It's all About Extraversion: Why Facebook Friend Count doesn't Count ...

Journal of Research in Personality 53 (2014) 64–67

Contents lists available at ScienceDirect

Journal of Research in Personality journal homepage: www.elsevier.com/locate/jrp

Brief Report

It’s all about Extraversion: Why Facebook friend count doesn’t count towards well-being Jan-Erik Lönnqvist a,⇑, Juha V.A. Itkonen b a b

Swedish School of Social Science, University of Helsinki, Finland Department of Political and Economic Studies, University of Helsinki, Finland

a r t i c l e

i n f o

Article history: Available online 6 September 2014 Keywords: Well-being Life Satisfaction Happiness Personality Extraversion Social networks Social relationships Facebook

a b s t r a c t Social relationships are assumed to be among the most important sources of well-being. However, previous studies showing that objectively measured social network size contributes to well-being have not included personality traits. We investigated, in a sample of Facebook users (N = 4701), whether the effects of Extraversion on well-being are mediated by a larger network. Although network size was initially associated with heightened well-being, this association was rendered insignificant when Extraversion was controlled for. Extraversion had a direct positive effect on ratings of Life Satisfaction and Happiness, but none of this effect was mediated by Facebook network size. The previously reported on associations between social network size and well-being could, at least in some contexts, be an artifact of Extraversion. Ó 2014 Elsevier Inc. All rights reserved.

1. Introduction Social relationships are considered one of the most important sources of well-being (Argyle, 1999; Myers & Diener, 1995). Psychological research on this topic has primarily focused on perceived social support, but social networks research has shown that also structural characteristics of the network, such as number of social relationships – the focus of the present research – contribute to well-being (Zhu, Woo, Porter, & Brzezinski, 2013). However, both lines of research have been hampered by methodological issues related to the use of self-reports. More specifically, estimates of network size have typically relied on name-generation techniques, which are highly susceptible to methodological artifacts (Marsden, 2005; Paik & Sanchagrin, 2013). However, online social network sites (SNSs) have recently allowed researchers to circumvent this problem by providing objective estimates of network size. Studies conducted on Facebook, the most popular SNS, with over 1.3 billion monthly users as of March 2014 (Key Facts – Facebook Newsroom), and site of the present research, have confirmed that social network size contributes to well-being (Kim & Lee, 2011). However, previous research employing objective estimates of network size has overlooked the potential role of personality, known to be an important determinant of both online

⇑ Corresponding author. Address: Swedish School of Social Science, University of Helsinki, PO Box 16, 00014 Helsinki, Finland. E-mail address: jan-erik.lonnqvist@helsinki.fi (J.-E. Lönnqvist). http://dx.doi.org/10.1016/j.jrp.2014.08.009 0092-6566/Ó 2014 Elsevier Inc. All rights reserved.

(Stopfer, Egloff, Nestler, & Back, 2013) and offline network size (Asendorpf & Wilpers, 1998), as well as of well-being (Weiss, Bates, & Luciano, 2008). The Five-Factor Model (FFM; Costa & McCrae, 1992) is currently the most popular framework of personality structure. According to this model, personality traits are biologically based and highly stable psychological tendencies with intrinsic paths of development. High Emotional Stability and high Extraversion are the two FFM traits that most consistently contribute to heightened well-being. Studies relying on self-reports of social network size have shown that Extraversion contributes to well-being by leading to desired social relationships (Argyle, 1999; Zhu et al., 2013). We initially sought to replicate this result using an objective measure of network size; i.e., number of Facebook friends. However, the results of the mediation analyses that we conducted did not conform to our expectations. Instead, they prompted to us the question of whether Facebook network size in any way contributes to greater well-being when Extraversion is controlled for.

2. Material and methods 2.1. Participants and procedure Data on social networks was collected using a Facebook application programmed for purposes of the present research project. Potential participants were directed to our survey web-page, which straightforwardly explained the purpose of the research

J.-E. Lönnqvist, J.V.A. Itkonen / Journal of Research in Personality 53 (2014) 64–67

(e.g., the description of the research began with the sentence ‘‘This is a new type of scientific survey that uses a Facebook app to link friendship data to questions on social relations, values, views, and personality’’) and allowed Facebook users to install our application after authentication. Before the application could be installed, the user had to explicitly agree (by accepting both the terms of use and the privacy and the data protection information) to share specified personal information (i.e., a full list of their Facebook friends and some background or control variables, such as gender and hometown) with the researchers. As an incentive to participate, everyone who completed the questionnaire, which took on average 16 min, was given feedback on his or her personal values profile. As additional incentives, consecutive raffles for an iPad and for a 500€ Amazon gift card were added at later stages of data collection. Participants received one raffle ticket for participating in the survey, and an additional ticket for each friend that participated. The application had three functions: (i) it retrieved data on the Facebook user’s contact network and some background or control variables, (ii) it administered a questionnaire survey including measures of personality, attitudes, and personal values, and (iii) it allowed the user to share the application in his or her contact network, either by posting the link to the application along with comments on the user’s own Facebook page (for a contact to see), or by sending a private invitation to selected Facebook friends. Participants were primarily recruited via invitations sent to e-mail lists. Several invitations to the University of Helsinki (Finland) and Aalto University (Finland) student and staff email lists were sent between June 2012 and May 2014. The number of participants that responded to the Finnish language questionnaire was 4701 (66% women; mean age 32.6 (SD = 11.4); nationality Finnish; 43.9% full-time and 14.0% part-time employed, 26.6% students; the questionnaire was also available in Swedish, English, and German, but extremely few responded in these languages, and this data was in no way analyzed). Sample size was determined by the time interval in which the study was performed – we started as soon as soon as the application was operational and included all responses that had been gathered at the time the analyses for the present paper were run.

65

between the personality traits, as well as all other variables that we report on, are provided in the Supplementary online material. 2.2.2. Well-being Well-being comprises both an emotional aspect of affect balance – referring to the level of positive and negative emotions – and a cognitive aspect of satisfaction with life, conceptualized as a sense of satisfaction with one’s life (Diener, 1984). The more cognitive aspects of well-being were measured with the ten-point Life Satisfaction scale from the European Social Survey: ‘All things considered, how satisfied are you with your life as a whole these days? 1 means you are ’’completely dissatisfied’’ and 10 means you are ’’completely satisfied’’. Where would you put your satisfaction with your life as a whole?’. The mean score on the Life Satisfaction scale was 7.69 (SD = 1.43). The more emotional aspect of well-being was measured using the four-point Happiness scale from the World Values Survey: ‘Taking all things together, would you say you are: Very happy, Quite happy, Not very happy, or Not at all happy?’. The mean score on the Happiness scale was 3.08 (SD = 0.59). No other well-being or personality measures than those reported on in the present paper were administered. 2.2.3. Number of Facebook friends Participants granted our Facebook application the right to retrieve a full list of their Facebook friends. The average number of Facebook friends was 251.93 (SD = 173.60). 3. Results The mediation analyses shown in Fig. 1 revealed that Extraversion predicted both number of Facebook friends and our two measures of well-being. However, the effects of Extraversion on well-being were not mediated by number of Facebook friends. Bootstrap analyses computed using percentile bootstrap estimates from 100,000 resamples (Hayes & Scharkow, 2013) yielded, in the prediction of Life Satisfaction and Happiness with Extraversion, indirect effects of 0.00 (CI = 0.02–0.02; where CI is the 95%

2.2. Measures 2.2.1. Personality Personality was measured by having participants complete, in self-report format, the National Character Survey (NCS; Terracciano et al., 2005; for the approved Finnish translation, see Realo et al. 2009). This measure – designed to mimic the original 240 item NEO PI-R (Costa & McCrae, 1992) – consists of 30 bipolar items, of which each measures a facet of the FFM (Costa & McCrae, 1992). Cross-instrument correlations between the NCS personality factors and longer measures of the FFM personality factors tend to vary between .70 and .80 (Konstabel, Lönnqvist, Walkowitz, Konstabel, & Verkasalo, 2012). Participants were instructed to rate themselves on a five-point scale using the 30 NCS items and at the top of the questionnaire was printed ‘I am. . .’. For instance, the two poles of the Extraversion Warmth facet were ‘Friendly, warm, affectionate’ and ‘Cool, aloof, reserved’. Scores on the five factors were calculated as the means of the six relevant facets. The scales measuring Extraversion (M = 3.47, SD = 0.65, a = 0.72), Emotional Stability (M = 3.57, SD = 0.66, a = 0.75), Openness to Experience (M = 3.89, SD = 0.56, a = 0.56), Agreeableness (M = 3.70, SD = 0.57, a = 0.69), and Conscientiousness (M = 3.67, SD = 0.63, a = 0.76) all had their mean scores near the mid-point of the scales, showed reasonable standard deviations, and the alpha internal consistency reliabilities indicated nothing unusual. The intercorrelations

Fig. 1. The mediation models show the effects of Extraversion on Life Satisfaction (top panel) and on Happiness (bottom panel), with and without number of friends included in the model. Age and sex were controlled for in all models. Asterisks indicate significant paths (⁄⁄⁄p < .001).

66

J.-E. Lönnqvist, J.V.A. Itkonen / Journal of Research in Personality 53 (2014) 64–67

Table 1 Standardized regression coefficients from regression models predicting Life Satisfaction and Happiness with age, sex, number of Facebook friends, and FFM personality traits (N = 4701). Life Satisfaction

Age Sex (woman = 1, man = 0)

Happiness

Model 1

Model 2

Model 3

Model 4

Model 1

Model 2

Model 3

Model 4

0.05** (0.01) 0.09*** (0.01)

0.07*** (0.01) 0.09*** (0.01) 0.11*** (0.02)

0.04* (0.01) 0.04** (0.01) 0.00 (0.02) 0.32*** (0.01)

0.01 (0.01) 0.13*** (0.01) 0.00 (0.01) 0.11*** (0.01) 0.49*** (0.02) 0.03* (0.01) 0.03* (0.01) 0.04** (0.01)

0.03 (0.01) 0.07*** (0.01)

0.00 (0.02) 0.08*** (0.01) 0.09*** (0.02)

0.03* (0.01) 0.03* (0.01) 0.00 (0.02) 0.29*** (0.01)

0.08*** (0.01) 0.11*** (0.01) 0.00 (0.01) 0.09*** (0.02) 0.45*** (0.02) 0.00 (0.01) 0.03 (0.01) 0.01 (0.01)

Number of friends Extraversion Emotional Stability Openness to experience Agreeableness Conscientiousness R2 Adjusted R2

0.01 0.01

0.02 0.02

0.11 0.11

0.31 0.31

0.01 0.00

0.01 0.01

0.09 0.08

0.24 0.24

Note: Numbers in parentheses are standard deviations. * p < .05. ** p < .01. *** p < .001.

Confidence Interval for the unstandardized coefficient) and 0.00 (CI: 0.01–0.01), respectively. The direct effects of Extraversion in the prediction of Life Satisfaction and Happiness were 0.70 (CI: 0.64–0.77) and .26 (CI: 0.23–0.29), respectively. The indirect effects of Extraversion were thus trivial as compared to the direct effects. This finding prompted to us the question of whether number of Facebook friends in any way contributes to greater well-being when Extraversion is controlled for. We investigated this in a set of regression analyses. In our first regression models (see Models 1 and 2 in Table 1), Life Satisfaction was predicted by age (b = .07), sex (b = .09; women were more satisfied or happy in all analyses), and number of Facebook friends (b = .11), whereas Happiness was predicted by sex (b = .08) and number of Facebook friends (b = .09). Our second set of regression models included Extraversion (see Models 3 in Table 1). In these models, Life Satisfaction was predicted by age (b = .04) sex (b = .04) and Extraversion (b = .32), but not by number of Facebook friends (b = .00). Similarly, Happiness was predicted by age (b = .03), sex (b = .03) and Extraversion (b = .29), but not by number of Facebook friends (b = .00).1 These results show that number of Facebook friends contributes to neither Life Satisfaction nor Happiness when Extraversion is controlled for.2 Our final models (see Models 4 in Table 1) show that Extraversion itself retains, albeit in reduced form, predictive power on Life Satisfaction and on Happiness even when other personality traits are controlled for, and also serve to rule out potential suppressor effects – also if entered together with all of the FFM personality traits, number of Facebook

1 The results were virtually identical in regression analyses that were ran without age and gender. When entered alone, number of Facebook friends predicted both Life Satisfaction (b = .08) and Happiness (b = .09). However, when entered together with Extraversion, number of Facebook friends predicted neither Life Satisfaction (b = .02; Extraversion b = .33) nor Happiness (b = .01; Extraversion b = .29). 2 There is some content overlap between the Positive Emotions facet of Extraversion and the Happiness scale. The two poles of the Positive Emotions facet of Extraversion are ‘‘Somber, dull, sober’’ and ‘‘Happy, cheerful, joyous’’. To alleviate concerns regarding the influence of content overlap, we ran the regression analyses predicting Happiness without this facet of Extraversion, and the results remained virtually identical – i.e., when entered together with this modified Extraversion scale, number of Facebook friends did not predict Happiness (b = .01; modified Extraversion b = .30).

friends has no predictive power on Life Satisfaction (b = .00) or on Happiness (b = .00). 4. Discussion McCrae and Costa (1991) distinguished between the direct and indirect effect of personality may have on well-being. Personality traits may directly affect well-being by affecting tendencies to experience positive or negative emotions. Or the effects may be indirect, with certain traits being instrumental in creating conditions that promote happiness or unhappiness. Based on previous research, one could have expected Extraversion to have both a direct (e.g., McCrae & Costa, 1991) and an indirect effect on wellbeing. Regarding the indirect effect, Extraversion could have been expected to lead to desired social relationships, which in turn could have been expected to contribute to well-being (Argyle, 1999; Zhu et al., 2013). By contrast, our results suggest that Extraversion does not have such an indirect effect on well-being. Furthermore, the previously reported on associations between number of social relationships and well-being may merely be an artifact of Extraversion, which affects both number of relationships and well-being. One reason our results differ from those of previous research could be that we had access to an objective measure of social network size; i.e., number of Facebook friends. Studies that have suggested an indirect effect of Extraversion on well-being via social relationships have relied on participants’ self-reported network size. Generally, having the same participants provide all measures may result in artifactual covariance between the measures. For instance consistency or social desirability motives may contribute such self-report bias (for a review, see Podsakoff, MacKenzie, Lee, & Podsakoff, 2003). Subjective estimates of network size are also hampered by more specific problems, such as biases related to the interpretation of the items, recall issues, and question-order effects (for a review, see Marsden, 2005). Although our assessment of social network size did not share method bias with our measures of personality and well-being, the latter two, which were both assessed by means of self-reports, were both subject to self-report biases. In order to rule out the possible effects of common method

J.-E. Lönnqvist, J.V.A. Itkonen / Journal of Research in Personality 53 (2014) 64–67

bias, it would have been preferable to have also these constructs assessed independently (e.g. peer-ratings of personality or psychiatrist’s ratings of well-being would have been valuable additions to our data). Another reason that our results differ from those of previous research could be that we assessed number of Facebook friendships, whereas previous studies have typically focused on offline friendships. However, we believe that our results would generalize to offline friendships. Because Facebook is used to maintain and strengthen relationships that have developed offline, not to meet new people (Ellison, Steinfield, & Lampe, 2007), number of Facebook friends is likely to be a good proxy for offline social network size. More generally, offline and online worlds have become highly similar. We thus believe that the reason that our results differ from those of previous research is that we employed an objective measure of network size, not because we assessed Facebook friendships rather than offline friendships. However, we do acknowledge that we used a very specific measure of network size, and that the results would not necessarily generalize if other measures of network size were to be used. Future research should try to find alternative, objective offline measurers of network size. Such measures could include the unique number of cell phone contacts or number of unique individuals in a cell phone text archive, or number of Holiday cards received. Future research should also employ a more diverse set of network and well-being measures. Such measures could be used to investigate alternative aspects of the social network (e.g., subjective measure of support network size, a measure of the psychological aspects of social support, such as perceived social support and satisfaction with one’s social support) and well-being (e.g., well-being measures with more breadth or reliability; non-self-report measures of well-being; separate measures of positive and negative affect). Although the result that Facebook social network size was not associated with well-being once Extraversion was controlled for was unexpected, somewhat similar findings have been reported on in the broader literature on well-being and social activity. For instance, experience sampling research has shown that neither extraverts’ greater social activity nor their supposedly more positive reactions to social situations can account for their increased positive affect when compared to introverts (Lucas, Le, & Dyrenforth, 2008). Such results, obtained with a very different methodology than the one we employed, could be interpreted as supporting the general conclusion that social activity – regardless of how it is conceptualized or measured – is not a strong mediator between Extraversion and well-being. However, as noted above, more research using a wider variety of network and well-being measures is necessary before such general conclusions are warranted. Seligman (l990) contended that today’s epidemic levels of depression stem partly from impoverished social connections in increasingly individualistic Western societies. By contrast, our results suggest that number of social relationships is of little if any significance for well-being. One interpretation of this apparent contradiction could be that the quality of social relationships has

67

deteriorated. Future research should investigate the interplay between personality traits and quality of social relationships in the prediction of well-being. Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.jrp.2014.08.009. References Argyle, M. (1999). Causes and correlates of happiness. In D. Kahneman, E. Diener, & N. Schwarz (Eds.), Well-being: The foundations of hedonic psychology (pp. 353–373). New York, NY: Russell Sage Foundation. Asendorpf, J. B., & Wilpers, S. (1998). Personality effects on social relationships. Journal of Personality and Social Psychology, 74, 1531–1544. Costa, P. T., Jr., & McCrae, R. R. (1992). NEO-PI–R professional manual. Odessa, FL: Psychological Assessment Resources. Diener, E. (1984). Subjective Well-being. Psychological Bulletin, 95, 542–575. Ellison, N. B., Steinfield, C., & Lampe, C. (2007). The benefits of Facebook ‘‘friends’’: Social capital and college students’ use of online social network sites. Journal of Computer-Mediated Communication, 12, 1143–1168. Hayes, A. F., & Scharkow, M. (2013). The relative trustworthiness of inferential tests of the indirect effect in statistical mediation analysis: Does method really matter? Psychological Science, 24, 1918–1927. Key Facts – Facebook Newsroom. . Retrieved 3.06.14. Kim, J., & Lee, J. E. (2011). The Facebook paths to happiness: Effects of the number of Facebook friends and self-presentation on subjective well-being. Cyberpsychology, Behavior and Social Networking, 6, 359–364. Konstabel, K., Lönnqvist, J.-E., Walkowitz, G., Konstabel, K., & Verkasalo, M. (2012). The ‘Short Five’ (S5): Measuring personality traits using comprehensive single items. European Journal of Personality, 26, 13–29. Lucas, R. E., Le, K., & Dyrenforth, P. S. (2008). Explaining the extraversion/positive affect relation: Sociability cannot account for extraverts’ greater happiness. Journal of Personality, 76, 385–414. Marsden, P. (2005). Recent developments in network measurement. In P. J. Carrington, J. Scott, & S. Wasserman (Eds.), Models and methods in social network analysis (pp. 8–30). New York, NY: Cambridge University Press. McCrae, R. R., & Costa, P. T. Jr., (1991). Adding Liebe und Arbeit: The full five-factor model and well-being. Personality and Social Psychology Bulletin, 17, 227–232. Myers, D. G., & Diener, E. (1995). Who is happy? Psychological Science, 6, 10–19. Paik, A., & Sanchagrin, K. (2013). Social isolation in America: An artifact. American Sociological Review, 78, 339–360. Podsakoff, P. M., MacKenzie, S. B., Lee, J. Y., & Podsakoff, N. P. (2003). Common method biases in behavioral research: A critical review of the literature and recommended remedies. Journal of Applied Psychology, 88, 879–903. Realo, A., Allik, J., Verkasalo, M., Lönnqvist, J. E., Kwiatkowska, A., Kööts, L., et al. (2009). Mechanisms of the national character stereotype: How people in six neighboring countries of Russia describe themselves and the typical Russian. European Journal of Personality, 23, 229–249. Seligman, M. E. P. (1990). Learned optimism. New York, NY: Knopf. Stopfer, J. M., Egloff, B., Nestler, S., & Back, M. D. (2013). Being popular in online social networks: How agentic, communal, and creativity traits relate to judgments of status and liking. Journal of Research in Personality, 47, 592–598. Terracciano, A., Abdel-Khalek, A. M., Ádám, N., Adamovová, L., Akn, C.-K., Ahn, H.-N., et al. (2005). National character does not reflect mean personality trait levels in 49 cultures. Science, 310, 96–100. Weiss, A., Bates, T. C., & Luciano, M. (2008). Happiness is a personal(ity) thing: The genetics of personality and well-being in a representative sample. Psychological Science, 19, 205–210. Zhu, X., Woo, S. E., Porter, C., & Brzezinski, M. (2013). Pathways to happiness: From personality to social networks and perceived support. Social Networks, 35, 382–393.