Computers in Human Behavior 19 (2003) 659–671 www.elsevier.com/locate/comphumbeh

Loneliness and social uses of the Internet§ Janet Morahan-Martina,1,*, Phyllis Schumacherb,1 a

Department of Psychology, Bryant College, 1150 Douglas Pike, Smithfield, RI 02917 1284, USA Department of Mathematics, Bryant College, 1150 Douglas Pike, Smithfield, RI 02917 1284, USA

b

Abstract Loneliness has been associated with increased Internet use. Lonely individuals may be drawn online because of the increased potential for companionship, the changed social interaction patterns online, and as a way to modulate negative moods associated with loneliness. Online, social presence and intimacy levels can be controlled; users can remain invisible as they observe others’ interactions, and can control the amount and timing of their interactions. Anonymity and lack of face-to-face communication online may decrease self-consciousness and social anxiety, which could facilitate pro-social behavior and enhance online friendship formation. Support for this model was found in a survey of 277 undergraduate Internet users that was used to assess differences between lonely and not-lonely individuals in patterns of Internet use. Loneliness was assessed on the UCLA Loneliness Scale; students in the highest 20% (Lonely) were compared with all other students (Non-lonely). Lonely individuals used the Internet and e-mail more and were more likely to use the Internet for emotional support than others. Social behavior of lonely individuals consistently was enhanced online, and lonely individuals were more likely to report making online friends and heightened satisfaction with their online friends. The lonely were more likely to use the Internet to modulate negative moods, and to report that their Internet use was causing disturbances in their daily functioning. # 2003 Elsevier Ltd. All rights reserved. Keywords: Internet; Loneliness; CMC; Disturbed use of Internet; Social behaviors

§ Portions of this paper were presented at the 107th Convention of the American Psychological Association, August 1999, Chicago, IL. This study involved a re-analysis of data published earlier by the same authors: Morahan-Martin, J., and Schumacher, P. (2000). Incidence and Correlates of pathological internet use among college students. Computers and Human Behavior, 16, 13–29. * Corresponding author. Tel.: +1-401-232-6268; fax: +1-401-232-6319. E-mail address: [email protected] (J. Morahan-Martin). 1 Both authors contributed equally to this manuscript.

0747-5632/03/$ - see front matter # 2003 Elsevier Ltd. All rights reserved. doi:10.1016/S0747-5632(03)00040-2

660

J. Morahan-Martin, P. Schumacher / Computers in Human Behavior 19 (2003) 659–671

1. Introduction Although the Internet is widely used to communicate with others (Pew Internet & American Life, 2002), loneliness has long been associated with excessive use of the Internet. Many early Internet users were computer hackers, who were characterized as preferring the machine to people. Often social outcasts, they turned first to the computer and then to the Internet in lieu of people (Shotten, 1991; Turkle, 1984). ‘‘The Internet gave them a vehicle for both communication and recreation, often providing a safe and fun social life that was denied in real life’’ (Morahan-Martin, 1999, p. 431). Many turned to the Internet to escape from the pressures and discomfort of their lives. Online, they found a world where they could be accepted and where their technical skills could be used to gain prestige. As use of the Internet spread to a broader population, early chroniclers of life online, such as Rheingold (1993) and Turkle (1995), continued to draw anecdotal links between loneliness and Internet use and abuse. Quantitative studies that followed confirmed that loneliness was associated with both increased Internet use (Kraut, Patterson, Landmark, Kiesler, Mukophadhyay, & Scherlis, 1998; Lavin, Marvin, McLarney, Nola, & Scott, 1999) and compulsive use of the Internet (Loytskert & Aiello, 1997; MorahanMartin & Schumacher, 2000; Young, 1998). However, the direction of this relationship is uncertain. Two opposing hypotheses have been proposed to explain the relationship between loneliness and Internet use: excessive Internet use causes loneliness vs. lonely individuals are more likely to use the Internet excessively. 1.1. Internet use causes loneliness Those who accept the first hypothesis and view the Internet as causing loneliness argue that time online interrupts real life relationships. Internet use isolates individuals from the real world and deprives them of the sense of belonging and connection with real world contacts. Thus, loneliness can be a byproduct of excessive Internet use because users are spending time online, often investing in online relationships, which are artificial and weak, at the expense of real life relationships. Further, online communication fosters technological alienation, creating barriers between participants, even those who know each other in other contexts. The implicit assumption of those supporting this hypothesis is that online relationships are weak and superficial compared with those in real life. Support for this argument is found in the HomeNet study, which documented increases in loneliness resulting from Internet use (Kraut et al., 1998). This study provided free or reduced cost computers, training, and Internet access for two years to 93 families consisting of 169 individuals. Participants allowed their Internet use to be monitored and provided self-reports on their psychological and social characteristics before beginning Internet use as well as at 1-year intervals after they began using the Internet. Participants’ reported loneliness at the beginning of the study did not predict amount of subsequent Internet use, but greater Internet use was associated with increased levels of loneliness. The authors attribute the increases in loneliness to decreases in family communication, social activities, happiness, and the

J. Morahan-Martin, P. Schumacher / Computers in Human Behavior 19 (2003) 659–671

661

number of individuals in one’s social network, which also were associated with increased Internet use. Additional confirmation of the adverse effect of Internet use on social relationships has been found in other studies as well. A study by Stanford Institute for the Quantitative Study of Society of a representative sample of 4113 American adults found social isolation increased with Internet use. A quarter of respondents who use the Internet more than 5 hours a week believe their time online reduces their time with friends and family either in person or on the phone, and 10% say they spend less time attending social events outside the home (O’Toole, 2000). Another, smaller study of adolescents found greater Internet use associated with weaker relationships with their parents and friends, although directionality could not be determined (Sanders, Field, Diego, & Kaplan, 2000). Two limitations should be noted. First, the results of other studies contradict these findings. Two studies of representative samples of the US population found no decline among Internet users either in pro-social behavior or in communication with family, friends or professional colleagues (Robinson, Kestnbaum, Neustadtl, & Alvarez, 2000; UCLA Center for Communication Policy, 2000), and other research has found that Internet use improves existing relationships and builds new social networks (Activmedia, 1998, Katz & Aspden, 1997). Yet, even these studies have documented a negative social impact of the Internet for a small percentage of users. Second, it may be inappropriate to causally associate reduced social contact with loneliness. Research on loneliness has found that loneliness is not related to the degree of social contact and activity. Typically, loneliness is not defined in terms of an individual’s social network, but as the subjective experience associated with a perceived lack of interpersonal intimacy (Chelune, Sultan, & Williams, 1980). 1.2. Lonely drawn to the Internet The second hypothesis is that lonely individuals are more likely to be drawn to the Internet and to use the Internet excessively because of the expanded social networks provided online and the altered patterns of online communication. Lonely individuals may be drawn to some forms of interactive social activities online because of the possibilities of belonging, companionship, and communities that they provide. Social interaction is altered online in ways that may be particularly attractive to those who are lonely. For some lonely individuals, this may lead to increased Internet use, and even to the development of Internet-related problems in their lives (Morahan-Martin, 1999). What about Internet use may be appealing to those who are lonely? ‘‘Loneliness occurs when a person’s network of social relationships is smaller or less satisfying than the person desired’’ (Peplau, Russell, & Heim, 1979, p. 55). Although being lonely can be a transient state, considerable research has documented it also is a stable personality trait, i.e. loneliness is a chronic state (Weeks, Michela, Peplau, & Bragg, 1980). To some extent, arguments that excessive Internet use causes loneliness are more state dependent in their view of loneliness while arguments that lonely people are drawn to the Internet are more trait dependent in their view of

662

J. Morahan-Martin, P. Schumacher / Computers in Human Behavior 19 (2003) 659–671

loneliness. Research on chronically lonely people has documented characteristic behavioral and cognitive factors that may contribute to their loneliness and: . . .explain why feelings of loneliness tend to be stable over time. Lonely people fail to develop close friendships because they approach conversations with negative expectancies and because they lack the necessary social skills. [These, in turn, cause them to]. . .have difficulty establishing social relations that might help them develop their social skills and thereby break out of their loneliness cycle (Burger, 1997, pp. 372–373). Lonely people are more likely than the non-lonely to be socially inhibited and anxious, self-conscious (Solano & Koester, 1989), and sensitive to rejection (Russell, Peplau, & Cutrona, 1980). They have poor social skills (Vitkus & Horowitz, 1987), have difficulty making friends, initiating social activity, and participating in groups (Horowitz & de Sales French, 1979), are less likely to be intimate and self-disclose (Berg & Peplau, 1982; Williams & Solano, 1983), and have low self-esteem (Burger, 1997). Not surprisingly, loneliness is associated with depression (Weeks et al., 1980) and other pathologies (Booth, 2000). The Internet provides an ideal social environment for lonely people to interact with others. Not only does it provide a vastly expanded social network, but also it provides altered social interaction patterns online that may be particularly attractive to those who are lonely. Online anonymity, lack of physical presence and lurking allow users to control social interaction. They can choose not only with whom and when to communicate, but also have time to compose messages. Internet communication facilitates disinhibition, self-disclosure, intimacy, and enhanced selfpresentation, as well as provides an arena for the practice and development of social skills. Much has been written about the phenomena of online disinhibition, i.e. weakened social restraints online (Joinson, 1998). Online anonymity, lack of face-to-face and real world contact, and the online culture all promote disinhibition, defined by Joinson (1998) as ‘‘an apparent reduction in concerns for self-presentation and the judgement of others’’ (p. 44). Enhanced disinhibition may counter factors associated with loneliness by reducing social inhibition, anxiety, and self-consciousness, thus promoting greater levels of self-disclosure and intimacy. Self-presentation online is altered not only by disinhibition, but also by the reduction online of role constraints, social expectations, and interpersonal barriers found in face-to-face (ftf) communication. These, combined with reduced social anxiety, can allow lonely individuals to self-present more idealized versions of self as well as role-play different online personae, with individuals feeling more themselves online than off (McKenna, 1998). Altered self-presentations online in turn can lead to an altered and enhanced sense of self in real life. In this way, the Internet can be an ‘‘identity workshop’’ where lonely individuals work through issues in the safety of the online environment (Turkle, 1995). Additionally, some lonely individuals may use the Internet as an escape: to reduce stress and alleviate negative feelings that are associated with loneliness (Booth,

J. Morahan-Martin, P. Schumacher / Computers in Human Behavior 19 (2003) 659–671

663

2000). The Internet has a downside for the lonely as well. Despite the potential positive impact of Internet use for those who are lonely, they appear to be more vulnerable than others to developing problems in their lives from Internet use such as work, school or social disruption (Loytskert & Aiello, 1997; Morahan-Martin & Schumacher, 2000; Young, 1998). This study is a comparison of lonely and non-lonely individuals in their use of the Internet and how Internet use affects social interaction. It is hypothesized that, compared with others, lonely people are more likely to spend time online, use e-mail, and use the Internet for emotional support. Consistent with the model presented earlier, lonely people are predicted to prefer Internet communication to ftf communication, enjoy the anonymity of Internet communication, and lurk more than others do. Increased pro-social behavior and role playing of lonely people is anticipated, resulting in their making and enjoying online friendships and feeling more themselves online than off as compared with the non-lonely. Lonely people are predicted to be more likely than others to use the Internet to modulate negative moods that are associated with loneliness. Finally, it is anticipated that lonely people are more likely than others to report disturbances in their everyday functioning resulting from their use of the Internet.

2. Method 2.1. Participants and procedure A survey that included Internet use and behaviors as well as the UCLA Loneliness Scale (Russell, 1996) was given to 283 undergraduates in courses requiring Internet use. Of these, 277 (97.88%) had Internet experience and were included in the study; 150 (54.15%) were male and 127 (45.85%) were female. The sample included 39 (14.1%) freshmen, 52 (18.8%) sophomores, 94 (33.9%) juniors, and 92 (33.2%) seniors. The average age of the participants was 20.72 years (S.D.=2.35). Participants had used the Internet for an average of 20.08 months (S.D.=13.8) and reported average weekly use as 3.45 hours (S.D.=4.24). 2.2. Measures The questionnaire completed by the participants included sections assessing demographics, Internet experience, reasons for Internet use, Internet behaviors, and loneliness. Demographic characteristics. The demographics section of the questionnaire included questions on participants’ gender, age, and year in college. Internet experience. Participants were asked if they had used the Internet. Those who had used the Internet were asked how many months they had used the Internet as well as how many hours per week they used the Internet and how many hours per week they spent using the Internet for e-mail.

664

J. Morahan-Martin, P. Schumacher / Computers in Human Behavior 19 (2003) 659–671

Reasons for Internet use. Participants were asked to indicate how often they had used the Internet for 17 specific reasons: communicating with friends and family, required course work, recreation or relaxation, work, meeting new people, talking to others who share common interests, staying abreast of new developments in areas of interest, sharing ideas or fantasies, wasting time, finding information for own use, emotional support, gambling, net resources intended for adults only, games, Virtual Reality, and browsing. Frequency of use for each reason was rated from 1 to 5 (1=never, 2=rarely, 3=sometimes, 4=often, 5=always). Scale of Internet Behaviors. Internet behaviors were assessed by 38 Likert-type questions that explored social aspects of Internet use, feelings of competency online, and negative impact of Internet use. A four-point scale was used with 1=strongly disagree and 4=strongly agree. UCLA Loneliness Scale. Loneliness was assessed using the UCLA Loneliness Scale, version 3 (Russell, 1996). This scale includes 20 Likert-type questions on a four-point scale, with 1=strongly disagree and 4=strongly agree. The UCLA Loneliness Scale has a good reported validity and reliability with college students (Coefficient a=0.92) (Russell, 1996). The total of the responses to the 20 questions was used to determine loneliness. The mean score was 38.35 (S.D.=8.51). Students in the top 20% (n=55), scored 47 or higher and were considered lonely (L), were compared with all other students (n=217), who were considered non-lonely (NL). The mean score for the lonely was 50.69 (S.D.=3.66). For the non-lonely, the mean score was 35.22 (S.D.=6.25). There were no gender differences in loneliness.

3. Results 3.1. Internet use and experience and loneliness No difference was found between the lonely and the non-lonely in how many months they had been using the Internet (M=20.08, S.D.=13.8, t=0.17, P=0.87). However, the two groups differed significantly in the average weekly hours online (t=2.59, P=.012), with lonely users (M=5.39, S.D.=6.57) reporting greater use than non-lonely users (M=3.00, S.D.=3.30). The lonely also used e-mail significantly more hours per week (M=3.58, S.D.=5.12) than the non-lonely (M=1.93, S.D.=2.56, t= 2.56, P=0.024). 3.2. Reasons for use and loneliness A MANOVA of reasons for use by loneliness found significant overall differences [F (16, 173)=1.66, P=0.025], with significance found for six of the17 given reasons for use. Lonely users were more likely than the non-lonely users to use the Internet for the following reasons: to relax [F (1, 188)=11, P=0.001], for work [F (1,188)=7.27, P=0.008], to meet people [F (1,188)=3.7, P=0.056], for emotional support [F (1,188)=10.04, P=0.002], talking to others who share same interests [F (1,188)=6.81, P =0.01], and to waste time [F (1, 188)=10.07, P=0.002].

J. Morahan-Martin, P. Schumacher / Computers in Human Behavior 19 (2003) 659–671

665

3.3. Internet behaviors and loneliness To explore further differences in specific Internet behaviors between lonely users and others, a MANOVA was used for the 38 questions assessing Internet-related behaviors on the Scale of Internet Behaviors. The overall results were significant (F (38,199)=1.52, P=0.035). Follow-up univariate ANOVAs were significant on 25 of 38 questions. The results are presented in Tables 1–3. As can be seen in Table 1, lonely users differed from others in social aspects of Internet use, with significance found in 14 out of 19 questions. Compared with others, lonely users were more likely to prefer communicating online to face-to-face communication, to find online anonymity liberating, to like the speed of communicating online, and to have lurked online. Lonely users were more likely than non-lonely users to agree that when online: they were more themselves than in real life, they opened up more to people than in other forms of communication, they were friendlier, they had shared intimate secrets, and they had pretended to be someone else. They also agreed more than the non-lonely that: their online friends understood them better, most of their friends they knew from online, going online made it easier to make friends, they had more fun with the people they met online, and they had a network of friends made online. The lonely also are more likely to use the Internet to modulate moods and be negatively impacted by their Internet use, as demonstrated in Table 2. The lonely agreed more that they had gone online when they were down or anxious, and when they were feeling isolated, and that they felt totally absorbed when online. Internet behavior for the lonely was causing disruptions in their life. Specifically, they agreed more than the non-lonely that they: felt guilty about time spent online, had been told they spent too much time online, routinely cut short sleep to be online, had missed social engagements to be online, had missed work or school to be online, had been unable to cut back time online, found it hard to stop thinking about going online, and had tried to hide how much time they had spent online. However, as seen in Table 3, there were no differences between the lonely and others associated with competency and convenience of Internet use.

4. Discussion This study confirmed most hypotheses. Lonely individuals differed markedly from the non-lonely in how they used the Internet socially. Compared with others, they used the Internet and e-mail more and were more likely to use the Internet for emotional support, to meet new people and to interact with others with similar interests. Social behavior of the lonely consistently was enhanced online in ways consistent with the disinhibition effect. Lonely individuals were more likely to prefer online to ftf communication, to enjoy the anonymity of online communication, and to lurk. Online, they felt more themselves, opened up more, shared intimate secrets, were friendlier, and role-played more than others. Compared with others, those who were lonely felt that going online had made it easier for them to make friends, that

666

J. Morahan-Martin, P. Schumacher / Computers in Human Behavior 19 (2003) 659–671

Table 1 Means, standard deviations, and F values of social aspects of Internet use by loneliness Individual items

Group

Mean

S.D.

F

My online friends understand me better than other people.

L NL

1.58 1.20

0.07 0.03

25.48**

I am more myself online that in real life.

L NL

1.71 1.25

0.08 0.04

25.45**

I open up more to people online than in other communication modes.

L NL

2.08 1.57

0.11 0.05

18.69**

Most of my friends I know from online.

L NL

1.58 1.21

0.08 0.04

18.15**

I prefer communicating online to face-to-face communication.

L NL

1.83 1.40

0.10 0.05

15.93**

I am friendlier online than in real life.

L NL

1.88 1.44

0.10 0.05

15.39**

The anonymity of being online is liberating.

L NL

2.25 1.71

0.12 0.06

15.12**

I have shared intimate secrets online.

L NL

1.73 1.35

0.10 0.05

11.71**

I have lurked online but never entered a conversation online.

L NL

1.88 1.47

0.11 0.06

10.49**

Going online has made it easier for me to make friends.

L NL

1.79 1.47

0.09 0.05

9.21**

I have more fun with the people I know online than others.

L NL

1.50 1.27

0.08 0.04

6.80**

I have a network of friends made online.

L NL

1.79 1.48

0.11 0.05

6.51**

Sometimes I pretend I am someone I am not while online.

L NL

1.75 1.46

0.11 0.05

5.93*

I like the speed of communicating online

L NL

2.85 2.58

0.12 0.06

4.33*

I prefer telephoning to communicating online. (reverse)

L NL

2.54 2.78

0.13 0.06

3.06

Online communication lets me control when I want to communicate.

L NL

2.83 2.61

0.12 0.06

3.05

I have pretended to be somebody of the opposite sex while online.

L NL

1.33 1.23

0.08 0.04

1.33

Being online has made it easier to communicate with people I know.

L NL

2.75 2.56

0.14 0.07

1.50

I feel less connected interpersonally when I communicate online. (reverse)

L NL

2.40 2.27

0.11 0.06

1.00

Items rated on a scale of 1–4: 1=strongly disagree and 4=strongly agree. L=lonely (n=55); NL=non-lonely (n=217). * P