A Social Skill Account of Problematic Internet Use

Transcription

1 Social Skill and Problematic Internet Use A Social Skill Account of Problematic Internet Use By Scott E. Caplan This study integrates research on social skill and self-presentation into the recently introduced cognitive-behavioral theory of generalized problematic Internet use. The model proposed and tested here predicted that individuals who lack selfpresentational skill are especially likely to prefer online social interaction over faceto-face communication. Further, the model predicted that a preference for online social interaction fosters compulsive Internet use, which results in negative outcomes. Participants in this study completed measures of social skill, preference for online social interaction, compulsive Internet use, and negative outcomes associated with Internet use. Structural equation modeling analysis techniques tested the proposed model. The analysis indicated a good fit between the hypothesized model and the current data. Research on interpersonal communication in face-to-face (FtF) settings is not only relevant but especially useful for advancing our understanding of problematic Internet use (PIU). PIU is a multidimensional syndrome consisting of cognitive and behavioral symptoms that result in negative social, academic, or professional consequences (Caplan 2002, 2003; Davis, 2001; Davis, Flett, & Besser 2002; Morahan- Martin & Schumacher, 2003). One particularly noteworthy finding to emerge from the growing literature on PIU is that individuals who report negative outcomes associated with their Internet use are especially drawn to the interpersonal uses of the Internet (Caplan, 2002, 2003; McKenna & Bargh, 2000; Morahan-Martin & Schumacher, 2000; Young, 1998; Young & Rogers, 1998). For example, Morahan-Martin and Schumacher (2000) found that problematic Internet users were more likely than nonproblematic users to use the Internet for meeting new people, seeking emotional support, and playing socially interactive games. In another study, Young (1998) observed that, whereas nondependent Internet users spent most of their time online using and the Web, dependent users spent most of their time online using synchronous interpersonal communication applications (i.e., chatrooms and interactive multiplayer games). In one Scott E. Caplan (PhD, Purdue University) is an assistant professor in the Department of Communication at the University of Delaware. The author thanks R. Lance Holbert for his comments and suggestions on this article. Copyright 2005 International Communication Association 721

2 Journal of Communication, December 2005 review, Wallace (1999) reported that synchronous [computer-mediated communication] spaces are not the only compelling Internet environments, but they do seem to be chief culprits in excessive Internet use (p. 182). Although such findings raise interesting questions for researchers in a variety of disciplines, understanding the observed relationship between online social interaction and PIU should be of particular interest to communication researchers. Surprisingly, however, communication researchers have devoted little attention to the subject. A search of the PsychINFO database 1 revealed that, between 1996 and 2004, researchers published approximately 67 journal articles dealing with PIU, but only four (6%) were published in major communication journals (i.e., Caplan, 2003; Kubey, Lavin, & Barrows, 2001; LaRose & Eastin, 2002; LaRose, Lin, & Eastin, 2003). In response, the current article sought to demonstrate that communication researchers stand to make a valuable contribution to the emerging PIU literature. The study reported below integrates research on FtF communication with the recently introduced cognitive-behavioral theory of generalized PIU 2 (articles by Caplan, 2002, 2003; Davis, 2001; Davis et al., 2002) to propose a social skill account PIU. The model presented here hypothesizes that a social skill deficit, along with exposure to the Internet, predisposes an individual to develop a preference for online, rather than for FtF, social interaction, which then leads to compulsive Internet use, resulting in negative outcomes (see Figure 1). The following sections elaborate each of the above claims in further detail and report results from a study that tested the model. Cognitive-Behavioral Model of PIU Davis introduced a cognitive-behavioral model of PIU, contending that psychosocial problems (e.g., loneliness, depression) predispose some Internet users to exhibit maladaptive cognitions and behaviors that result in negative outcomes (Davis, 2001; Davis et al., 2002; Caplan, 2002, 2003). Results from a number of studies support the claim that psychosocial problems such as loneliness and depression are associated with PIU (Amichai-Hamburger & Ben-Artzi, 2003; Caplan, 2002, 2003; LaRose et al., 2003; Kubey et al., 2001; Morahan-Martin, 1999; Morahan-Martin & Schumacher, 2000, 2003; Shapira, Goldsmith, Keck, Khosla, & McElroy, 2000). According to Davis (2001), such psychosocial problems increase the likelihood that an individual will experience cognitive and behavioral symptoms of generalized PIU that, in turn, will result in negative outcomes. One example of a maladaptive cognitive symptom of PIU is experiencing more positive feelings about oneself when online compared to when offline (Davis, 1 The PsychINFO search looked for records of journal articles containing any of the following terms: Internet abuse, Internet addiction, problematic Internet use, pathological Internet use, Internet dependence, addictive Internet use, excessive Internet use, and compulsive Internet use. 2 Davis (2001) differentiates between specific and generalized PIU. Whereas specific PIU is cast as one of the many possible manifestations of a broader behavioral disorder, generalized PIU is a pathology associated with the unique social context avaiable on the Internet. The scope of the current study is limited to generalized PIU because of its relevance to interpersonal communication research. 722

3 Social Skill and Problematic Internet Use 2001; Caplan, 2003). With regard to maladaptive behavior, Davis (2001) proposed that PIU involves compulsive Internet use to the point that the individual suffers negative outcomes at work, school, or in personal relationships. The following sections elaborate on, and make hypotheses about, how a social skill deficit increases one s vulnerability for developing symptoms of PIU. Social Skill Deficits and Vulnerability to Cognitive Symptoms of PIU Caplan (2003) proposed that one important cognitive symptom of PIU is a preference for online, over FtF, social interaction. According to Caplan (2003) preference for online social interaction is a cognitive individual-difference construct characterized by beliefs that one is safer, more efficacious, more confident, and more comfortable with online interpersonal interactions and relationships than with traditional FtF social activities (p. 629). In a recent study, Caplan (2003) found that, consistent with the cognitive-behavioral model, preference for online social interaction (POSI) mediated the relationship between psychosocial problems (i.e., loneliness) and negative outcomes associated with Internet use among a sample of college students. A closer look at the research on social skill in FtF contexts helps explain how a lack of skill might predispose an individual to develop a preference for online social interaction. An abundance of research has indicated that social skill deficits in FtF contexts are associated with the same psychosocial problems that Davis (2001) argued precipitate the development of PIU, including depression (Segrin, 1998; Segrin & Flora, 2000; Wierzbiki, 1984; Wierzbiki & McCabe, 1988), loneliness (Jones, Hobbs, & Hockenbery, 1982; Spitzberg & Canary, 1985; Spitzberg & Hurt, 1989), and social anxiety (Leary & Kowalski, 1995; Segrin, 1996; Segrin & Kinney, 1995). To explain the relationship between well-being and social skill, scholars have proposed the social-skills-deficit vulnerability hypothesis, which argues that an individual s psychosocial well-being is threatened by a lack of interpersonal competence (Lewinsohn, Mischel, Chaplin, & Barton, 1980; Segrin, 1990, 1992, 1993, 1996; Segrin & Flora, 2000; Youngren & Lewinsohn, 1980). Segrin (1990) has maintained that a social skill deficit enhances the likelihood of eliciting punishment from the social environment (p. 293). Examples of such punishments include interpersonal rejection, embarrassment, and relationship failure. In general, social skill reflects the ability to interact with other people in a way that is both appropriate and effective (Segrin & Givertz, 2003, p. 136; also see Bedell & Lennox, 1997; Spitzberg, 2003; Spitzberg & Cupach, 1989; McFall, 1982). Riggio (1989; Riggio & Zimmerman, 1991) identified six different dimensions of social skill 3 : emotional expressivity, emotional sensitivity, emotional control, so- 3 For a thorough discussion of each of the six dimensions of social skills, see Riggio (1989) and Riggio and Zimmerman (1991). Emotional expressivity skill refers to one s ability to send nonverbal messages that express emotion, attitude, and conversational cues. Emotional sensitivity skill refers to competence with interpreting others nonverbal expressions. Emotional control skill involves one s ability to convey emotions on cue or to hide unwanted emotions. Social expressivity skill pertains to verbal message production ability. Social sensitivity skill refers to one s verbal message reception and interpretation skills. Social control skill, which is most relevant for the current study, involves one s selfpresentational competence. 723

4 Journal of Communication, December 2005 cial expressivity, social sensitivity, and social control. The current study focuses specifically on social-control skill, an individual s competence at self-presentation, role taking, and impression management in FtF interpersonal interactions. In most interpersonal encounters, people engage in a wide range of strategic self-presentation behaviors to make positive impressions of themselves on others (Goffman, 1959). Riggio (1989) explained that persons whose self-presentational skills are well developed are generally adept, tactful, and self-confident in social situations and can fit in comfortably in just about any type of social situation (p. 3). However, when individuals lack confidence in their self-presentational skills, they experience social anxiety (Leary, 1983; Leary & Kowalski, 1995; Segrin & Kinney, 1995; Schlenker & Leary, 1985). For example, in one study, Segrin and Kinney (1995) found that, although socially anxious individuals behavior did not differ significantly from that of nonanxious individuals, socially anxious participants negatively misperceived their own social skills. In order to increase their perceived self-presentational ability and to decrease social risk, people with social-control skill deficits are likely to seek out communicative channels (such as CMC) that minimize potential costs and enhance their limited abilities. O Sullivan (2000) maintained that CMC represents a useful tool for managing self-relevant information in pursuit of self-presentational goals (p. 404). In one study, O Sullivan examined preference for different interpersonal communication channels (e.g., CMC, FtF, telephone) and found that people s preferences varied depending on the extent to which they appraised a self-presentational situation as threatening or supportive. Specifically, O Sullivan s participants preferred mediated interpersonal channels when their self-presentation was threatened. As the following discussion explains, individuals with a social-control skill deficit should be more likely than others to be attracted to the unique features of online social interaction. The model presented here (Figure 1) hypothesizes that individuals who perceive themselves as having low self-presentational skill should be more motivated than their counterparts to prefer online social interaction because they perceive their self-presentational skill in online social interaction to be greater than in FtF interaction (see McKenna & Bargh, 1999). FtF interaction and synchronous computer-mediated communication differ from one another in ways that should be especially appealing to those who perceive themselves to have low self-presentational skills (for detailed reviews of these differences, see Hancock & Dunham, 2001; Ramirez, Walther, Burgoon, & Sunnafrank, 2002; Walther & Parks, 2002). Compared to FtF conversation, synchronous online social interaction often entails greater anonymity (see Bargh, McKenna, & Fitzsimmons, 2002; McKenna & Bargh, 1999, 2000; McKenna, Greene, & Gleason, 2002), greater control over self-presentation and impression formation (Bargh et al., 2002; Cornwell & Lundgren, 2001; Noonan, 1998; O Sullivan, 2002; Walther, 1996), and less perceived social risk (Morahan-Martin & Schumacher, 2000; Walther, 1996). In general, the literature suggests that CMC interactants are more selective and strategic in their selfpresentation, form more idealized impressions of their partners, and, consequently, engage in more intimate exchanges than people in FtF situations (Lea & Spears, 1992; Reicher, Spears, & Postmes, 1995; Spears & Lea, 1992, 1994; Tidwell & 724

5 Social Skill and Problematic Internet Use Walther, 2002; Walther, 1993, 1996; Walther & Burgoon, 1992). As Davis et al. (2002) argued, for some individuals, the Internet becomes a buffer for threatening social interactions (p. 332). Results of a study by Morahan-Martin & Schumacher (2000) also support these assertions. In a study of PIU and loneliness, the researchers reported that [problematic Internet] users gained social confidence online. They are friendlier, more open, and more themselves, and they report it easier to make friends when online. They have more fun with people online than [nonproblematic] users and are more likely to share intimate secrets online as well. (p. 26). The next section explains the hypothesized influence of preference for online social interaction on maladaptive behavioral symptoms of PIU (see Caplan, 2002, 2003; Davis, 2001). Compulsive Internet Use In addition to clarifying the associations among social skill, POSI, and PIU, a second goal of the current study was to test the hypothesis that compulsive Internet use mediates the influence of POSI on negative outcomes. When individuals with a self-presentational skill deficit develop a preference for online social interaction, they may devote increasingly more resources (e.g., time, money, and attention) to their online social lives; their careers and FtF relationships may quickly become secondary priorities. For example, Davis et al. (2002) argued that for socially reticent individuals, online social interaction is a social liberating experience [that] might motivate increased dependence on the Internet as a means of communicating with others (p. 332). Davis et al. (2002) go on to explain that the process occurs as individuals confine themselves to virtual communications and reduce time spent engaging in face-to-face interactions (p. 332). In other words, when individuals primary use of the Internet is synchronous online social interaction, it becomes harder for them to control their Internet use and devote sufficient effort to their FtF relationships. Compulsive and excessive Internet use are closely related, but conceptually distinct, behavioral patterns. Excessive Internet use refers to a quantity, or degree, of online activity that exceeds what a person thinks of as a normal, usual, or planned amount of time online, whereas compulsive use involves difficulty with impulse control (Caplan, 2002). Spending an excessive amount of time online is not necessarily indicative of a problem many functional Internet behaviors require excessive time online. For example, a student may spend what many would consider to be an excessive amount of time online, but such effort may be necessary in order for the student to successfully complete an assignment. Compulsive, rather than excessive, Internet use is more likely to result in negative outcomes. Indeed, a number of researchers have maintained that PIU primarily involves compulsive Internet use (for a review, see Beard & Wolf, 2001; see also Caplan, 2003; Davis, 2001; Griffiths, 2000; Shapira et al., 2000; Young, 1998; Young & Rogers, 1998). In one study, Caplan (2003) compared the extent to which excessive Internet use and compulsive Internet use predicted negative out- 725

6 Journal of Communication, December 2005 comes associated with PIU. Caplan (2003) found that both excessive and compulsive Internet use were significant predictors of negative outcomes associated with Internet use but also reported that excessive use was one of the weakest predictors of negative outcomes, whereas preference for online social interaction, compulsive use, and withdrawal [another cognitive predictor] were among the strongest (pp ). In another study, Shapira et al. (2000) interviewed people suffering from PIU (defined as engaging in Internet use that was uncontrollable, distressing, time-consuming, or resulted in negative outcomes) and found all (100%) subjects problematic Internet use met the DSM-IV (American Psychiatric Association, 1994) criteria for an impulse control disorder (p. 267). Shapira et al. (2000) concluded that, based on the current limited empirical evidence, problematic Internet use may best be classified as an impulse control disorder (p. 207). The model presented above seeks to explain why PIU is associated with interpersonal Internet use. The model predicts that individuals who perceive themselves to lack self-presentational skill are especially likely to perceive online social interaction more favorably than they perceive FtF communication. Over time, the model predicts, a preference for online social interaction leads to compulsive Internet use that results in negative outcomes. Focus of Current Study The conceptual model proposed and tested in the current study hypothesized that a self-presentational skill deficit increases the likelihood that an individual will develop a preference for online social interaction (see Figure 1). The model further predicted that POSI increases the likelihood that an individual will engage in compulsive Internet use that results in negative outcomes. Specifically, the current study tested the following direct effect hypotheses: H1: One s perceived level of self-presentational skill has a direct negative influence on POSI. H2: POSI has a direct positive influence on level of compulsive Internet use. H3: Level of compulsive Internet use has a direct positive influence on negative outcomes associated with Internet use. The current study also tested the following indirect effect hypotheses: H4: POSI has an indirect positive influence on negative outcomes that is mediated by compulsive Internet use. H5: One s perceived level of self-presentational skill has an indirect negative effect on compulsive Internet use that is mediated by POSI. Methods Participants Participants were 251 undergraduate students with a variety of different majors who were enrolled in an introductory communication class (70% female) ranging in age from 18 to 32 years old (M = 19.8; Median =19; SD = 1.42 years). To test the 726

7 Social Skill and Problematic Internet Use model presented above, the current study required a homogenous sample of individuals who frequently use the Internet. As Calder, Phillips, and Tybout (1981) explained, homogenous samples decrease the chance of making false conclusions about whether there is covariation between the variables under study (p. 200). Participants were students at a university where Internet use is extremely high, almost every student (approximately 98%) owns his or her own computer, and all residence hall rooms have direct Internet connections. Variables and Measures Self-presentational social skill. The current study employed the social control subscale of the widely used Social Skill Inventory (SSI; Riggio, 1989) to operationalize self-perceived skill at interpersonal impression management. The social control subscale of the SSI consists of three sets of 15 self-report items assessing one s skill in role playing and social self-presentation. 4 Examples of items from the social control subscale include, When in a group of people, I have trouble thinking of the right things to talk about, I can easily adjust to being in just about any social situation, and I often find myself in awkward social situations (see Riggio, 1989, for wording of all items on the SSI). Participants rated their agreement with each item on a scale ranging from 1 (strongly disagree) to 5 (strongly agree). The means of each of the three sets of social-control skill items served as indicators of the latent construct self-presentational skill in an SEM analysis. As with previous estimates of the subscale s interitem reliability (e.g., Riggio, 1989), Cronbach s alpha coefficient in the current study was high (α =.87). Preference for online social interaction. The items employed as indicators of POSI were based on a measure reported by Caplan (2003). Specifically, participants rated their agreement on a scale ranging from 1 (strongly disagree) to 5 (strongly agree) with the following statements: I prefer communicating with other people online rather than face-to-face, I feel like I have more control over conversations online than I do in face-to-face conversations, and Meeting and talking with people is better when done online rather than in face-to-face situations. In the current study, reliability among the scale items was α =.76. Each item served as an indicator of the latent POSI in the SEM analysis reported below. Compulsive Internet use. Compulsive Internet use was measured with the compulsive use subscale of the Generalized Problematic Internet Use Scale (GPIUS; Caplan, 2002). The compulsive use subscale asks participants to rate the extent to which they agree with the following statements: I have made unsuccessful attempts to control my Internet use, I am unable to reduce the amount of time I spend online, and I have tried to stop using the Internet for long periods of time. Participants rated their agreement with each item on a scale ranging from 1 (strongly disagree) to 5 (strongly agree). Each item served as 4 Riggio (1999) explained that, due to the length of the SSI, the items for each of the [sub]scales were grouped into sets ( parcels of five items each), resulting in a total of 18 item sets... [and that]... care was taken to ensure that the sets were heterogeneous regarding the content of items within a particular set (p. 15). 727

8 Journal of Communication, December 2005 an indicator of the latent compulsive Internet use variable in the SEM analyses. The reliability of the compulsive Internet use subscale for the current study was (α =.87). Negative outcomes of Internet use. Negative outcomes were measured with a subscale of items from the GPIUS (Caplan, 2002). The negative outcomes subscale asks participants to rate the extent to which they agreed or disagreed with the following items: I have missed class or work because of being online and I have gotten in trouble at work or school because of my Internet use. Participants rated their agreement with each item on a scale ranging from 1 (strongly disagree) to 5 (strongly agree). Each item served as an indicator of the latent negative outcomes variable in the SEM analyses. In the current study, the reliability coefficient for the negative outcomes subscale items was α =.76. Analyses The descriptive statistics for each observed variable, along with results of a correlation analysis of these variables, appear in Tables 1 and 2. An SEM analysis simul-taneously tested all five hypotheses presented above. SEM analysis can involve both observed variables and latent variables and three types of relationships between variables: direct effects, indirect effects, and total effects (Holbert & Stephenson, 2002; Hoyle, 1995). Indirect effects involve the influence of an independent variable on a dependent variable that is mediated by some third variable (Baron & Kelly, 1986; Hoyle, 1995). The current study assessed direct effects and decomposed total effects in order to determine indirect effects. The SEM analysis reported below was computed with AMOS 5.0 (Arbuckle, 2003). Results The SEM analysis reported here employed a full-information maximum likelihood estimation procedure. Figure 1 illustrates the specific structural model representing hypotheses that were tested in the current study, along with the standardized estimated path coefficients obtained with the current study s SEM analysis. Several indices suggest the model fit the data well, CFI =.95 ECVI =.58; SRMR =.08; BIC = ; RMSEA =.07 (90% C.I: ); p close to fit =.03. For purposes of model comparison, χ 2 = 94.81, df = 41, p <.001. Direct Effect Hypotheses The data supported H1; participants self-presentational skill was a significant negative predictor of POSI, β = -.40, p < The results also supported H2; participants level of POSI was a positive predictor of their level of compulsive Internet use, β =.48, p <.001. Additionally, the data supported H3; participants level of compulsive Internet use was a positive predictor of negative outcomes, 5 All β coefficients reported in the current article are standardized. 728

9 Social Skill and Problematic Internet Use Table 1. Descriptive Statistics and Item Wording Variable/Item wording Mean SD Social-Control Skill Subscale Social-Control Skill Subscale Social-Control Skill Subscale POSI1 I prefer communicating with other people online rather than face-to-face POSI2 I feel like I have more control over conversations online than I do in face-to-face conversations POSI3 Meeting and talking with people is better when done online rather than in face-to-face situations Compulsive use 1 I have made unsuccessful attempts to control my Internet use Compulsive use 2 I am unable to reduce the amount of time I spend online Compulsive use 3 I have tried to stop using the Internet for long periods of time Negative outcome1 I have missed classes or work because of online activities Negative outcome 2 I have missed social engagements because of online activities β =.40, p <.001. Overall, all the results were consistent with the direct effect hypotheses presented in Figure 1. Indirect Effect Hypotheses H4 and H5 were both tested with a distribution of product coefficients (P) test developed by MacKinnon, Lockwood, and Hoffman (1998). Monte Carlo analyses of different indirect effects tests indicate that the distribution of product coefficients test has greater statistical power and does a better job of maintaining an accurate Type I error rate than other available procedures (MacKinnon, Lockwood, Hoffman, West, & Sheets, 1998; also see Holbert & Stephenson, 2003). The current data supported the hypothesized negative indirect effect of social skill on compulsive Internet use, mediated by POSI (H4), P = , p <.001. The data also supported the hypothesized positive indirect effect of POSI on negative outcomes, mediated by compulsive use (H5), P = 9.23, p <.001. Finally, the predictors in the model (i.e., self-presentational skill, POSI, and compulsive use) together accounted for 10% of the variance in negative outcome scores. 729

10 Journal of Communication, December 2005 Social control skill -0.40*** Preference for online social interaction 0.44*** Compulsive Internet use 0.32*** Negative outcomes ***p <.001 Figure 1. Hypothesized model and standardized path estimates obtained in current study. Discussion As noted earlier, one particularly noteworthy finding to emerge from the growing literature on PIU is that individuals who report negative outcomes associated with their Internet use are especially drawn to the interpersonal uses of the Internet. The purpose of the current study was to employ theory and research on communication in FtF contexts to help understand why PIU is related to interpersonal Internet use. Toward this end, the current study introduced and tested a social skill account of PIU. In general, the model predicted that individuals who perceive themselves as lacking self-presentational skill would be especially likely to perceive online social interaction more favorably than FtF communication. Moreover, the model predicted that a preference for online social interaction leads to compulsive Internet use that results in negative outcomes. Overall, the results of the SEM analysis indicated that the current data did fit the conceptual model presented in Figure 1. The following paragraphs address the specific findings associated with each of the five hypotheses advanced earlier. The first hypothesis predicted that one s self-presentational skill would have a negative direct influence on POSI. The data reported above support this hypothesis, self-presentational social skill was a strong negative predictor of POSI. A second goal of the study was to test the hypothesis that compulsive Internet use stems from a heightened preference for online social interaction. Specifically, H2 predicted that one s preference for online social interaction would have a direct positive influence on level of compulsive Internet use. The results reported above support this hypothesis: Preference for online social interaction was a significant positive predictor of compulsive use. 730

11 Social Skill and Problematic Internet Use Table 2. Zero-order Pearson Correlation Matrix for Observed Variables (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (1) SSISC1 1 (2) SSISC2.71** 1 (3) SSISC3.70**.68** 1 (4) POSI1 -.25** -.21** -.28** 1 (5) POSI2 -.19** -.23** -.28**.52** 1 (6) POSI ** -.25** -.28**.54*.59** 1 (7) Compulsive use **.17**.26**.26** 1 (8) Compulsive use ** -.20**.18**.32*.24**.72** 1 (9) Compulsive use ** -.19** -.26**.23**.25**.32**.51**.46** 1 (10) Neg. outcome *.26**.25*.34**.18**.21**.26** 1 (11) Neg. outcome *.34**.19**.32**.18**.18**.22**.68** SSISC = Social-Control Skill Subscale; *p < 05. **p <.01. Next, H3 predicted that one s level of compulsive Internet use would have a direct positive influence on the likelihood one would experience negative outcomes. The data supported H3: Compulsive use was a significant positive predictor of negative outcomes. Thus, the current data supported all three direct effect hypotheses. Finally, H4 and H5 both predicted relationships with indirect effects. The data supported H4, which proposed that POSI has an indirect positive influence on negative outcomes that is mediated by compulsive Internet use. Additionally, the results supported H5, which predicted that one s perceived level of self-presentational skill has an indirect negative effect on compulsive Internet use that is mediated by POSI. In sum, the results reported above support the processes specified in the cognitive-behavioral model of PIU, in which a deficit of self-presentational social skill leads to a preference for less risky social interaction online. On the face of it, it seems that participants who prefer online social interaction might consider CMC a functional alternative to their FtF interactions. However, the results presented above suggest otherwise; participants preference for online social interaction predicted the degree to which they reported an inability to control impulses to use CMC. The following sections identify limitations of the current study and offer recommendations for future research. Although encouraging, the results reported above require some qualification. The study of computer-mediated interpersonal communication is quite new. As such, the current study represents an initial step toward developing and testing a new theory of online social interaction and PIU. The model tested here represents a nascent theory, and the study s goal was to support the most basic aspects of the theory. Thus, limitations of the current study are worth addressing in order to recommend directions for future research. 731

12 Journal of Communication, December 2005 First, given that the current study employed a cross-sectional rather than longitudinal design, it is impossible to determine causality with the current data. However, SEM does allow researchers to determine whether or not data support hypothesized causal associations. In the current study, the SEM results did offer substantial support to the hypothesized model. Second, as with most early tests of nascent theories, the current study employed a student sample. Although future studies should begin to specify the theoretical boundaries of the model presented above, results from a student sample are useful for initial hypothesis testing and theory building, especially when examining multivariate relationships 6 (Calder et al., 1981; Basil, Brown, & Bocarnea, 2002; Shapiro, 2002). A third limitation of the current study was that the model accounted for only 10% of the variance in negative outcome of Internet use scores. Researchers need to begin to identify other exogenous variables that increase the explanatory power of the model proposed here. For example, the current study employed one cognitive and one behavioral predictor of negative outcomes. Future studies may benefit from identifying multiple cognitive and behavioral predictors. Additionally, future research should begin to investigate the relative influence of both generalized PIU and specific PIU symptoms (Davis, 2001) on negative outcomes. Although the current study was focused on generalized PIU because of its association with interpersonal Internet use, one question for researchers to address is how generalized and stimuli-specific Internet use processes are similar or different from one another. In addition to its relevance to research on PIU, the current study also raises several interesting questions for interpersonal communication researchers. For example, do people perceive themselves as having different levels of social skill online than they have in FtF interactions? In other words, do people have an online social skill that differs from their FtF social skill? Are online social interactions and relationships as fulfilling and as functional for satisfying one s interpersonal needs as traditional FtF relationships are? Finally, are there functional or positive effects of preference for online social interaction? Research designed to address the above questions will help advance the growing literature on the association between Internet use, psychosocial well-being, and interpersonal communication. 6 Shapiro (2002) argued that rejecting a study that seeks to expand theory and that detects a potentially important effect on the basis of a nonrandom sample usually reflects a misunderstanding of the nature of generalizability. If a study detected an important effect, no matter what kind of sample is used, it is clearly true for some group of people, in some setting, at some time, for some message. The next step may be to conduct theoretically driven boundary search to determine to whom the effect applies and to whom it does not (p. 499). Moreover, Calder et al. (1981) explained that when making theoretical generalizations, a representative sample is not required because statistical generalization of the findings is not the goal. It is the theory that is applied beyond the research setting. The research sample need only allow a test of the theory. And, any sample within the theory s domain (e.g., any relevant sample), not just a representative one, can provide such a test (p. 200). 732

13 Social Skill and Problematic Internet Use References American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders (4th ed.; DSM-IV). Washington DC: American Psychiatric Association. Amichai-Hamburger, Y., & Ben-Artzi, E. (2003). Loneliness and Internet use. Computers in Human Behavior, 19, Arbuckle, J. L. (2003). AMOS for Windows. Analysis of moment structures (Version 5.0). Chicago: Small Waters. Bargh, J. A., McKenna, K. Y. A., & Fitzsimons, G. M. (2002). Can you see the real me? Activation and expression of the true self on the Internet. Journal of Social Issues, 58, Basil, M. D., Brown, W. J., & Bocarnea, M. C. (2002). Differences in univariate values versus multivariate relationships: Findings from a study of Diana, Princess of Wales. Human Communication Research, 28, Beard, K. W., & Wolf, E. M. (2001). Modification in the proposed diagnostic criteria for Internet addiction. CyberPsychology & Behavior, 4, Bedell, J. R., & Lennox, S. S. (1997). Handbook for communication and problem-solving skills training: A cognitive-behavioral approach. Oxford, UK: Wiley. Calder, B. J., Phillips, L. W., & Tybout, A. M. (1981). Designing research for application. Journal of Consumer Research, 8, Caplan, S. E. (2002). Problematic Internet use and psychosocial well-being: Development of a theorybased cognitive-behavioral measurement instrument. Computers in Human Behavior, 18, Caplan, S. E. (2003). Preference for online social interaction: A theory of problematic Internet use and psychosocial well-being. Communication Research, 30, Cornwell, B., & Lundgren, D. C. (2001). Love on the Internet: Involvement and misrepresentation in romantic relationships in cyberspace vs. realspace. Computers in Human Behavior, 17(2), Davis, R. A. (2001). A cognitive-behavioral model of pathological Internet use. Computers in Human Behavior, 17(2), Davis, R. A., Flett, G. L., & Besser, A. (2002). Validation of a new scale for measuring problematic Internet use: Implications for pre-employment screening. CyberPsychology & Behavior, 5, Goffman, E. (1959). The presentation of self in everyday life. New York: Doubleday Griffiths, M. (2000). Does Internet and computer addiction exist? Some case study evidence. CyberPsychology and Behavior, 3, Hancock, J. T., & Dunham, P. J. (2001). Impression formation in computer-mediated communication revisited: An analysis of the breadth and intensity of impressions. Communication Research, 28, Holbert, R. L., & Stephenson, M. T. (2002). Structural equation modeling in the communication sciences, Human Communication Research, 28, Holbert, R. L., & Stephenson, M. T. (2003). The importance of indirect effects in media effects research: Testing for mediation in structural equation modeling. Journal of Broadcasting & Electronic Media, 47, Jones, W. H., Hobbs, S. A., & Hockenbury, D. (1982). Loneliness and social skill deficits. Journal of Personality & Social Psychology, 42, Kubey, R. W., Lavin, M. J., & Barrows, J. R. (2001). Internet use and collegiate academic performance decrements: Early findings. Journal of Communication, 51,

14 Journal of Communication, December 2005 LaRose, R., & Eastin, M. S. (2002). Is online buying out of control? Electronic commerce and consumer self-regulation. Journal of Broadcasting & Electronic Media, 46, LaRose, R., Lin, C. A., & Eastin, M. S. (2003). Unregulated Internet usage: Addiction, habit, or deficient self-regulation? Media Psychology, 5, Lea, M., & Spears, R. (1992). Paralanguage and social perception in computer-mediated communication. Journal of Organizational Computing, 2, Leary, M. L. (1983). Understanding social anxiety: Social, personality, and clinical perspective. Newbury Park, CA: Sage. Leary, M. R., & Kowalski, R. M., (1995). Social anxiety. New York: Guilford Press. Lewinsohn, P. M., Mischel, W., Chaplin, W., & Barton, R. (1980). Social competence and depression: The role of illusory self-perceptions. Journal of Abnormal Psychology, 89, MacKinnon, D. P., Lockwood, C., & Hoffman, J. (1998, June). A new method to test for mediation. Paper presented at the annual meeting of the Society for Prevention Research, Park City, UT. MacKinnon, D. P., Lockwood, C. M., Hoffman, J. M., West, S. G., & Sheets, V. (2002). A comparison of methods to test mediation and other intervening variable effects. Psychological Methods, 7, McFall, R. M. (1982). A review and reformulation of the concept of social skills. Behavioral Assessment, 4, McKenna, K. Y. A., & Bargh, J. A. (1999). Causes and consequences of social interaction on the Internet: A conceptual framework. Media Psychology, 1, McKenna, K. Y. A., & Bargh, J. A. (2000). Plan 9 from cyberspace: The implications of the Internet for personality and social psychology. Journal of Personality & Social Psychology, 75, McKenna, K. Y. A., Green, A. S., & Gleason, M. E. J. (2002). Relationship formation on the Internet: What s the big attraction? Journal of Social Issues, 58, Morahan-Martin, J. (1999). The relationship between loneliness and Internet use and abuse. CyberPsychology and Behavior, 2, Morahan-Martin, J., & Schumacher, P. (2000). Incidence and correlates of pathological Internet use among college students. Computers in Human Behavior, 16, Morahan-Martin, J. & Schumacher, P. (2003). Loneliness and social uses of the Internet. Computers in Human Behavior, 19, Noonan, R. J. (1998). The psychology of sex: A mirror from the Internet. In J. Gackenbach (Ed.), Psychology and the Internet: Intrapersonal, interpersonal, and transpersonal implications (pp ). San Diego, CA: Academic Press. O Sullivan, P. B. (2000). What you don t know won t hurt me: Impression management functions of communication channels in relationships. Human Communication Research, 26, Ramirez, J. A., Walther, J. B., Burgoon, J. K., & Sunnafrank, M. (2002). Information-seeking strategies, uncertainty, and computer-mediated communication toward a conceptual model. Human Communication Research, 28, Riggio, R. (1989). The social skills inventory manual: Research edition. Palo Alto, CA: Consulting Psychologists Press. Riggio, R., Throckmorton, B., & DePaola, S. (1990). Social skills and self-esteem. Personality & Individual Differences, 11, Riggio, R., & Zimmerman, J. (1991). Social skills and interpersonal relationships: Influences on social support and support seeking. In W. H. Jones & D. Perlman (Eds.), Advances in personal relationships (Vol. 2, pp ). London: Jessica Kingsley Press. 734

15 Social Skill and Problematic Internet Use Schlenker, B. R., & Leary, M. R. (1985). Social anxiety and communication about the self. Journal of Language & Social Psychology, 45, Segrin, C. (1990). A meta-analytic review of social skill deficits in depression. Communication Monographs, 57, Segrin, C. (1993). Social skills deficits and psychosocial problems: Antecedent, concomitant, or consequent? Journal of Social & Clinical Psychology, 12, Segrin, C. (1996). The relationship between social skills deficits and psychosocial problems: A test of a vulnerability model. Communication Research, 23, Segrin, C. (1998). Interpersonal communication problems associated with depression and loneliness. In P. A. Andersen, & L. K. Guerrero (Eds.), Handbook of communication and emotion: Research, theory, applications, and contexts (pp ). San Diego, CA: Academic Press. Segrin, C., & Flora, J. (2000). Poor social skills are a vulnerability factor in the development of psychosocial problems. Human Communication Research, 26, Segrin, C., & Givertz, M. (2003). Methods of social skills training and development. In J. O. Greene & B. R. Burleson (Eds.), Handbook of communication and social interaction skills (pp ). Mahwah, NJ: Erlbaum. Segrin, C., & Kinney, T. (1995). Social skills deficits among the socially anxious: Rejection from others and loneliness. Motivation & Emotion, 19, Shapira, N. A., Goldsmith, T. D., Keck, P. E., Jr., Khosla, U. M., & McElroy, S. L. (2000). Psychiatric features of individuals with problematic Internet use. Journal of Affective Disorders, 57, Shapiro, M. A. (2002). Generalizability in communication research. Human Communication Research, 28, Spears, R., & Lea, M. (1992). Social influence and the influence of the social in computer-mediated communication. In M. Lea (Ed.), Contexts of computer-mediated communication: Contexts of computer-mediated communication (pp ). Hertfordshire, UK: Harvester Wheatsheaf. Spears, R., & Lea, M. (1994). Panacea or panopticon? The hidden power in computer-mediated communication. Communication Research, 21, Spitzberg, B. H. (2003). Methods of interpersonal skill assessment. In J. O. Greene & B. R. Burleson (Eds.), Handbook of communication and social interaction skills (pp ). Mahwah, NJ: Erlbaum. Spitzberg, B. H., & Canary, D. J. (1985). Loneliness and relationally competent communication. Journal of Social & Personal Relationships, 2, Spitzberg, B. H., & Cupach, W. R. (1989). Handbook of interpersonal competence research. New York: Springer-Verlag. Spitzberg, B. H., & Hurt, T. (1989). The relationship of interpersonal competence and skills to reported loneliness across time. In M. Hojat & R. Crandall (Eds.), Loneliness: Theory, research, and applications (pp ). Newbury Park, CA: Sage. Tidwell, L. C., & Walther, J. B. (2002). Computer-mediated communication effects on disclosure, impressions, and interpersonal evaluations: Getting to know one another a bit at a time. Human Communication Research, 28, Turkle, S. (1995). Life on the screen. New York: Simon & Schuster. Wallace, P. M. (1999). The psychology of the Internet. New York: Cambridge University Press. Walther, J. B. (1993). Impression development in computer-mediated interaction. Western Journal of Communication, 57, Walther, J. B. (1996). Computer-mediated communication: Impersonal, interpersonal, and hyperpersonal interaction. Communication Research, 23,

16 Journal of Communication, December 2005 Walther, J. B., & Burgoon, J. K. (1992). Relational communication in computer-mediated interaction. Human Communication Research, 19, Walther, J. B., & Parks, M. R. (2002). Cues filtered out, cues filtered in: Computer-mediated communication and relationships. In M. L. Knapp & J. A. Daly (Eds.), Handbook of interpersonal communication (3rd ed., pp ). Thousand Oaks, CA: Sage. Wierzbicki, M. (1984). Social skills deficits and subsequent depressed mood in students. Personality & Social Psychology Bulletin, 10, Wierzbicki, M., & McCabe, M. (1988). Social skills and subsequent depressive symptomatology in children. Journal of Clinical Child Psychology, 17, Young, K. S. (1998). Internet addiction: The emergence of a new clinical disorder. CyberPsychology & Behavior, 1, Young, K. S., & Rogers, R. C. (1998). The relationship between depression and Internet addiction. CyberPsychology & Behavior, 1, Youngren, M. A., & Lewinsohn, P. M. (1980). The functional relation between depression and problematic interpersonal behavior. Journal of Abnormal Psychology, 89,