was not significant. Whether participants were first-year students explained an additional four percent of the model, F(3, 57)=5.10, p=0.00.
The effect of network size remained consistent when adding covariates for ethnicity and whether participants identified as female. There were no associations between gender, ethnicity, and percent of network messaged (Models 4–5). In Model 5, the effect of study duration became significant b=0.49, t(52)=2.25, p=0.03, and the effect of whether participants were first-year students no longer remained significant. The variance explained by the final model was 26 percent, F(8, 52)=2.24, p=0.04.
Table 1: Hierarchical regression predicting Percent of network messaged.
Notes: Asian is the omitted race category. Standard errors in parentheses. Betas are unstandardized. †p<0.10; *p<0.05; **p<0.01.
Unstandardized Betas Model 1 Model 2 Model 3 Model 4 Model 5 Network size -.002** -.002** -.001* -.001* -.001* (-.001) (-.001) (-.001) (-.001) (-.001) Study duration (days) .39† .39† .40† .49* (.21) (.21) (.21) (.22) First-year student -.53† -.54† -.53 (.32) (.31) (.33) Female -.22 -.33 (.31) (.32) Ethnicity African American -1.17 (.91) Hispanic/Latino(a) -.23 (.40) Other or did not disclose .43 (.73) Constant 2.51 -.25 -.16 -.16 -.55 (.31) (1.52) (1.49) (1.50) (1.55) N 61 61 61 61 61 R2 .12 .17 .21 .22 .26 ΔR2 — .05 .04 .01 .04 F 8.31 6.05 5.10 3.92 2.24
To find the best fitting model, we tested a quadratic term for percent of friends messaged. The interaction term was not significant (see Table A.2 in the Appendix), suggesting that a linear model is more appropriate than a curvilinear one.
Facebook is a highly popular and commonly used SNS. Numerous studies have found that teen and young adult users value its importance in staying up-to-date on social events, interacting with friends, and having private and easily accessible communication with others (e.g., Reich, et al., 2012; Subrahmanyam, et al., 2008; Anderson, et al., 2012). As such, Facebook could strengthen relationships with close friends by allowing communication with more people as network size grows.
However, we found that on average users interact directly with a very small percentage of their network. Moreover, in line with Mayhew and Levinger’s hypothesis, the number of people messaged in a week does not grow proportionately as the network increases. This results in direct communication with a smaller percentage of the network, even after controlling for time in the study, whether or not participants were first year students, gender, and ethnicity. Given that users, on average, have trouble identifying over a quarter of their Facebook friends by name and that users with more friends identify a smaller percentage of their network correctly, this is not surprising (Croom, et al., 2016). Thus, users may reserve their time and effort for closer friends in their network.
Interestingly, time in the study predicted direct communication with a larger percentage of the network. This suggests that people have limited resources each day to directly and privately communicate with friends on Facebook, but distribute those resources differentially, day-to-day, to reach more people in their Facebook network.
Limitations and future directions
Given our unobtrusive data collection method, some limitations exist. First, due to privacy issues, we were not able to capture posts on friends’ timelines. This limitation of the data likely underestimated the number of direct interactions with friends. However, it is also possible that the number of timeline posts decreases at a slower rate as network size increases because timeline posts are shorter and less personal (Bazarova and Choi, 2014) and therefore may require less time and resources to compose. Future studies could use diary methods to allow users to report when they posts on their friends’ timelines.
Second, it is also worthwhile to note that participants could be interacting with their Facebook friends through different platforms, as college students report changing the platforms they use with their friends as the relationship develops (Yang, et al., 2013). Thus, future work might want to consider the use of other SNS or messaging platforms.
Our findings suggest that individuals do not use Facebook to interact directly with hundreds or even dozens of friends. The affordances of SNS such as Facebook may not be that they increase connectedness by enabling individuals to easily interact directly with many. Rather, as previous research suggests, it may increase connectedness by enabling individuals to maintain many ties with casual friends and acquaintances through interacting passively or publicly. As Burke and Kraut (2014) found, users felt closer to Facebook friends even if their interactions consisted of passively viewing profiles, news feed stories, and photos.
Users with large networks may still benefit from having many passive or inactive weak ties. Such weak ties can facilitate connections across networks, the spread of information and influence, community organization, and mobility opportunities (Coleman, 1988; Granovetter, 1973). Likewise, more contemporary researchers find that through weak ties with casual friends and acquaintances, university students gain new information and perspectives on Facebook (e.g., Ellison, et al., 2011, 2007; Steinfield, et al., 2008), which then increase feelings of connectedness on campus.
However, having large networks does not seem to facilitate opportunities for direct interactions with more individuals per day or week. This suggests that there may be a limit to the number of people individuals can interact with directly, even when communication is easy and always available. While the ability to connect to hundreds of people with ease increases, our time and ability to support intimate connections with others may not. The amount of direct communication on Facebook, which is typically private and personal, does not increase in proportion to network size growth.
While these findings are not surprising, they do reify that even with increasing technological innovation and time online, patterns from studies of off-line networks still hold true. A theory proposed over 40 years ago about face-to-face social networks, still applies to digitally mediated networks of today.
About the authors
Joanna C. Yau is a doctoral candidate in Education at the University of California, Irvine. Her interdisciplinary research uses techniques from psychology, education, and human-computer interaction to understand the effects of technology and media on interpersonal relationships, well-being, and learning.
Direct comments to: jcyau [at] uci [dot] edu
Stephanie M. Reich is an Associate Professor of Education, Informatics, and Psychology and Social Behavior at the University of California, Irvine. Her research explores direct and indirect influences on children’s health, education, and well-being, specifically through the family, online, and school environments.
E-mail: smreich [at] uci [dot] edu
Yiran Wang received her Ph.D. in Informatics from the University of California, Irvine. Her research interests are in human computer interaction (HCI) and social computing and she studies the use and effects of social media in college life. Currently, she conducts user experience research at Google.
E-mail: yiranw2 [at] uci [dot] edu
Melissa Niiya earned an MA in Education from UC Irvine and a BFA in Writing for Screen and Television from the University of Southern California. She works at Portland Public Schools, where she supports data-driven policy-making. Her research interests include technology in literacy education, using digital technologies to close opportunity gaps, and using data to promote equitable decision-making.
E-mail: m [dot] k [dot] niiya [at] gmail [dot] com
Gloria Mark is a Professor in the Department of Informatics, University of California, Irvine. Her research focuses on studying how the use of digital technology impacts our lives in real-world contexts. Her current projects include precision tracking of people’s digital media use: how it affects multitasking, focus of attention, interruptions, mood, and stress. She uses sensors and other mixed methods to study this. She received her Ph.D. in psychology from Columbia University. Prior to UCI she worked at the German National Research Center for Information Technology (GMD, now Fraunhofer Institute) and has been a visiting researcher at Microsoft Research, IBM, and National University of Singapore. She is a research affiliate at the MIT Media Lab. She has published in top conferences and journals in the fields of human-computer interaction (HCI) and computer-supported cooperative work. She is the general chair for the ACM CHI 2017 conference, and is on the editorial board of ACM TOCHI and Human-Computer Interaction. Her work has appeared in the popular press such as the New York Times, Atlantic, BBC, NPR, Time, and Wall Street Journal and she has presented her work at SXSW and the Aspen Ideas Festival.
E-mail: gmark [at] uci [dot] edu
This study was funded by the National Science Foundation under grant number 1218705.
Beth Anderson, Patrick Fagan, Tom Woodnutt, and Tomas Chamorro-Premuzic, 2012. “Facebook psychology: Popular questions answered by research,” Psychology of Popular Media Culture, volume 1, number 1, pp. 23–37.
doi: http://dx.doi.org/10.1037/a0026452, accessed 20 April 2018.
Valerio Arnaboldi, Andrea Guazzini, and Andrea Passarella, 2013. “Egocentric online social networks: Analysis of key features and prediction of tie strength in Facebook,” Computer Communications, volume 36, numbers 10–11, pp.1,130–1,144.
doi: https://doi.org/10.1016/j.comcom.2013.03.003, accessed 20 April 2018.
AWARE Framework, 2013. “AWARE” [Mobile app], at http://www.awareframework.com/, accessed 20 April 2018.
Natalya N. Bazarova, 2012. “Public intimacy: Disclosure interpretation and social judgments on Facebook,” Journal of Communication, volume 62, number 5, pp. 815–832.
doi: https://doi.org/10.1016/j.comcom.2013.03.003, accessed 20 April 2018.
Natalya N. Bazarova and Yoon Hyung Choi, 2014. “Self-disclosure in social media: Extending the functional approach to disclosure motivations and characteristics on social network sites,” Journal of Communication, volume 64, number 4, pp. 635–657.
doi: https://doi.org/10.1111/jcom.12106, accessed 20 April 2018.
Erin M. Bryant and Jennifer Marmo, 2012. “The rules of Facebook friendship: A two-stage examination of interaction rules in close, casual, and acquaintance friendships,” Journal of Social and Personal Relationships, volume 29, number 8, pp. 1,013–1,035.
doi: https://doi.org/10.1177/0265407512443616, accessed 20 April 2018.
Moira Burke and Robert E. Kraut, 2014. “Growing closer on Facebook: Changes in tie strength through social network site use,” CHI ’14: Proceedings of the SIGCHI Conference on Human Factors in Computing Systems, pp. 4,187–4,196.
doi: https://doi.org/10.1145/2556288.2557094, accessed 20 April 2018.
James S. Coleman, 1988. “Social capital in the creation of human capital,” American Journal of Sociology, volume 94, pp. S95–S120.
doi: https://doi.org/10.1086/228943, accessed 20 April 2018.
Charles Croom, Bay Gross, Larry D. Rosen, and Brad Rosen, 2016. “What’s her Face(book)? How many of their Facebook ‘friends’ can college students actually identify?” Computers in Human Behavior, volume 56, pp. 135–141.
doi: https://doi.org/10.1016/j.chb.2015.11.015, accessed 20 April 2018.
Stuart C. Dodd, 1957. “A power of town size predicts an internal interacting: A controlled experiment relating the amount of an interaction to the number of potential interactors,” Social Forces, volume 36, number 2, pp. 132–137.
doi: https://doi.org/10.2307/2573848, accessed 20 April 2018.
Maeve Duggan, Nicole B. Ellison, Cliff Lampe, Amanda Lenhart, and Mary Madden, 2015. “Social media update 2014,” Pew Research Center (9 January), at http://www.pewinternet.org/2015/01/09/social-media-update-2014/, accessed 15 July 2017.
R.I.M. Dunbar, 2016. “Do online social media cut through the constraints that limit the size of offline social networks?” Royal Society Open Science, volume 3 (20 January), at http://rsos.royalsocietypublishing.org/content/3/1/150292, accessed 20 April 2018.
doi: https://doi.org/10.1098/rsos.150292, accessed 20 April 2018.
R.I.M. Dunbar, Valerio Arnaboldi, Marco Conti, and Andrea Passarella, 2015. “The structure of online social networks mirrors those in the offline world,” Social Networks, volume 43, pp. 39–47.
doi: https://doi.org/10.1016/j.socnet.2015.04.005, accessed 20 April 2018.
Nicole B. Ellison, Charles Steinfield, and Cliff Lampe, 2011. “Connection strategies: Social capital implications of Facebook-enabled communication practices,” New Media & Society, volume 13, number 6, pp. 873–892.
doi: https://doi.org/10.1177/1461444810385389, accessed 20 April 2018.
Nicole B. Ellison, Charles Steinfield, and Cliff Lampe, 2007. “The benefits of Facebook ‘friends:’ Social capital and college students’ use of online social network sites,” Journal of Computer-Mediated Communication, volume 12, number 4, pp. 1,143–1,168.
doi: https://doi.org/10.1111/j.1083-6101.2007.00367.x, accessed 20 April 2018.
Mark S. Granovetter, 1973. “The strength of weak ties,” American Journal of Sociology, volume 78, number 6, pp. 1,360–1,380.
doi: https://doi.org/10.1086/225469, accessed 20 April 2018.
Eszter Hargittai, 2007. “Whose space? Differences among users and non-users of social network sites,” Journal of Computer-Mediated Communication, volume 13, number 1, pp. 276–297.
doi: https://doi.org/10.1111/j.1083-6101.2007.00396.x, accessed 20 April 2018.
Russell A. Hill and R.I.M. Dunbar, 2003. “Social network size in humans,” Human Nature, volume 14, number 1, pp. 53–72.
doi: https://doi.org/10.1007/s12110-003-1016-y, accessed 20 April 2018.
Reynol Junco, 2013. “Comparing actual and self-reported measures of Facebook use,” Computers in Human Behavior, volume 29, number 3, pp. 626–631.
doi: https://doi.org/10.1016/j.chb.2012.11.007, accessed 20 April 2018.
KidLogger, 2013. “Do you know what your kids are doing online?” at http://kidlogger.net/, accessed 20 April 2018.
Jane Lewis and Anne West, 2009. “‘Friending’: London-based undergraduates’ experience of Facebook,” New Media & Society, volume 11, number 7, pp. 1,209–1,229.
doi: https://doi.org/10.1177/1461444809342058, accessed 20 April 2018.
Adriana M. Manago, Tamara Taylor, and Patricia M. Greenfield, 2012. “Me and my 400 friends: The anatomy of college students’ Facebook networks, their communication patterns, and well-being,” Developmental Psychology, volume 48, number 2, pp. 369–380.
doi: https://doi.org/10.1037/a0026338, accessed 20 April 2018.
Bruce H. Mayhew and Roger L. Levinger, 1976. “Size and the density of interaction in human aggregates,” American Journal of Sociology, volume 82, number 1, pp. 86–110.
doi: https://doi.org/10.1086/226271, accessed 20 April 2018.
Laurie R. McNelles and Jennifer A. Connolly, 1999. “Intimacy between adolescent friends: Age and gender differences in intimate affect and intimate behaviors,” Journal of Research on Adolescence, volume 9, number 2, pp. 143–159.
Debra L. Oswald and Eddie M. Clark, 2003. “Best friends forever? High school best friendships and the transition to college,” Personal Relationships, volume 10, number 2, pp. 187–196.
doi: https://doi.org/10.1111/1475-6811.00045, accessed 20 April 2018.
Malcolm R. Parks, 2007. Personal relationships and personal networks. Mahwah, N.J.: Erlbaum.
Tiffany A. Pempek, Yevdokiya A. Yermolayeva, and Sandra L. Calvert, 2009. “College students’ social networking experiences on Facebook,” Journal of Applied Developmental Psychology, volume 30, number 3, pp. 227–238.
doi: https://doi.org/10.1016/j.appdev.2008.12.010, accessed 20 April 2018.
Stephanie M. Reich, 2010. “Adolescents’ sense of community on Myspace and Facebook: A mixed–methods approach,” Journal of Community Psychology, volume 38, number 6, pp. 688–705.
doi: https://doi.org/10.1002/jcop.20389, accessed 20 April 2018.
Stephanie M. Reich, Kaveri Subrahmanyam, and Guadalupe Espinoza, 2012. “Friending, IMing, and hanging out face-to-face: overlap in adolescents’ online and offline social networks,” Developmental Psychology, volume 48, number 2, pp. 356–368.
doi: https://doi.org/10.1037/a0026980, accessed 20 April 2018.
Sam G.B. Roberts and R.I.M. Dunbar, 2011. “Communication in social networks: Effects of kinship, network size, and emotional closeness,” Personal Relationships, volume 18, number 3, pp. 439–452.
doi: https://doi.org/10.1111/j.1475-6811.2010.01310.x, accessed 20 April 2018.
Sam G.B. Roberts, R.I.M. Dunbar, Thomas V. Pollet, and Toon Kuppens, 2009. “Exploring variation in active network size: Constraints and ego characteristics,” Social Networks, volume 31, number 2, pp. 138–146.
doi: https://doi.org/10.1111/j.1475-6811.2010.01310.x, accessed 20 April 2018.
Charles Steinfield, Nicole B. Ellison, and Cliff Lampe, 2008. “Social capital, self-esteem, and use of online social network sites: A longitudinal analysis,” Journal of Applied Developmental Psychology, volume 29, number 6, pp. 434–445.
doi: https://doi.org/10.1111/j.1475-6811.2010.01310.x, accessed 20 April 2018.
Kaveri Subrahmanyam, Stephanie M. Reich, Natalia Waechter, and Guadalupe Espinoza, 2008. “Online and offline social networks: Use of social networking sites by emerging adults,” Journal of Applied Developmental Psychology, volume 29, number 6, pp. 420–433.
doi: https://doi.org/10.1016/j.appdev.2008.07.003, accessed 20 April 2018.
Barry Wellman, 2001. “Computer networks as social networks,” Science, volume 293, number 5537 (14 September), pp. 2,031–2,034.
doi: https://doi.org/10.1126/science.1065547, accessed 20 April 2018.
Barry Wellman, Janet Salaff, Dimitrina Dimitrova, Laura Garton, Milena Gulia, and Caroline Haythornthwaite, 1996. “Computer networks as social networks: Collaborative work, telework, and virtual community,” Annual Review of Sociology, volume 22, pp. 213–238.
doi: https://doi.org/10.1146/annurev.soc.22.1.213, accessed 20 April 2018.
Chia-chen Yang and B. Bradford Brown, 2013. “Motives for using Facebook, patterns of Facebook activities, and late adolescents’ social adjustment to college,” Journal of Youth and Adolescence, volume 42, number 3, pp. 403–416.
doi: https://doi.org/10.1007/s10964-012-9836-x, accessed 20 April 2018.
Chia-chen Yang, B. Bradford Brown, and Michael T. Braun, 2013. “From Facebook to cell calls: Layers of electronic intimacy in college students’ interpersonal relationships,” New Media & Society, volume 16, number 1, pp. 5–23.
doi: https://doi.org/10.1177/1461444812472486, accessed 20 April 2018.
Table A1: Descriptive statistics across friend network size quartiles.
Note: Standard deviations are listed in parentheses.
Q1 Q2 Q3 Q4 Number of friends Mean 193.29 (53.73) 346.67 (55.50) 477.93 (47.40) 870.80 (172.74) Median 194 350 471 829 Range 92–259 262–419 420–571 650–1207 Number of friends messaged Mean 5.00 (3.72) 6.53 (4.07) 6.07 (3.10) 9.93 (5.93) Median 4 6 6 9 Range 0–11 0–17 2–13 2–23 Percent of friends messaged Mean 2.54 (1.83) 1.86 (1.02) 1.28 (0.66) 1.15 (0.65) Median 2.07 1.86 1.31 1.09 Range 0–5.26 0–4.06 0.44–2.76 0.20–2.15 Time spent on Facebook (min.) Mean 162.70 (178.68) 188.20 (134.64) 239.53 (294.67) 192.28 (145.79) Median 104.32 156.33 109.46 122.5 Range 2.20–712.72 17.78–555.63 2.20–968.83 32.40–482.12 n 17 15 14 15
Table A2: Ordinary least squares regression predicting percentage of network messaged with quadratic term.
Notes: Asian is the omitted race category. Standard errors in parentheses. Betas are unstandardized. *p<0.01.
Unstandardized Betas Network size -.004 (.003) Network size squared .00 (.00) Study duration (days) .47* (.22) First-year student -.46 (.34) Female -.34 (.32) Ethnicity African American -.1.12 (.91) Hispanic/ Latino(a) -.39 (.44) White -..34 (.48) Other .46 (.68) Constant .22 (1.78) n 61 R2 .37 F 2.07
Received 12 December 2017; revised 9 February 2018; accepted 10 February 2018.
Copyright © 2018, Joanna C. Yau, Stephanie M. Reich, Yiran Wang, Melissa Niiya, and Gloria Mark. All Rights Reserved.
More friends, more interactions? The association between network size and interactions on Facebook
by Joanna C. Yau, Stephanie M. Reich, Yiran Wang, Melissa Niiya, and Gloria Mark.
First Monday, Volume 23, Number 5 - 7 May 2018