Computers in Human Behavior 27 (2011) 2271–2283

Contents lists available at ScienceDirect

Computers in Human Behavior journal homepage: www.elsevier.com/locate/comphumbeh

Conceptualizing personal web usage in work contexts: A preliminary framework Sunny Jung Kim ⇑, Sahara Byrne Department of Communication, Cornell University, United States

a r t i c l e

i n f o

Article history: Available online 9 August 2011 Keywords: Personal web usage in work contexts Cyberloafing Problematic internet use Non-work-related internet uses Internet addiction Internet deviant behaviors

a b s t r a c t As the internet became the primary method of task-related communication within organizations, a social phenomenon was born: internet users going online for non-work-related purposes when supposedly working. However, there is little consensus on how to conceptualize this broad range of phenomena. Not only do many conceptual terms exist in the literature without clear distinctions, but also the degree to which specific behaviors belong under each concept remains unclear. In this article, we analyze each broad concept on specific dimensions found in the literature, including formal definitions, causes, and outcomes. We then provide a typology integrating this knowledge. Based on an empirical investigation of this typology, an initial framework of personal web usage in work contexts is proposed. Ó 2011 Elsevier Ltd. All rights reserved.

1. Introduction The internet has undoubtedly become a vital tool for taskoriented communication (Davis, Flett, & Besser, 2002; Hilts, 2008). Even though individuals within organizations, such as employees and students, currently enjoy a number of advantages in the era of computer-mediated communication, internet users in such settings may also find themselves distracted from their work by the temptation to use the internet for personal reasons. Non-work-related use of the internet has raised social concerns, especially for corporate and educational institutions. By some estimations, more than half of web activities are personal (Greengard, 2000), which translates to time employees and students are not spending on their work. For example, Kandell (1998) found that because college students are developing a solid sense of identity and intimate relationships, they tend to become dependent on social networking sites and other personal internet use. Dependency on the internet may cause students to go online during class, and may distract them from processing important course materials. Other studies have found that although internet-dependent students use the internet to gain social support and to reduce psychological symptoms such as depression and low self-esteem (LaRose, Lin, & Eastin, 2003), using the internet does not enhance their psychological well-being or performance on course work. Negative outcomes such as decline in study habits, missed classes, and grade drops were reported by students engaged in excessive internet use on a daily basis (Scherer, 1997; Young, 1996). On the other hand, some scholars have argued that, when appropriately used, non-work-related internet use can have a posi⇑ Corresponding author. E-mail address: [email protected] (S.J. Kim). 0747-5632/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.chb.2011.07.006

tive influence on work productivity by providing refreshment and increasing creativity (Terr, 1999). In spite of these positive beliefs about non-work-related internet activities, these behaviors often lead to negative consequences such as productivity loss, resource wasting, increased risk of computer hacking, potential legal liability risk, online copyright infringement, and distribution of viruses (Chou, Sinha, & Zhao, 2008; Greenfield & Davis, 2002; Mills, Hu, Beldona, & Clay, 2001). Various concepts and terms have been used to describe the phenomenon, including non-work-related computing (NWRC), cyberloafing, cyberslacking, cyberbludging, on-line loafing, internet deviance, problematic internet use (PIU), personal web usage (PWU) at work, internet dependency, internet abuse, internet addiction, and internet addiction disorder (IAD). Because these diverse terms have been used in research literature without clear distinctions, it is uncertain whether scholars are describing a single phenomenon, several related phenomena, or distinct phenomena. This weak conceptual foundation will eventually lead to confusion as the field grows. Furthermore, it is not clear what specific behaviors constitute each concept. There is insufficient empirical evidence that examines how the overarching concepts and specific behaviors fit together. For example, should checking personal email be included in problematic internet use, but not internet abuse? Is visiting adult sites at work a sign of internet addiction disorder, whereas visiting Facebook is not? No study has comprehensively examined perceptions of the phenomenon by identifying to what degree specific internet activities should be included under each overarching concept. Finally, research on such behaviors does not concisely capture the variety of non-work-related internet activities. Although previous work has measured individual’s engagement in many types of internet deviant behaviors (see Blanchard & Henle, 2008), those studies were conducted only in the context of cyberloafing and

2272

S.J. Kim, S. Byrne / Computers in Human Behavior 27 (2011) 2271–2283

before the dramatic rise of social networking sites (SNS). Therefore SNS behaviors, such as interacting on Facebook, were not measured in the previous studies. Clarifying the conceptual definitions and articulating the behaviors that represent these concepts are crucial steps before further developing theoretical applications and practical implications. Accordingly, the main goals of this study are to (1) provide a typology of concepts describing internet usage for non-work-related purposes when supposedly working, (2) identify associations between each concept and specific internet activities that capture a contemporary trend of internet use, and (3) argue for an initial framework driven by our empirical investigation on how these concepts relate to one another. 1.1. Conceptualizing non-work-related internet deviant behaviors Although some of the concepts mentioned above, such as internet addiction (Beard, 2002; Block, 2001; Chou & Hsiao, 2000; Nalwa & Anand, 2003; Soule, Shell, & Kleen, 2003; Young, 1999, 2004; Young and Rogers,1998) and problematic internet use (Davis et al., 2002; Jia & Jia, 2009; Shapira et al., 2003) are discussed in the literature as having mostly negative implications and connotations, other concepts such as cyberloafing (Terr, 1999) and personal web usage (Anandarajan & Simmers, 2004a) carry both positive and negative connotations. While the reasons for these distinctions seem apparent, clearly understanding differences and similarities across these concepts is not simple. For example, non-work-related computing (NWRC) and personal web usage (PWU) are difficult to distinguish. One reason is that scholars in this field often use specific terms without explaining why they decided on one term over another existing option. The taxonomy developed by Chou et al. (2008) represents a rare example of studies making an effort to distinguish concepts from one another. The development of a clear and detailed conceptual framework can fruitfully inform future research in the field of personal web use in work contexts. In this article, the seven terms that have been systematically tested at least once through empirical research will be explicated in reference to four dimensions: definitions, causes, outcomes, and exchangeable terms (see Table 1). Analysis of the literature on these four dimensions affords an opportunity to identify differences or similarities among the descriptions of each concept. In addition to explicating definitions, as we explore causes and outcomes in overarching concepts within the literature, we seek to identify concepts that seem similar by definition. For example, the reasons why individuals engage in non-work-related internet activities vary across the overarching concepts, ranging from convenient access (Young, 1999) to coping with depression (Ceyhan and Ceyhan, 2007; Davis et al., 2002). Also, in terms of perceived outcomes, certain concepts such as internet addiction disorder (IAD) and problematic internet use (PIU) are defined with an emphasis on problematic outcomes of using the internet excessively (Anandarajan & Simmers, 2004b; Mills et al., 2001; Oravec, 2004; Terr, 1999), whereas concepts like cyberloafing acknowledge positive consequences such as refreshment (Oravec, 2004; Terr, 1999). The following section describes seven overarching concepts and proposes a conceptual typology based on a four-dimensional explication of these concepts. 1.1.1. Personal web usage (PWU) 1.1.1.1. Definition. Personal web usage (PWU) refers to ‘‘any voluntary act of using web access during work to surf non-work-related websites for non-work-related purposes’’ (Anandarajan & Simmers, 2004a, p. 2). There are numerous PWU activities reported to date, including visiting news sites, online shopping, job search, video games, video streaming, music downloads, and online communities (Fox, 2007).

1.1.1.2. Causes. Referring to Bandura’s (1986, 1991) concept of fundamental incentives that lead to behavior adoptions, researchers found that boredom increases self-reactive incentives in order to have psychological equilibrium, thus decreasing self-regulation on work. Reduced self-regulation increases PWU engagement while working, and as the behavior increases, it can become problematic (Eastin, Glynn, & Griffiths, 2007). 1.1.1.3. Outcomes. PWU was found to have both positive and negative influences on work performance. Anandarajan and Simmers (2004a) discussed PWU from three different perspectives. First, from the organization’s perspective, PWU is usually considered to lead to negative consequences such as productivity losses, increased security costs, legal liability risks, and network overload (Anandarajan & Simmer, 2004a). Second, on an individual level, PWU is perceived as a dysfunctional work behavior. Third, some researchers claim that PWU is actually a positive ‘‘cyber-activity,’’ providing ‘‘a necessary break from drudgery or intense endeavor. . . so that it might increase productivity’’ (Friedman, 2000, p. 1563). The third view advocates that PWU is a way of managing one’s personal life and this can be constructive by contributing to knowledge gains that are necessary to be productive (Anandarajan & Simmers, 2004a, 2004b). 1.1.2. Cyberloafing 1.1.2.1. Definition. Cyberloafing is any voluntary, aimless, and undirected way of using web access and engaging in non-work-related activities on a regular basis, partially due to a lack of self-control at work. The notion of cyberloafing emerged to refer to deviant work behaviors (Kamins, 1995; Manrique deLara, Tacoronte, & Ding, 2006), meaning ‘‘the behavior that departs from norms of a reference group’’ (Blanchard & Henle, 2008, p. 1071; Warren, 2003). From this perspective, certain types of cyberloafing might be considered acceptable if the behaviors do not deviate from the reference group’s norms. Or, those behaviors could be considered deviant behaviors when they violate the norms. 1.1.2.2. Causes. Other than internal causes such as a lack of selfcontrol and procrastination (Lavoie & Pychyl, 2001), when employees feel that they are not being treated well, they tend to engage in cyberloafing behaviors (Lim, Teo, & Loo, 2002; Manrique de Lara et al., 2006). Perceived organizational control and fear of formal punishment were found to decrease cyberloafing engagement (Manrique de Lara et al., 2006). 1.1.2.3. Outcomes. There have been continual debates on whether cyberloafing results in positive or negative consequences, or both. Some researchers believe cyberloafing is a ‘‘way of idling on the job’’ that leads to counterproductive and harmful consequences (Lim, 2002) such as productivity losses, lost wages due to decreased work productivity (Blanchard and Henle, 2008; Scheuermann & Langford, 1997; Stewart, 2000), waste of bandwidth (Johnson & Indvik, 2003), increased risk of computer hacking, online copyright infringement, and increased risk of computer viruses (Blanchard & Henle, 2008; Chou et al., 2008; Greenfield & Davis, 2002; Mills et al., 2001). On the other hand, some researchers argue that cyberloafing is not always inappropriate, but is rather innocuous if properly controlled (Blanchard & Henle, 2008); they acknowledge the internet can provide internet users with creativity, flexibility, and ‘camaraderie’ (Blanchard & Henle, 2008; Anandarajan & Simmers, 2004a, 2004b; Belanger & Van Slyke, 2002; Block, 2001; Greenfield & Davis, 2002; Oravec, 2004; Stanton, 2002). Blanchard and Henle (2008) claim that cyberloafing can be divided into two sorts: minor cyberloafing and serious cyberloafing. Minor cyberloafing (e. g., reading personal emails) is considered

Table 1 Typology of concepts. Cyberloafing

Non-work-related computing (NWRC)

Internet abuse

Problematic internet use (PIU)

Internet addiction

Internet addiction disorder (IAD)

Definition

‘‘Any voluntary act of using their organization’s internet access during office hours to surf non-work-related websites for non-work purposes’’ (Anandarajan & Simmers, 2004a; Lim et al., 2002)

‘‘Employees’ usage of organizational IT resources for personal purpose, goal, and aim not directly related to organizational goals’’ (Chou et al., 2008; Lee et al., 2005)

‘‘Non-work-related internet use (Hilts, 2008), milder form of internet addiction’’ (Young, 2004). This can be led by ‘‘high degree of Internet addiction’’ (Chen et al., 2007)

A very broad concept that covers internet addiction, internet addiction disorder (IAD), and pathological internet use (Davis et al., 2002; Shapira et al., 2003)

‘‘Repetitive habit pattern’’ and, ‘‘Loss of control that continues despite volitional attempts to abstain or moderate use’’ (Marlatt et al., 1988)

‘‘A broad term covering a variety of behavioral and impulse-control problems’’ (Soule et al., 2003). It can involve cybersexusal addiction and cyber-relationship addiction (Armstrong, 2001; Young, 1999, 2004)

Causes

-

‘‘Act of employees using their organizations’ internet access for personal reasons during work hours’’ and often viewed as ‘‘slacking, aimless, and dysfunctional work behavior’’ (Anandarajan & Simmers, 2004a), except personal learning cyberloafing (Blanchard & Henle, 2008) - Lack of self-control - Lack of external locus of control (i.e., lack of belief in powerful others) (Blanchard & Henle, 2008)

- Lack of self-control - Workplace subjective norms - Personality - Perceived usefulness - Unconscious personal habit - Perceived internet accessibility (Lee et al., 2005)

-

- Lack of self-control - Diminished impulse control - Loneliness - Depression - Social comfort - Distraction - Procrastination (Davis et al., 2002; Lavoie & Pychyl, 2001)

- Lack of self-control - Internet dependency - Anonymity - Convenience - Escapism - Low-self esteem - Isolation - Intense loneliness (Chen et al., 2007; Soule et al., 2003; Young, 1999)

-

Outcomes Exchangeable terms

Positive/negative - Non-work-related Internet Usage (Eastin et al., 2007)

Negative

Negative - Pathological Internet Use (Nalwa & Anand, 2003; Young, 1999, 1996)

Lack of self-contro Workplace boredom Deficient self-regulation Habit Stress relieving (Eastin, Glynn, & Griffiths, 2007)

Positive/negative - Online-loafing - Cyberslacking - Deviant workplace behavior - Cyberbludging (Blanchard & Henle, 2008; Mills et al., 2001)

Positive/negative - Web-based internet abuse - Web-abuse - Junk computing (Bock & Ho, 2009; Chou et al., 2008)

Lack of self-control Opportunity and access Affordability Anonymity Convenience Escapism Social acceptability Long working hours A tendency of believing external locus of control - Low-self esteem - Internet addiction (Chen et al., 2007; Griffiths, 2003) Negative - Non-work-related computing (NWRC) - Problematic Internet use in the workplace - PWU in the workplace

Lack of self-control Dependency on the internet Procrastination Isolation Loneliness (Soule et al., 2003)

- Internet overuse - Pathological Internet Use (Niemz, Griffiths, & Banyard, 2005)

S.J. Kim, S. Byrne / Computers in Human Behavior 27 (2011) 2271–2283

PWU

2273

2274

S.J. Kim, S. Byrne / Computers in Human Behavior 27 (2011) 2271–2283

innocuous whereas serious cyberloafing (e. g., adult websites, downloading copyrighted files) is perceived to be problematic and inappropriate. Research has found that minor cyberloafing is more prevalent than serious cyberloafing (Blanchard & Henle, 2008; Lavoie & Pychyl, 2001; Lim et al., 2002; Manrique deLara et al., 2006). However, when time and consequences are taken into consideration, minor cyberloafing can be just as harmful as serious cyberloafing. For example, if a student spends 5 h a day in minor cyberloafing activities, its can bring seriously negative consequences on his/her school performance.

claimed as reasons that increase internet abuse behaviors (Griffiths, 2003). Internal causes include escapism (Griffiths, 2003), low self-esteem (Chen, Chen, & Yang, 2007), and loss of self-control (Padilla-Walker, Nelson, Carroll, & Jensen, 2009).

1.1.3. Non-work-related computing (NWRC) 1.1.3.1. Definition. NWRC indicates the use of the organization’s internet services or resources for personal aims that are not directly related to the organization’s goals (Chou et al., 2008; Wong, Lee, & Lim, 2005). Different from cyberloafing, which sees the behaviors as slacking and aimless, NWRC has been used to describe behaviors that are more focused on personal goals and aims.

1.1.5. Problematic internet use (PIU) 1.1.5.1. Definition. Problematic internet use (PIU) is a broad negative concept that describes ‘‘an individual’s inability to control his or her internet use’’ (Shapira et al., 2003, p. 208), behavioral addiction that focuses on particular online activity (Davis et al., 2002), dysfunctional use of the internet (Shapira et al., 2003), psychological dependency (Bradley, 1990), and impulsive control disorder (Shapira et al., 2003; Young, 1999; Young & Rogers, 1998).

1.1.3.2. Causes. Lack of self-control, unconscious personal habit, perceived internet service accessibility, environmental conditions and individual behavioral styles have all been claimed as causes for NWRC (Bock & Ho, 2009). However, comparatively insufficient empirical evidence made it difficult to further investigate this concept. 1.1.3.3. Outcomes. Similar to PWU or cyberloafing, NWRC is discussed with both positive and negative sides. Some researchers argue that NWRC allows more freedom for internet usage, and this can lead users to feel happy and gratified, and consequently more productive (Chou et al., 2008). On the contrary, others focus on negative consequences of NWRC on individuals, such as productivity loss (Bock & Ho, 2009), and on organizations, such as security concerns, reduced bandwidth, and legal issues (Case & Young, 2002). 1.1.4. Internet abuse 1.1.4.1. Definition. Internet abuse refers to non-work-related and unauthorized use of internet access for individual pleasure (Lee, Lim, & Wong, 2005). Regarding the tone of this concept, some researchers argue that specific and self-explanatory terms such as ‘excessive internet use’ could be more appropriate than socially negative terms like internet abuse (Beard, 2002; Bryant & Zillmann, 2002). Other researchers argue that internet abuse is a mild form of internet addiction (Griffiths, 2003; Young, 2004). Griffiths (2003) adopted six subtypes of internet addiction to develop a typology of internet abuse. These subtypes include cybersexual internet abuse (e.g., wasting time on online pornographic sites), internet activity abuse (e.g., gaming, online gambling, online stocks, e-auction, travel booking), online friendship/ relationship abuse (e.g., creating deceptive personas), online information abuse (e.g., wasting time for searching irrelevant information), criminal internet abuse (e.g., online sexual harassment), and miscellaneous internet abuse (e.g., activity not included in the previous subtypes such as creating fake celebrity images). Other than these six subtypes of internet abuse behaviors, it is noteworthy that minor cyberloafing behaviors, such as sending non-work-related emails, have been identified as internet abuse behaviors (Case & Young, 2002; Chou et al., 2008; Greenfield & Davis, 2002; Lim et al., 2002). This may imply that these activities can be considered internet abuse when performed with malicious intentions. 1.1.4.2. Causes. External causes such as affordability (e.g., free access to the internet), anonymity, convenience (i.e., easy access to the internet), long working hours, and social acceptability are

1.1.4.3. Outcomes. Negative outcomes of internet abuse include wasted time (Malachowski, 2005), productivity loss (Sharma & Gupta, 2003), security risks, network congestion, and legal liabilities in relation to sexual harassment and online crimes (Chou et al., 2008, p. 420; Panko & Beh, 2002).

1.1.5.2. Causes. PIU is mainly concerned with internal causes such as unstable emotional and psychological states. Researchers report diminished impulse control (Davis et al., 2002), depression, loneliness (Davis et al., 2002), a tendency to procrastinate (Davis et al., 2002), and social rejection (Lavoie & Pychyl, 2001) as primary reasons that increase PIU engagement. For problematic internet users, the internet is used to avoid stressful events or heavy tasks, to avoid feeling pressure, to gain social comfort. 1.1.5.3. Outcomes. PIU deals with negative personal (e.g., productivity loss), psychological (e.g., loneliness), and social consequences (e.g., job turnover) (Brenner, 1997; Davis et al., 2002; Kraut et al., 1998; Young, 1996). Although some people use the internet to engage in social activity and widen their social sphere, PIU researchers argue that it ironically decreases time spent on face-to-face interactions. Consequently, problematic internet users begin to feel even lonelier due to a lack of human face-to-face interaction (Davis et al., 2002). 1.1.6. Internet addiction 1.1.6.1. Definition. According to Mosby’s Medical, Nursing & Allied Health Dictionary, addiction refers to ‘‘compulsive and uncontrollable dependence on a substance, habit, or practice to such a degree that cessation causes severe emotional, mental, or physiologic reactions.’’ Marlatt, Baer, Doncan, and Kivlahan (1988) define media addiction as ‘‘a repetitive habitual pattern that increases the risk of disease and/or associated personal and social problems’’ (p. 224). Young first introduced the notion of internet addiction in 1996 at the Annual Meeting of the American Psychological Association. The concept often involves ‘‘loss of control despite volitional attempts to abstain or moderate use,’’ and is considered a type of ‘‘behavioral addiction,’’ or ‘‘process addiction’’ (Chou & Hsiao, 2000; Soule et al., 2003). 1.1.6.2. Causes. Based on Griffiths’s six criteria for behavioral addiction, researchers developed criteria to examine internet addiction among students (Young, 2004). The criteria include salience (i.e., using the internet becomes the most important activity in my life), mood modification (i.e., I use the internet to escape from the real world and to be aroused), tolerance (i.e., I spend increasing amounts of time online to achieve desired effects), withdrawal symptoms (i.e., I feel irritable or unpleasant when I am offline), conflict (i.e., my internet use causes conflicts with friends and family), and relapse (i.e., I tried to discontinue or reduce my internet use, but it turned out to be impossible). These criteria reflect

S.J. Kim, S. Byrne / Computers in Human Behavior 27 (2011) 2271–2283

relevant causes of internet addiction although some of them are closer to symptoms than causes. 1.1.6.3. Outcomes. Only negative consequences at an individual level have been reported. The outcomes are missing deadlines, losing interest in hobbies, losing sleep, preferring to be online, lacking interpersonal skills, and job loss due to consistently reduced work productivity (Fox, 2007; Scherer, 1997; Young, 2004). 1.1.7. Internet addiction disorder (IAD) 1.1.7.1. Definition. Internet addiction disorder (IAD) refers to compulsive internet use. There has been a debate on whether or not IAD should be included as a diagnosis in the Diagnostic and Statistical Manual of Mental Disorders (DSM-V). If IAD is included in the 5th edition of DSM, expected to be published in 2012, insurance will pay for internet addiction counseling. Young (1999, 2004) and Armstrong (2001) claim that IAD entails several subtypes (Soule et al., 2003, p. 65) such as cybersexual addiction (e.g., addiction to online porn or adult chat rooms), cyber-relationship addiction (e.g., online-based friendship established via online chatting), net compulsion (e.g., compulsive trading, auction, and gambling), information overload (e.g., compulsive Web surfing or database surfing), and computer addiction (e.g., game playing or programming in a compulsive manner). These five types of IAD are very similar to the five sub-types of internet abuse developed by Griffiths (2003). The difference is that internet abuse is also concerned with potential crime-related problems, while internet addiction disorder (IAD) does not include those issues. That might be because IAD has been developed within media psychology to cure individuals suffering from compulsive internet usage. 1.1.7.2. Causes. Block (2008) claims that people with IAD suffer from emotional problems such as depression. IAD users tend to use the virtual cyber-world to escape unpleasant moods or psychologically stressful situations. Procrastination, isolation, and loneliness are found to be significant causes for IAD that lead to compulsive internet use (Soule et al., 2003). 1.1.7.3. Outcomes. There is as yet insufficient empirical evidence that explores behavioral consequences resulting from IAD. 1.2. Research questions on the overarching concepts Table 1 presents a summary of the literature review in the form of a typology. The table allows us to compare across the four comparative dimensions of definitions, causes, outcomes and exchangeable terms. While helpful, when reading across columns, it is still unclear exactly how the concepts relate to one another. There are conceptual overlaps, and distinctions need to be made. Therefore, to guide an empirical investigation of the beliefs and attitudes associated with each overarching concept, we generated a set of research questions designed to provide insight into these distinctions. RQ1: In general, is each concept associated with more positive attitudes or more negative attitudes? RQ2: Which concepts are associated with positive consequences, such as enhanced productivity, creativity, being entertained, and attaining social comfort? RQ3: Which concepts are associated with negative consequences, such as productivity loss, low self-esteem, depression, and reduced time for social activity? RQ4: Do individuals perceive that the concepts have distinct causal roots? If so, are these causes internal (e. g., need for escape or as a solution to boredom) or external (e. g., easy access to the internet or distance from the professor or boss)?

2275

1.3. Internet deviant behaviors (IDBs) The typology and research questions focus on understanding the overarching concepts that describe personal use of the internet at work. Next, we unravel the specific behaviors that should, or should not, be included under each of those concepts. Scholars have argued that using the internet for non-work-related purposes while working in an unregulated manner is a type of ‘‘deviant work behavior’’ (Manrique deLara, Tacoronte, & Ding, 2006). Deviant work behaviors refer to any activities that depart from the expected behaviors of a reference group (Blanchard & Henle, 2008, p. 1071; Warren, 2003). Applying this definition to the subject matter at hand, we define internet deviant behaviors (IDB) as specific internet uses that violate organizational norms. Previous research on IDBs has mainly focused on cyberloafing (Blanchard & Henle, 2008). In this work, behaviors have been placed within two categories, minor cyberloafing (e.g., checking non-work emails) and serious cyberloafing (e.g., online gambling). However, this categorization has not been applied to any other overarching concepts. Also, the work in this area occurred before the rise of popular social media activities (e.g., Facebook and YouTube), thus it does not include social-mediarelated behaviors. Understanding the different forms of internet deviant behaviors (IDBs) within the seven concepts is important for organizations, policy makers, and program designers to develop appropriate internet policies, effective preventative campaigns, and preventative computer programs. That being said, our next primary goals are, first, to understand the degree to which IDBs are included within each of the overarching concepts and, second, to propose an inclusion of social media activities in the measurement of such behaviors. 1.3.1. Minor and serious IDBs Robinson and Bennett (1995) proposed a typology of deviant work behaviors in order to identify conceptual distinctions between serious deviant work behaviors and minor deviant work behaviors. Based on Robinson and Bennett’s typology, Blanchard and Henle (2008) adopted the categorizations and applied them to cyberloafing, calling the new categories minor cyberloafing and serious cyberloafing (see Appendix A). We use these categorizations and broaden the terminology in order to test affiliations between behaviors and each overarching concept; we call the categories minor internet deviant behaviors (MIDB) and serious internet deviant behaviors (SIDB) (see Appendix A). These two IDB categories are distinct from each other for several reasons. First, perceived outcomes and attitudes are qualitatively different (Blanchard & Henle, 2008). Minor IDBs (MIDB) include behaviors that are somewhat common and perceived as relatively innocuous (e.g., sending personal email, visiting mainstream news and financial sites), whereas serious IDBs (SIDB) entail behaviors perceived as potentially harmful, inappropriate, or illegal (e.g., online gambling and illegal file downloading). Due to the benign nature of MIDB, it is likely that individuals will associate the behaviors in the MIDB category with relatively positive overarching concepts. For example, checking news reports may be categorized under non-work-related computing (NWRC), but is less likely to be perceived as problematic internet use (PIU) behavior. Furthermore, depending on one’s group norms (Blanchard & Henle, 2008) or social responsibility and status, there may be variations in whether MIDB should be considered socially acceptable, whereas, in most cases, individuals agree that SIDB in work contexts are inappropriate and not normative. Second, the two categories are quantitatively different on occurrence rates. Researchers have found that internet users

2276

S.J. Kim, S. Byrne / Computers in Human Behavior 27 (2011) 2271–2283

within organizations engaged in Serious IDBs much less than Minor IDBs (Fox, 2007). According to internet addiction researchers, viewing pornography websites at work showed the lowest occurrence rates. However, over 85% of study participants reported that they often check non-work-related emails at work (Lim, 2002; Lim & Teo, 2005; Lim et al., 2002).

2001; Lim, 2002). Qualified participants were asked to send a request email to a researcher of the study. All participants were then randomly assigned to one of the seven conditions and received an email containing a link to the stimulus and a questionnaire corresponding to his or her assigned condition. 2.3. Manipulation

1.3.2. Social network site (SNS)-related IDB One limitation of previous work developed by Blanchard and Henle (2008) is that it does not account for behaviors related to Social Network Sites (SNS). Therefore, these studies may not accurately reflect current trends of IDBs. The use of SNS has increased rapidly, especially among young adults. Forty-six percent of American adult internet users use SNS on a regular basis (PEW Research Center, 2010), 73% of adult users have a Facebook account and 48% have a MySpace profile. Despite the popularity, there has been scant empirical research that examines the degree of individuals’ involvement in SNS in work contexts. Therefore, we generate a new category for SNS-related IDBs that includes networking (e.g., Facebook), video entertainment (e.g., Youtube, Hulu), and personals/online dating. Because SNS behaviors were not included in the original Blanchard and Henle work, it is theoretically unclear if some or all behaviors on SNS would be classified as minor or serious. Therefore, we maintain a separate category for behaviors related to Social Networking sites in addition to the other two categories (i.e., minor IDBs and serious IDBs). 1.3.3. Behavioral research questions The next step is to examine how each SNS-IDB, MIDB, and SIDB fits within the overarching concept. For example, it is unclear whether using SNS in class should be considered deviant or socially acceptable. If one considers using SNS while working as problematic, SNS behaviors will be more closely associated with negative overarching concepts than relatively positive ones. In order to examine these issues, we propose to examine the following research questions associated with IDBs. RQ5: Will Minor, Serious, and SNS IDBs differ in their association with the overarching concepts? RQ6: Will SIDB be more strongly associated with negative overarching concepts than positive overarching concepts? RQ7: Will MIDB be more strongly associated with positive overarching concepts than negative overarching concepts? RQ8: Will previously unmeasured behaviors related to SNS be associated with certain overarching concepts more than others?

2. Method 2.1. Participants To investigate the research questions above, 203 young adults were recruited through a computer-based research participation system at a northeastern university. The sample consisted of 124 females and 72 males. The participants ranged in age between 18 and 31 years old with a median age of 20. Participants reported their hours of internet use per day (less than 1 h = .5%, more than 1 h less than 3 h = 26%, more than 3 h less than 6 h = 51.5%, more than 6 h less than 9 h = 19%, more than 9 h = 3%). The participants received course credit for participating. 2.2. Procedure Data were obtained through an online-based experiment. This method of data collection is appropriate when the target population is made up of frequent internet users (Cheyne & Ritter,

To measure perceptual differences on each overarching concept and IDBs, each concept (PWU, cyberloafing, NWRC, internet abuse, PIU, Internet addiction, and IAD) became a condition. Participants were first asked to read a description of their assigned overarching concept ‘X’, which was formulated based on a review of current literature. After reading the corresponding description, participants were asked to rate a set of questions measuring general attitudes and beliefs about the assigned concept, as well as perceived causes and consequences of the assigned concept. The concept corresponding to the participant’s assigned condition was inserted into the stem ‘X’. This method ensured that the question wording was consistent across the seven different conditions. 2.4. Dependent variables Participants were asked to rate, ‘‘how strongly would you agree with the following statements based on the description you read of (X)?’’ on a 7-point Likert scale. To assess the dependent variable in RQ1, participants were asked to indicate general attitudes toward their overarching concept with two items reading, ‘‘engaging in (X) at work/school is a good thing’’ (M = 3.64, SD = 1.49), and ‘‘engaging in (X) at work/school is a bad thing’’ (M = 4.19, SD = 1.56). The latter item was reverse-coded, and then the two items were averaged to create a measure of general attitude toward the concept (M = 3.73, SD = 1.42, r = .74). A higher score indicated a more positive attitude. In order to measure the dependent variable addressed in RQ2, participants were asked to report the degree to which they perceived their assigned concept (X) would lead to positive outcomes. Five items measuring positive outcomes (adapted from Table 1) were combined to create a scale, positive outcomes fit (M = 3.84, SD = 1.18, a = .83). The specific items in the scale included enhanced productivity (M = 3.02, SD = 1.52), increased creativity (M = 3.57, SD = 1.64), being entertained (M = 4.81, SD = 1.4), receiving social comfort (M = 3.98, SD = 1.51) and one additional item asking whether or not (X) is likely to lead to positive outcomes (M = 3.83, SD = 1.63). In order to examine the dependent variable in RQ3, participants were asked to report the degree to which they perceived their assigned concept (X) would lead to negative outcomes. Eight items measuring negative outcomes (adapted from studies cited across Table 1) were compressed to create a scale indicating negative outcomes fit (M = 3.84, SD = .95, a = .82). The specific items in the scale included productivity loss (M = 5.00, SD = 1.50), becoming habit (M = 5.25, SD = 1.24), becoming depressed (M = 2.67, SD = 1.48), low-self esteem (M = 2.76, SD = 1.41), feeling isolated (M = 3.22, SD = 1.51), less time for social activity (M = 3.93, SD = 1.56), feeling lonely (M = 3.14, SD = 1.52), and one general question reading, ‘‘Engaging in (X) at school/work can have a negative effect on students’’ (M = 4.76, SD = 1.37). Perceived internal causes and external causes were tested as the dependent variables in RQ4. Participants were asked to indicate the degree to which they felt the assigned concept was caused by internal causes (e.g., dependency, procrastination, escapism, and boredom) and external causes (e.g., easy access to the internet, physical distance to your professor/boss). For internal causes, 15 items adapted from Table 1 were measured. In order to create scales for internal causal factors, a principal components analysis

2277

S.J. Kim, S. Byrne / Computers in Human Behavior 27 (2011) 2271–2283

(PCA) with Varimax rotation retaining Eigen values above 1 was conducted. Three factors emerged and items were combined to form a measure of each factor. Low self-esteem, depression, and feeling lonely loaded together. We combined these items to form a factor labeled loneliness (M = 3.57, SD = 1.28, a = .84). The second factor was lack of control (M = 4.10, SD = 1.19, a = .77). This factor included internet dependency, lack of self-control, inability to selfregulate, and diminished impulse control. The third factor, labeled as escapism (M = 5.49, SD = .82, a = .72), encompassed items such as relieving stress, boredom, the need to be entertained, escape and procrastinate. Two variables that did not load on any factor were kept as single measures. One item measured to what degree going online at work is caused by the need for rejuvenation. The item read ‘‘people engage in (X) at work/school because they want to be re-energized while working/studying’’ (M = 4.53, SD = 1.49). Another item measured to what degree going online at work is caused by personality factors. The item read, ‘‘People’s personalities lead them to do more (X)’’ (M = 4.73, SD = 1.26). For external causes, which have been rarely addressed in previous research, we asked participants to respond to four items. Those included a measurement of being anonymous (M = 4.82, SD = 1.43), group norms (M = 5.6, SD = 1.19), physical distance from a boss/professor (M = 5.59, SD = 1.15), and convenient access to the internet (M = 5.72, SD = 1.15). In our analyses on RQ5 throughout RQ8, we measured the degree to which all 26 IDB items should be included within their assigned concept (X). Participants were asked to make judgments on ‘‘how strongly do you consider each of the following behaviors to fall under the concept of (X)?’’ (see Appendix A for means and standard deviations of the items and reliability of combined scales). The previously validated 22 cyberloafing behavior items (Blanchard & Henle, 2008; Lim, 2002; Robinson & Bennett, 1995) were given to participants along with four additional SNS-related IDB items. Among those 26 items, 10 items were combined to form a Minor IDB fit scale. Twelve items were combined to form a Serious IDB fit scale, and the remaining four items were combined to form a SNS-IDB fit scale. These final three fit scales are minor IDB fit (M = 4.22, SD = 1.29, a = .89), serious IDB fit (M = 5.70, SD = 1.25, a = .95), and SNS-IDB fit (M = 5.75, SD = 1.21, a = .80). Items in each fit scale achieved Cronbah’s alpha greater than .80 at a 95% significance level (see Appendix A). 3. Results Our first research question assessed participants’ attitudes toward the overarching concepts. Taking each overarching concept as an independent variable and ‘‘attitude toward the concept’’ as the dependent variable, a one-way ANOVA revealed two homogeneous subsets (see Table 2). The first subset was composed of concepts associated with more positive attitudes (i.e., PWU, NWRC, and cyberloafing). The second subset was associated with less positive attitudes toward the concepts, including PIU, internet abuse, internet addiction, IAD and cyberloafing (F(6, 193) = 5.96, p < .001, g2p = .16, Tukey). Notice that cyberloafing fell within both subsets. This result implies that cyberloafing is the broadest concept on this particular dependent variable, in that individuals can have both positive and negative attitudes toward this concept. RQ2 asked which concepts are associated with positive outcomes, such as enhanced productivity, creativity, being entertained, and attaining social comfort. A one-way Analysis of Variance (see Table 2) revealed two homogeneous subsets (F(6, 194) = 6.16, p < .001, g2p = .16, Tukey). The concepts in the first subset (PWU, NWRC, cyberloafing, and PIU) were perceived to have positive consequences such as enhancing productivity. The concepts with low scores on perceived positive outcomes were

Table 2 Attitudes and perceived outcomes of the seven overarching concepts. Dependent variables Concept (M, SD)

* ** ***

Attitudes scale* More positive

Less positive

PWU (4.27, 1.57)

Cyberloafing (3.93, 1.41)

NWRC (4.69, 1.07) Cyberloafing (4,00, 1.24)

PIU (3.48, 1.36) Internet abuse (3.21, 1.05) Internet addiction (3.11, 1.56) IAD (3.22, 1.35)

Perceived positive outcomes** More positive outcomes PWU (3.15, .67) NWRC (3.39, .79) Cyberloafing (3.58, .57) PIU (4.09, .89)

Less positive outcomes Internet abuse (3.99, .80) PIU (4.09, .89) Internet addiction (4.48, 1.01) IAD (4.23, 1.10)

Perceived negative outcomes*** Less negative outcomes PWU (3.15, .67) NWRC (3.39, .79) Cyberloafing (3.58, .57)

More negative outcomes Internet abuse (3.99, .80) PIU (4.09, .89) Internet addiction (4.48, 1.01)

ANOVA F(6, 193) = 5.96, p < .001, g2p = .16. ANOVA F(6, 194) = 6.16, p < .001, g2p = .16. ANOVA F(6, 194) = 9.18, p < .001, g2p = .22.

internet addiction, IAD, internet abuse, and PIU. These concepts were loaded on the second subset. Notice that problematic internet use (PIU) fell in both subsets. This implies that PIU can perhaps be associated with relatively positive concepts but at the same time it entails all the other negative concepts such as internet abuse, internet addiction, and IAD. In RQ3, we measured the degree to which the overarching concepts are associated with negative consequences, such as productivity loss, low self-esteem, depression, and reduced time for social activity. A one-way ANOVA on the negative outcomes fit scale (see Table 2) revealed two homogeneous subsets (ANOVA F(6, 194) = 9.18, p < .001, g2p = .22, Tukey). Similar to the analysis of attitude toward the concepts tested on RQ1, the first subset included PWU, NWRC, and cyberloafing as these concepts were perceived to result in less negative consequences than the concepts in the second subset (internet abuse, PIU, internet addiction, and IAD). In this analysis, cyberloafing appeared only within the subset of more positive concepts. The RQ4 asked participants to rate the degree to which their assigned concept is caused by internal factors (such as the need to alleviate boredom) and external factors (such as easy internet access). In order to visually compare the degree to which internal and external causes explain or ‘‘fit’’ each concept, we created two different 95% CI graphs; one containing the five internal cause ‘‘fit’’ scales (Fig. 1), and the other with four external cause ‘‘fit’’ scales (Fig. 2). Graphical demonstrations can ‘‘convey an overall pattern of results at a quick glance’’ (APA, 2001, p. 176). The use of confidence intervals is considered ‘‘the best reporting strategy’’ (APA, 2001, p. 22) because it provides information with precision and inferential statistical significance (Belia, Fidler, Williams, & Cumming, 2005). When sample sizes are similar across conditions but not small (at least 10), 95% CIs of independent means that overlap by one fourth the average length of the intervals indicate p values almost equal to or slightly less than .05. When 95% CIs just touch, p value on the two-tail test is less than .01. This principle is called Rule of Thumb for 95% CIs (Cumming & Finch, 2005; Payton, Greenstone, & Schenker, 2003). Looking first at the internal cause fit scales in Fig. 1 (loneliness, lack of self-control, escapism, rejuvenation, and personality), the CIs of the loneliness fit do not overlap with any of the escapism fit CIs.

2278

S.J. Kim, S. Byrne / Computers in Human Behavior 27 (2011) 2271–2283

Fig. 1. 95% CI of internal causes across concepts.

Fig. 2. 95% CI of external causes across concepts.

This indicates that there is significant difference between these two fit scales across all conditions (p < .01 two-way). Inspection of confidence intervals yields the conclusion that there are many pairs of CIs with significant differences in how individuals attribute

the overarching concept to each internal cause (p < .05). For example, according to Fig. 1, the perception on how strongly loneliness causes PWU was significantly different from the perception on how strongly loneliness causes IAD (p < .05). Also two general

2279

S.J. Kim, S. Byrne / Computers in Human Behavior 27 (2011) 2271–2283

trends should be noted. First, when looking at escapism as a cause, the pattern indicates that more positive concepts appear to be caused by escapism and this perception decreases as the concepts become more negative. The opposite trend can be seen for loneliness. The more negative concepts, the more people perceive the concepts to be caused by loneliness. And perceiving loneliness as an internal cause decreases as the concepts become more positive. Fig. 2 displays 95% CIs of external causes, including group norms, physical distances from the boss/professor, anonymity, and convenient access to the internet. Based on Rule of Thumb for CIs (Cumming & Finch, 2005), PWU was distinctive from other concepts on perceived group norm (p < .05). In order to be more specific, we ultimately constructed a database of perceived causes (see Table 4). RQ5 examined the association between IDBs and each overarching concept. We did this to measure how distinctively each IDB-fit scale was associated with each concept. Referring to Fig. 3, the 95% confidence intervals indicate that the inclusion of Minor IDBs in the overarching concepts varies across conditions, whereas Serious IDBs tend to be included in all overarching concepts to a relatively high degree. This is also demonstrated in Table 3. For example, Minor IDBs were more likely to be included within personal web

usage, cyberloafing, and NWRC, whereas these behaviors were not included as primary behaviors of internet abuse, PIU, internet addiction and internet addiction disorder. This conceptual difference was significant (p < .05). RQ6 assessed if Serious IDBs were more strongly associated with relatively negative overarching concepts such as internet addiction and PIU than positive overarching concepts such as PWU. It was not the case. Participants in PWU, cyberloafing, PIU, internet addiction, and IAD conditions perceived that Serious IDBs should be equally included within their assigned concepts. The only significant difference was that NWRC appeared less likely than internet addiction to include serious IDBs. Therefore, these two concepts appear to anchor the top and bottom of the continuum of overarching concepts when it comes to the inclusion of Serious IDBs. RQ7 measured if Minor IDBs were more strongly associated with positive overarching concepts and not with negative overarching concepts (see Fig. 3). However, this finding was not the case with internet abuse. Although participants hold relatively negative attitudes toward internet abuse in our other findings, internet abuse still differed from internet addiction and internet

Fig. 3. 95% CI of IDB category fits across concepts.

Table 3 Perceptions on inclusion of three types of internet deviant behaviors within overarching concepts. PWUa Minor internet deviant behaviors* Serious internet deviant behaviors** SNS internet deviant behaviors***

Cyberloafing b

Non-work-related computing (NWRC)c

Internet abused

Problematic internet use (PIU)e

Internet addiction

5.11defg (1.12)

4.65efg (1.28)

4.70efg (1.25)

4.25afg (1.24)

3.76abc (0.96)

3.52abcd (1.14)

3.48abcd (1.10)

5.75 (1.79)

5.90 (1.23)

6.02 F(0.95)

5.94 (0.86)

5.53 (1.02)

5.32c (1.35)

5.37 (1.23)

5.98 (1.45)

5.97 (1.20)

5.90 (1.07)

5.85 (1.06)

5.73 (1.10)

5.44 (1.25)

5.36 (1.31)

Note: Possible range for means was 1–7. Standard deviations are indicated in brackets after the means. Corresponding letters indicate coefficients that are significantly different than labeled condition (p < .05 one-way). * a = .89. ** a = .95. *** a = .80.

f

Internet addiction disorder (IAD)g

2280

S.J. Kim, S. Byrne / Computers in Human Behavior 27 (2011) 2271–2283

Table 4 Means and 95% confidence intervals for internal causes and external causes. Loneliness fit

Lack of control fit

95% CI Nc PWU Cyberloafing NWRC Internet abuse PIU Internet addiction IAD Overall

95% CI

Rejuvenation fit

95% CI

Personality fit

95% CI

95% CI

M

Lower

Upper

M

Lower

Upper

M

Lower

Upper

M

Lower

Upper

M

Lower

Upper

28 30 29 29 29 28 28

3.36 3.65 3.69 3.91 3.91 4.33 4.27

2.97 3.31 3.27 3.48 3.49 3.98 3.84

3.76 3.99 4.11 4.35 4.34 4.69 4.70

3.59 4.28 3.65 4.10 4.33 4.58 4.17

3.11 3.88 3.25 3.69 3.88 4.10 3.67

4.07 4.69 4.05 4.51 4.78 5.06 4.67

5.65 5.68 5.46 5.35 5.30 5.38 5.17

5.38 5.32 5.15 5.06 4.99 5.04 4.77

5.93 6.03 5.77 5.64 5.60 5.72 5.57

5.00 4.50 4.97 4.14 4.34 4.39 4.37

4.34 3.95 4.54 3.59 3.76 3.78 3.79

5.66 5.05 5.39 4.68 4.93 5.00 4.95

5.11 4.83 4.72 4.24 4.79 4.61 4.81

4.60 4.49 4.19 3.77 4.30 4.07 4.34

5.62 5.17 5.26 4.71 5.28 5.15 5.29

201

3.87

3.72

4.02

4.10

3.93

4.27

5.43

5.31

5.55

4.53

4.32

4.74

4.73

4.55

4.91

Physical distance

Group norms

95% CI

PWU Cyberloafing NWRC Internet abuse PIU Internet addiction IAD Overall

Escapism fit

Anonymity

95% CI

Convenient access

95% CI

95% CI

Nc

M

Lower

Upper

M

Lower

Upper

M

Lower

Upper

M

Lower

Upper

28 30 29 29 29 28 28 201

5.86 5.70 5.62 5.48 5.55 5.75 5.18 5.59

5.45 5.31 5.16 4.98 5.14 5.36 4.68 5.43

6.26 6.09 6.08 5.99 5.97 6.14 5.67 5.75

6.30 5.77 5.62 5.38 5.11 5.36 5.68 5.60

5.99 5.35 5.21 4.80 4.64 4.88 5.27 5.43

6.60 6.18 6.03 5.96 5.57 5.83 6.09 5.76

4.64 4.77 4.90 4.79 4.41 5.21 5.00 4.82

3.97 4.26 4.28 4.37 3.86 4.72 4.47 4.62

5.31 5.27 5.52 5.22 4.97 5.71 5.53 5.01

5.82 5.93 5.76 5.43 5.66 5.82 5.57 5.72

5.28 5.51 5.43 4.95 5.25 5.41 5.07 5.55

6.36 6.36 6.09 5.91 6.07 6.23 6.07 5.88

determined the degree to which specific behaviors should be included under each concept. In this section, we discuss a framework that is based upon our findings and previous literature. Based on the explication of concepts as well as empirical evidence, Fig. 4 provides an illustration of how the concepts relate to one another. Our framework identifies an umbrella term for the overall phenomenon and presents three branches of sub-phenomena. The double-sided arrows indicate exchangeable terms that have been discussed in previous research and were shown in our data to be virtually synonymous.

addiction disorder when participants were asked to evaluate how strongly each MIDB should fall under the assigned concept. RQ8 asked if social network site (SNS)-related IDBs were associated with certain overarching concepts more than the others. SNSrelated IDBs were perceived as belonging under all concepts to an equal degree. There was no significant difference across overarching concepts on perceiving SNS-related IDBs as strong, indicative behaviors (see Table 3 and Fig. 3). 4. Discussion

4.1.1. The ‘‘umbrella’’ concept The first step in proposing a framework is arguing for a broad term that subsumes the other concepts, a so-called umbrella concept. Based on both previous research and our empirical evidence, personal web usage (PWU) is the most appropriate umbrella concept for several reasons. First, it is a relatively objective concept. PWU is not always, under the established definition, a bad thing (Anandarajan & Simmers, 2004a; Lim et al., 2002), nor does it always refer to

4.1. A conceptual framework of personal web usage in work contexts In the previous sections, we demonstrated the applicability of our conceptual typology and provided empirical evidence to better understand the social phenomenon of using the web for non-workrelated reasons when supposedly working. We investigated the phenomenon under two broad lenses: (a) we examined conceptual dimensions, assessing attitudes toward overarching concepts, and the perceived causes and outcomes of the concepts; and (b) we

PWU

Aimless

Cyberloafing

Cyberslacking

Problematic Strategic

None-Work-Related Computing

Internet Abuse

Problematic Internet Use

Internet Addiction

Internet Addiction Disorder

Pathological Internet Use

Note: A double-headed arrow connects interchangeable concepts. Fig. 4. A conceptual framework of personal web usage in work contexts. Note: A double-headed arrow connects interchangeable concepts.

S.J. Kim, S. Byrne / Computers in Human Behavior 27 (2011) 2271–2283

distracting behaviors in an effort to cope with one’s psychological state (e.g., stress or boredom). Even though our data indicated that cyberloafing held both positive and negative attitudes (see Table 2) and PWU was perceived to be slightly more positive than negative, by definition, cyberloafing entails negative connotations such as ‘‘being slacking’’ and ‘‘aimless,’’ but PWU does not necessarily hold these attributes. Compared to concepts like internet addiction that mainly discuss negative consequences, PWU does not neglect the possibility that personal web usage while working can be harmless, or even positive and conducive. Supported by our empirical evidence on perceived positive outcomes, PWU appeared to have positive consequences (M = 4.51, SD = 1.01). Second, PWU includes various types of internet activities. Some researchers prefer to use the term ‘internet abuse’ to indicate inappropriate use of internet while working (Chou et al., 2008). Many similarities are found in previous research between internet abuse and PWU. Users’ internet behaviors reported by some studies on internet abuse are the same as those found by PWU research, thus creating potential confusion. However, our empirical evidence showed that PWU was distinct from internet abuse on Minor IDBs; MIDB were inclusive to PWU but not to internet abuse. Table 3 indicates that PWU is very similar to cyberloafing and NWRC in terms of the degree to which MIDB should be included as representing behaviors. Although not statistically significant (p > 0.05), the pattern indicates that individuals perceived PWU to be more inclusive (M = 5.11, SD = 1.12) of MIDB than both NWRC (M = 4.69, SD = 1.25), and cyberloafing (M = 4.65, SD = 1.28) (see Table 3 and Fig. 3). The final reason why PWU is a plausible umbrella concept is the existence of relevant empirical research as a conceptual foundation. Non-work-related computing (NWRC) is often used synonymously with PWU, but has been neither empirically nor systematically tested. Considering limited knowledge of NWRC, we suggest that using PWU as an umbrella construct is useful and it is a reasonable place to begin a framework of the overall phenomenon. Therefore, we built three branches of the framework under the umbrella concept, PWU (see Fig. 4). 4.1.1.1. Aimless branch. The first branch is called the Aimless Branch. Concepts in this branch may not necessarily be problematic to individuals or to organizations, but they are still considered aimless and slacking due to a student/employee’s lack of selfcontrol. We pose that cyberloafing falls under this branch. First, conceptually, previous research argues that cyberloafing is slacking, dysfunctional and aimless behaviors (Anandarajan & Simmers, 2004a; Blanchard & Henle, 2008). Second, our empirical evidence supports this argument. Cyberloafing was rated high on a lack of self-control (M = 4.28, SD = 1.08), and it was distinctive from NWRC (M = 3.65, SD = 1.06) on self-control, F(6, 194) = 2.75, p < .05. 4.1.1.2. Strategic branch. Next we propose a Strategic Branch. Related concepts describe non-work internet behaviors that can be constructive and lead to positive outcomes. Strategic internet uses are goals-driven, even if these might not be directly related to work (e.g., visiting a bank website to pay a bill). Among the seven concepts, NWRC is best suited to this branch, as it refers to nonwork-related internet behaviors for personal ‘‘aims and goals.’’ Specifically, NWRC was perceived to be distinctive from all other concepts except PWU on positive attitudes, F(6, 193) = 5.96, p < .001, positive outcomes F(6, 194) = 6.16, p < .001, a lack of control, F(6, 194) = 2.75, p < .05, and negative outcomes, F(6, 194) = 9.18, p < .001. Therefore, NWRC is placed under the Strategic Branch alone. 4.1.1.3. Problematic branch. The last branch of the framework focuses on internet use that can be problematic at an individual or

2281

a societal level. Certain concepts such as internet addiction and internet addiction disorder (IAD) emerged from addiction research. These concepts explain that internet misuse is a pathological symptom and behavioral addiction (Beard, 2002). Both existing research and our own data suggest that PIU, internet abuse, internet addiction, and IAD compose a problematic branch of PWU. Previous research states that PIU is a broad term that embraces both internet addiction and IAD (Davis et al., 2002; Shapira et al., 2003). Interestingly, ANOVA homogeneous subsets revealed that one subset includes NWRC, PWU, cyberloafing and PIU, and the other subset entails internet abuse, internet addiction, IAD, and PIU (see Table 3). It is reasonable to place PIU as a sub-category of PWU that entails internet abuse, internet addiction, and IAD.

5. Conclusions, implications, and questions for future research Internet access has become ubiquitous in corporate and academic settings. Considering the increased number of employees and students who often use computer-mediated communication technology for both work and leisure, it is important to acknowledge that the boundaries between work and leisure have become blurry. One can imagine that it is just as easy to slip into leisurely behaviors while one should be working as it is to slip into work behaviors when one is not supposed to be working. In today’s work environment, it is worthwhile to note that breaking the boundary may not always be perceived as deviant. Students and employees with flexible work time and settings may use the internet to their advantage, such as facilitating their work while simultaneously attaining leisure. However, the ambiguous boundary between leisure and work can be more of a problem when internet users carry over the flexible pattern of personal internet use into the classroom or the workplace, as this may present consequences for the instructor, the employer, or even the user. Taking this into consideration, this article sheds light on internet behaviors and contexts where non-work-related internet use is perceived to be inappropriate and relatively negative as well as contexts where it might not be so bad. On the basis of a review of existing literature, our study explicated the distinct meanings of each concept and provided a typology as well as a preliminary framework to prevent conceptual confusion in future research. We also provided empirical evidence on perceived differences across seven concepts and analyzed the fit between internet deviant behaviors and concepts in one specific population: college students. We found that people hold different attitudes toward different concepts. Some concepts such as PWU and NWRC are perceived more positively than other concepts such as internet abuse. While certain concepts such as cyberloafing acknowledge that the phenomenon can cause both positive and negative consequences, other concepts such as internet addiction mainly focus on negative perspectives of the behavior. Indeed, participants perceived internet abuse, internet addiction, PIU, and IAD to be associated with undesirable and negative consequences more than NWRC, PWU and cyberloafing. In terms of the causes of IDBs, individuals tended to perceive that the behaviors were attributable to external causes (e.g., convenient internet access) more than to internal causes (e.g., lack of self-control). Finally, we defined internet deviant behavior types and measured to what degree those behaviors should be included in each concept. Interestingly, it was the minor internet deviant behaviors that showed the most variation in perceptions of inclusiveness to their concept. Specifically, while serious IDBs were considered part of all concepts, minor IDBs were not perceived to be part of the problematic branch. However, it is important to acknowledge the limitations of our experiment. First, we measured perceptions instead of actual behaviors. Designing a field experiment or lab experiment (such

2282

S.J. Kim, S. Byrne / Computers in Human Behavior 27 (2011) 2271–2283

as using an eye-tracking technique) to observe how people become distracted from work and when they go online for non-work-related purposes could be a more accurate way of measuring their IDBs than judging their perceptions based on self-reports. Second, the perceptions about these concepts are group sensitive; students and employees might have different perceptions that yield different experimental results. For this reason, we confined our initial study to a single population to avoid indecipherably mixed results caused by diverse populations in one study. Therefore, our study only reflects the perceptions of students. Our next goal is to examine this issue among different populations, including employees who are engaged in web-based communication on a regular basis at work, and supervisors who may hold distinct perceptions about internet deviant behaviors within work contexts. Third, individuals may have web access not only on computers but also on their portable internet-enabled gadgets such as smartphones. The PWU concepts do not specify the medium that people may use for the internet activities. Although diverse communication technology platforms are constantly being released, it is unknown whether or not the scope of the concepts should be extended to these mobile technologies. Finally, our measurement was methodologically limited. Although we expanded the scope of current IDB measurements by including social-network-site (SNS)-related behaviors, there are still other SNS-related activities that can be added to the current IDB measures. In spite of the aforementioned research limitations, our conceptual framework and empirical evidence have valuable implications for practitioners and policy makers. Identifying when and why individuals are engaged in non-work related internet activities is crucial for the development of effective interventions and internet use policies. Especially, when a campaigner promotes internet selfregulation behaviors, the present framework and the typology can be useful resources to better understand the phenomenon.

Table A1 Means and SD for dependent variables. Dependent variable

Mean

SD

Minor IDB scale (a = .89) Checking non-work emails Sending non-work emails Receiving non-work emails Visiting news sites Visiting stock sites Visiting sports sites Visiting financial/banking sites Shopping online Booking a future vacation and travel Job hunting

3.81 3.94 3.31 3.55 3.59 4.86 3.67 5.88 5.10 4.51

1.90 1.87 1.94 1.82 1.76 1.79 1.73 1.37 1.87 2.01

Serious IDB scale (a = .95) Visiting/creating personal web pages Online auction (eBay) Playing games online Visiting chat-rooms Newsgroups Virtual communities (second life) Blogging about your day Downloading music Online gambling Reading blogs Visiting adult sites/ pornography Instant messaging

5.57 5.81 6.18 6.05 4.60 6.09 5.20 5.48 6.30 5.23 6.44 5.42

1.66 1.56 1.36 1.53 1.79 1.57 1.75 1.64 1.42 1.66 1.38 1.62

5.76 5.57 5.66

1.44 1.64 1.52

5.99

1.54

SNS IDB scale (a = .80) Social networking (Facebook, MySpace) Twittering Video/Music entertainment (YouTube, Hulu) Visiting personals/online dating

Without a full understanding of the phenomenon, persuasion theories and campaign models could be less powerful in constructing effective campaign messages to prevent IDBs, as well as in identifying intervention goals and target populations. As web-based communication becomes prevalent and time spent on the web increases, it is crucial to understand internet use behaviors and how individuals acquire self-regulation skills. An informational campaign or a media literacy approach may serve to disseminate useful information on internet behavior regulation at individual, organizational, or societal levels. These practical implications and questions are not a primary concern of this article, but they deserve empirical study in future research. Appendix A See Table A1. References American Psychological Association (2001). Publication manual of the American Psychological Association (5th ed.). Washington, DC:Author. Anandarajan, M., & Simmers, C. A. (2004a). A multidimensional scaling approach to personal Web usage in the workplace. In M. Anandarajan & C. A. Simmers (Eds.), Personal web usage in the workplace. a guide to effective human resource management (pp. 61–78). Hershey, PA: Idea Group. Anandarajan, M., & Simmers, C. A. (2004b). Constructive and dysfunctional personal web usage in the workplace: Mapping employee attitudes. Idea Group Inc. Armstrong, L. (2001). How to beat addiction to Cyberspace. Vibrant Life, 17(4), 14. Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory. Englewood Cliffs, NJ: Prentice Hall. Bandura, A. (1991). Social cognitive theory of self-regulation. Organizational Behavior and Human Decision Processes, 50, 248–287. Beard, K. W. (2002). Internet addiction: Current status and implications for employees. Journal of Employment Counseling, 39, 2–11. Belanger, F., & Van Slyke, C. (2002). Abuse or learning? Communications of the ACM, 45, 617–638. Belia, S., Fidler, F., Williams, J., & Cumming, G. (2005). Researchers misunderstand confidence intervals and standard error bars. Psychological Methods, 10, 389–396. Blanchard, A. L., & Henle, C. A. (2008). Correlates of different forms of cyberloafing: The role of norms and external locus of control. Computers in Human Behavior, 24, 1067–1084. Block, W. (2001). Cyberslacking, business ethics and managerial economics. Journal of Business Ethics, 33(3), 225–231. Block, J. J. (2008). Issues for DSM-V: Internet addiction. American Journal of Psychiatry, 165(3), 306–307. Bock, G. W., & Ho, S. L. (2009). Non-work-related computing (NWRC). Communications of the ACM, 52(4), 124–128. Bradley, B. P. (1990). Behavioral addictions: Common features and treatment implications. British Journal of Addiction, 85, 1417–1419. Brenner, V. (1997). Psychology of computer use, parameters of Internet use, abuse and addiction: The first 90 days of the Internet usage survey. Psychological Reports, 80, 879–882. Bryant, J., & Zillmann, D. (Eds.). (2002). Media effects: Advances in theory and research. Mahwah, New Jersey: Lawrence Erlbaum Associates. Case, C. J., & Young, K. S. (2002). Employee Internet use policy: An examination of perceived effectiveness. Issues in Information Systems, 3, 82–88. Ceyhan, A. A., & Ceyhan, E. (2007). Loneliness, depression, and computer selfefficacy as predictors of problematic internet use. CyberPsychology and Behavior, 11(6), 699–701. Chen, J. V., Chen, C. C., & Yang, H. A. (2007). An empirical evaluation of key factors contributing to Internet abuse in the workplace. Industrial Management & Data Systems, 108(1), 87–106. Cheyne, C. L., & Ritter, F. E. (2001). Targeting audiences on the Internet. Communication of the ACM, 44, 94–98. Chou, C., & Hsiao, M. C. (2000). Internet addiction, usage, gratification, and pleasure experience. The Taiwan college students’ case. Computer & Education, 35, 65–80. Chou, C. H., Sinha, A. P., & Zhao, H. (2008). A text mining approach to Internet abuse detection. Information Systems and E-business Management, 6, 419–439. Cumming, G., & Finch, S. (2005). Inference by eye: Confidence intervals, and how to read pictures of data. American Psychologist, 60, 170–180. 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 and Behavior, 5(4), 331–345. Eastin, M. S., Glynn, C. L., & Griffiths, R. P. (2007). Psychology of communication technology use in the workplace. CyberPsychology and Behavior, 10, 436–443. Fox, A. (2007, December). Caught in the web. HR Magazine (pp. 35–39). Friedman, W. H. (2000). Is the answer to Internet addiction, Internet Gallup interdiction? In Proceedings of the 2000 Americas conference on information systems. AMCIS.

S.J. Kim, S. Byrne / Computers in Human Behavior 27 (2011) 2271–2283 Greenfield, D. N., & Davis, R. A. (2002). Lost in cyberspace. The web at work. Cyberpsychology and Behavior, 5(4), 347–353. Greengard, S. (2000). The high cost of cyber slacking. Workforce, 79(12), 22–24. Griffiths, M. (2003). Internet abuse in the workplace. Issues and concerns for employers and employment counselors. Journal of Employment Counseling, 40, 87–95. Hilts, M. L. (2008). Internet dependency, motivations for Internet use and their effects on work productivity: The 21st century addiction. Master’s thesis, Department of Communication, Rochester Institute of Technology, Rochester, NY. Jia, R., & Jia, H. (2009). Factorial validity of problematic Internet use scales. Computers in Human Behavior, 25, 1335–1342. Johnson, P. R., & Indvik, J. (2003). The organizational benefits of reducing cyberslacking in the workplace. Proceedings of the Academy of Organizational Culture, Communications and Conflict, 7(2), 53–60. Kamins, A. (1995, July). Cyber-loafing: Does employee time online add up to net losses? New York Daily News. Kandell, J. J. (1998). Internet addiction on campus: The vulnerability of college students. CyberPsychology and Behavior, 1, 11–17. Kraut, R., Lundmark, V., Patterson, M., Kiesler, S., Mukopadhyay, T., & Scherlis, W. (1998). Internet paradox: A social technology that reduces social involvement and psychological well-being? American Psychologist, 53, 1017–1031. LaRose, R., Lin, C. A., & Eastin, M. S. (2003). Unregulated Internet usage: Addiction, habit, or deficient self-regulation? Media Psychology, 5(3), 225–253. Lavoie, J., & Pychyl, T. A. (2001). Cyber-slacking and the procrastination super highway: A web-based survey of on-line procrastination, attitudes and emotion. Social Science Computer Review, 19(4), 431–444. Lee, O., Lim, K. H., & Wong, W. M. (2005). Why employees do non-work-related computing: An exploratory investigation through multiple theoretical perspectives. In Proceeding of the 35th Hawaii international conference on system sciences. Lim, K. G. (2002). The IT way of loafing on the job: Cyberloafing, neutralizing and organizational justice. Journal of Organizational Behavior, 23(5), 675–694. Lim, V. K. G., & Teo, T. S. H. (2005). Prevalence, perceived seriousness, justification and regulation of cyberloafing in Singapore: An exploratory study. Information & Management, 42, 1081–1093. Lim, V. K. G., Teo, T. S. H., & Loo, G. L. (2002). How do I loaf here? Let me count the ways. Communication of the ACM, 45(1), 66–70. Malachowski, D. (2005). Wasted time at work costing companies billions. . Manrique deLara, P. Z., Tacoronte, D. V., & Ding, J. T. (2006). Do current anticyberloafing disciplinary practices have a replica in research findings? A study of the effects of coercive strategies on workplace Internet misuse. Internet Research, 16(4), 450–467. Marlatt, G. A., Baer, J. S., Doncan, D. M., & Kivlahan, D. R. (1988). Addictive behaviors: Etiology and treatment. Annual Review of Psychology, 29, 223–252. Mills, J. E., Hu, B., Beldona, S., & Clay, J. (2001). Cyberslacking: A liability issue for wired workplaces. Cornell Hotel and Restaurant Administration Quarterly, 42(5), 34–48. Nalwa, K., & Anand, A. P. (2003). Internet addiction in students: A case of concern. Cyberpsychology and Behavior, 6(6), 653–656. Niemz, K., Griffiths, M., & Banyard, P. (2005). Prevalence of pathological Internet use among university students and correlations with self-esteem, the General Health Questionnaire (GHQ), and disinhibition. CyberPsychology and Behavior, 8(6), 562–569.

2283

Oravec, J. A. (2004). Bookmarking the world: Weblog applications in education. Journal of Adolescent & Adult Literacy, 45, 616–621. Padilla-Walker, L. M., Nelson, L. J., Carroll, J. S., & Jensen, A. C. (2009). More than just a game: Video game and Internet use during emerging adulthood. Journal of Youth & Adolescence, 39, 103–113. Panko, R. R., & Beh, H. G. (2002). Monitoring for pornography and sexual harassment. Communication of the ACM, 45(1), 84–87. Payton, M. E., Greenstone, M. H., & Schenker, N. (2003). Overlapping confidence intervals or standard error intervals: What do they mean in terms of statistical significance? Journal of Insect Science, 3(34), 1–6. PEW Research Center (2010). Searching, following and friending: How users monitor other people’s digital footprints online. Accessed 26.05.10. Robinson, S. L., & Bennett, R. J. (1995). A typology of deviant workplace behaviours: A multidimensional scaling study. Academy of Management Journal, 38, 555–571. Scherer, K. (1997). College life online: Healthy and unhealthy internet use. In Paper presented at the 10th annual meeting of the American Psychological Association, Chicago. Scheuermann, L. E., & Langford, H. R. (1997). Perceptions of Internet abuse, liability and fair use. Perceptual and Motor Skills, 85, 847–850. Shapira, N. A., Lessig, M. C., Goldsmith, T. D., Szabo, S. T., Lazoritz, M., Gold, Stein, et al. (2003). Problematic Internet use: Proposed classification and diagnostic criteria. Depression and Anxiety, 17(4), 207–216. Sharma, S. K., & Gupta, J. N. D. (2003). Improving workers’ productivity and reducing Internet abuse. The Journal of Computer Information Systems, 44(2), 74–78. Soule, L. C., Shell, W., & Kleen, B. A. (2003). Exploring Internet addiction: Demographic characteristics and stereotypes of heavy internet users. Journal of Computer Information Systems, 44(1), 64–73. Stanton, J. M. (2002). Company profile of the frequent Internet user: Web addict or happy employee? Communications of the Association for Computing Machinery, 45(1), 55–59. Stewart, F. (2000). Internet acceptable use policies: Navigating the management, legal, and technical issues. Information Systems Security, 9(3), 46–53. Terr, L. (1999). Beyond love and work: Why adults need to play (p. 226). Scribner. Warren, D. E. (2003). Constructive and destructive deviance in organizations. Academy of Management Review, 28, 622–632. Wong, W. M., Lee, O., & Lim, K. H. (2005). Managing non-work-related computing within an organization: The effects of two disciplinary approaches on employees’ commitment to change. In Proceedings of the ninth Pacific Asia conference on information systems (pp. 441–454), Bangkok, Thailand. Young, K. S. (1996). Addictive use of the Internet: A case that breaks the stereotype. Psychological Reports, 79, 899–902. Young, K. S. (1999). Evaluation and treatment of internet addiction. In VandeCreek, L., & Jackson, T. (Eds.), Innovations in clinical practice: A source book (Vol. 17, pp. 19–31). Sarasota, FL: Professional Resource Press. Young, K. S. (2004). Internet addiction: The consequences of a new clinical phenomenon. In Doyle, K. (Ed.), Psychology and the new media (Vol. 46, pp. 1– 14). Thousand Oaks: American Behavioral Scientist. Young, K. S., & Rogers, R. C. (1998). The relationship between depression and internet addiction. CyberPsychology and Behavior, 1, 25–28.