Reflection Paper (Assignment)

adolescence.selfinjury.pdf

Home >Psychology homework help >Reflection Paper (Assignment)

The Virtual Cutting Edge: The Internet and Adolescent Self-Injury

Janis L. Whitlock, Jane L. Powers, and John Eckenrode Cornell University

The 2 studies reported here use observational data from message boards to investigate how adolescents solicit and share information related to self-injurious behavior. Study 1 examines the prevalence and nature of these message boards, their users, and most commonly discussed topics. Study 2 was intended to explore the correlations between content areas raised for discussion. Both studies were intended to shed light on the role of message boards in spreading information about self-injurious practices and influencing help-seeking behavior. More than 400 self-injury message boards were identified. Most are populated by females who describe themselves as between 12 and 20 years of age. Findings show that online interactions clearly provide essential social support for otherwise isolated adolescents, but they may also normalize and encourage self-injurious behavior and add potentially lethal behaviors to the repertoire of established adolescent self-injurers and those exploring identity options.

Keywords: adolescence, self-injury, Internet, mental health

I think my greatest fear is to be forgotten. A teacher I had last year doesn’t even remember my name—it makes me think that no one remembers me. How do I know I exist? At least I know I exist when I cut. (Self-injury message board post)

Because adolescents use the Internet for the purpose of connect- ing with others at higher rates than any other age group (Lenhart, Rainie, & Lewis, 2001), a better understanding of how Internet use affects their social and emotional development is an important line of scientific inquiry. Indeed, a small but growing body of research is beginning to examine the implications of various electronic forums for social interaction (e.g., chat rooms, news groups, mes- sage boards) on adolescent behavior (e.g., Gross, 2004; Subrah- manyam, Greenfield, & Tynes, 2004; Tynes, Reynolds, & Green- field, 2004). The Internet may have particular relevance for adolescents who feel marginalized, because it provides a low-risk venue for finding others who share their perceived or real differ- ences and exchanging information that is difficult to convey in person or when using one’s real identity (McKenna & Green, 2002). Adolescents who intentionally injure themselves are one such group. Although research is nascent, adolescent self-injury appears to be increasingly common (Welsh, 2004; Yates, 2004) and, as this article shows, is a practice around which many virtual communities have formed. To date, almost nothing has been written about the existence of self-injury Internet forums, the types

of exchanges that occur there, or the ways in which these ex- changes may affect the development of adolescents and their ability to cope with distress.

Adolescent Internet Use

Computer access and use among adolescents have grown expo- nentially over the past decade (Becker, 2000). More than 80% of American youth 12 to 17 years of age use the Internet, and nearly half log on daily (Lenhart, Madden, & Hitlin, 2005). Once con- nected, adolescents engage in a wide variety of activities, includ- ing doing schoolwork, playing games, shopping, and downloading music. Research shows, however, that adolescents use the Internet primarily for social reasons (Gross, 2004; Roberts, Foehr, & Ride- out, 2005). The Internet has become a virtual meeting place where teens hang out with their peers to pass time. Many adolescents reportedly prefer being online to other media, including the tele- phone, TV, and radio (2002 Gallup Survey, cited in Heitner, 2002). According to data from the Pew Internet and American Life Project (Lenhart et al., 2005), the vast majority (89%) of teens use e-mail and 75% use instant messaging (IM), which allows them to have multiple simultaneous conversations with a defined group of peers. More than 50% of teens possess more than one e-mail address or screen name, which they can use to send private messages to friends or to participate anonymously in online fo- rums, such as chat rooms (Lenhart et al., 2001).

Some studies suggest that Internet use may facilitate social interaction by making it easier for individuals to connect with others they know as well as with strangers. It serves also as a powerful resource for youth desiring information about socially sensitive topics such as sexuality and interpersonal relations (Su- zuki & Calzo, 2004). This form of communication may be espe- cially advantageous for shy, socially anxious, or marginalized youth, enabling them to practice their social skills without the risks associated with “on the ground” interactions (Heitner, 2002; Mc- Kenna, Green, & Gleason, 2002; Subrahmanyam et al., 2004). Additionally, online communication may encourage more truthful

Janis L. Whitlock and John Eckenrode, Family Life Development Cen- ter and Department of Human Development, Cornell University; Jane L. Powers, Family Life Development Center, Cornell University.

This study was supported by funding from the Cornell College of Human Ecology Seed and Innovation Grant. We thank Amanda Purington, Alexis Matusiewicz, and the Cornell Self-Injury Study Team for their substantial work on data collection and analysis. We also thank Elizabeth R. Woods for her thoughtful reading of an earlier version of this article.

Correspondence concerning this article should be addressed to Janis L. Whitlock, Family Life Development Center, Cornell University, Beebe Hall, Ithaca, NY 14853-4401. E-mail: [email protected]

407

exchanges; many people report a greater willingness to share thoughts and feelings online than they would in face-to-face situ- ations (Lenhart et al., 2001; McKenna & Bargh, 2000). Clearly, the Internet is transforming the social world of adolescents by influ- encing how they communicate, establish and maintain relation- ships, and find social support. Nevertheless, the developmental consequences of adolescent Internet use is an area about which little is known (Wartella, Caplovitz, & Lee, 2004).

The Internet and Adolescent Social and Emotional Development

Three central tasks are integral to healthy social development during adolescence: (a) to establish caring, meaningful relation- ships; (b) to find acceptance and belonging in social groups; and (c) to establish interpersonal intimacy (Baumeister & Leary, 1995; Reis & Shaver, 1988; Sullivan, 1953). Peers play a crucial role in this process, because a positive relationship with peers is important for psychological well-being and social adjustment (Bishop & Inderbitzen, 1995; Hartup, 1996), whereas peer rejection is linked to serious problems, including delinquency, drug abuse, and de- pression (Hartup, 1996; Merten, 1996).

Early studies of the influence of online interactions on adoles- cent development suggested that high levels of Internet use may inhibit healthy social development by linking frequent use to social isolation and depression, especially among teenagers (Kraut et al., 1998; Nie & Erbring, 2000). However, these findings have been disputed. A follow-up study conducted by Kraut et al. (2002) found that the effects documented in their earlier study had largely dissipated. They did find, however, that effects of Internet use on depression differed for introverts and extroverts: Extroverts were more likely to feel greater social connection as a result of Internet use, whereas introverts became more depressed and withdrawn. Heitner (2002) found that adolescents who use the Internet to connect with others in real-time social exchanges tended to possess higher peer status, more social skills, and greater social integration than their more socially introverted and withdrawn peers, who spent most of their Internet time in solitary activities. Additionally, adolescents who used chat rooms exhibited lower peer status and had fewer social skills than those who did not. Similarly, Gross, Juvonen, and Gable (2002) found that teenagers with strong social connections use e-mail and IM to reinforce preexisting bonds, whereas those with less developed social networks use the anon- ymous features of the Internet to find new friends and social outlets, perhaps compensating for what they lack offline. This suggests that chat rooms and similar venues in which adolescents share experiences anonymously may provide a safe forum for less socially adept adolescents to practice social interaction.

Research finds that online exchange decreases social isolation among adolescents and helps them connect with people and ex- plore their identity (Maczewski, 2002; Suzuki & Calzo, 2004). This helps to explain how the Internet may serve as a virtual peer support group where adolescents under stress can express feelings and exchange information about modes of coping. Adolescence is also a time of increased feelings of distress (e.g., depressed mood) and increased access to modes of coping with stress that are independent of parents (Arnett, 1999; Compas, 1987; Petersen, Kennedy, & Sullivan, 1991). To the extent that Internet use can reduce feelings of social isolation, help normalize feelings of

distress through a process of self-disclosure and social comparison, and serve as a venue for giving and receiving social support, it may also provide a positive coping resource for distressed youth. Al- ternatively, Internet use may maintain or increase distress if the information exchanged reinforces negative views of self or sug- gests destructive or otherwise ineffective coping strategies.

This article examines the role of the Internet among adolescents who use self-injurious behavior (SIB) as a method of coping with distress. The opportunity to explore different identities and roles through the Internet may be particularly important for individuals with stigmatized identities, such as self-injurious youth, who feel that important aspects of their selves need to be concealed in their day-to-day lives. These youth may be especially motivated to participate in electronic forums that allow them to express them- selves in a safe and anonymous environment and to find support from others who share their sense of marginalization and under- stand their behavior. To date, however, not only is very little known about self-injury in the adolescent population, but nothing is known about how this group uses the Internet to connect with others about their practice.

Adolescent Self-Injury as a Developmental Phenomenon

The increasing number of stories in the mainstream press and popular media, as well as the growing number of anecdotal reports by physicians, therapists, and school counselors, suggest that self- injury may be “the next teen disorder” (Welsh, 2004). Although operationally elusive, scholars differentiate self-injury from cul- turally sanctioned forms of self-mutilation, such as piercing or tattooing, by intention rather than form. Alternatively called de- liberate self-harm, self-injury, self-mutilation, or cutting, self- injurious behavior typically refers to a variety of behaviors in which an individual purposefully inflicts harm to his or her body for purposes not socially recognized or sanctioned and without the obvious intention of committing suicide (Alderman, 1997; Favazza, 1996). Self-injury, which is most often associated with the term cutting, also includes intentional carving or cutting of the skin and subdermal tissue, scratching, burning, ripping or pulling skin or hair, swallowing toxic substances, bruising, and breaking bones. Although not typically a suicidal gesture, self-injury is statistically associated with suicide and can result in unanticipated severe harm or fatality (Claes, Vandereycken, & Vertommen, 2003; Favazza, DeRosear, & Conterio, 1989).

There are currently no reliable estimates of the prevalence of self-injury among the general U.S. adolescent population. The vast majority of research on self-injury has been conducted in clinical populations or using small, unrepresentative community samples. These studies generally find that cutting and other forms of self- injury are evident in approximately 20% of the clinical population (Deiter, Nicholls, & Pearlman, 2000) and are linked to high levels of pathology (Brodsky, Cloitre, & Dulit, 1995; Ross & Heath, 2003). The few studies that have been conducted in U.S. commu- nity samples of young adults and adolescents are limited by small convenience-based samples and vary in estimates of self-injury prevalence from 4% to 38% (Briere & Gil, 1998; Favazza, 1992; Gratz, Conrad, & Roemer, 2002; Kokaliari, 2005; Muehlenkamp & Gutierrez, 2004). Large studies in Britain estimate that approx- imately 10% of youth 11 to 25 years of age self-injure. A British report on the national scope of the problem documents a dramatic

408 WHITLOCK, POWERS, AND ECKENRODE

increase in disclosures of self-injury to national children’s help lines over the 5 years before the study, noting a 65% increase in the last 2 years (Young People and Self Harm: A National Inquiry, 2004).

The reasons for this apparent increase are unclear. Although general awareness of self-injury may lead to increased willingness to disclose the behavior, it may also reflect a true increase in incidence. Social contagion has been identified by some (Yates, 2004; Rosen & Walsh, 1989), because self-injury follows epidemic-like patterns in institutional settings such as hospitals and detention facilities (Matthews, 1968; Ross & McKay, 1979; Taiminen, Kallio-Soukainen, Nokso-Koivisto, Kaljonen, & Hele- nius, 1998). The possibility that self-injury is communicable may reflect a pattern similar to what Brumberg (1992) argued with regard to the spread of anorexia nervosa in the 1980s, when heightened cultural visibility through the mass media rendered anorexia nervosa an available emotional outlet for individuals with receptive predispositions.

Although no single self-injurer profile has emerged, there is general consensus that self-injury is most common among adoles- cents. Self-injurious behavior may parallel other problem behav- iors, which begin during early adolescence, peak during mid- to late adolescence, and then decline in adulthood (Briere & Gil, 1998; Favazza, 1999; Favazza & Conterio, 1989). However, there is some evidence that self-injurious behavior may follow two distinct patterns similar to the life course-persistent and adolescence-limited trajectories evident in antisocial behavior (Moffit, 1993). One of these begins in early childhood and persists through adulthood. The other follows a typical adolescence-limited course that emerges in early adolescence and declines in late adolescence or early adulthood (Dubo, Zanarini, Lewis, & Wil- liams, 1997; Nixon, Cloutier, & Aggarwal, 2002). Although some research finds females more likely to injure themselves than males (Conterio & Lader, 1998; Favazza, 1999), other studies suggest that the gender gap may be narrower (Briere & Gil, 1998; Deiter et al., 2000; Dulit, Fyer, Leon, Brodsky, & Frances, 1994; Galley, 2003; Martin, Rozanes, Pearce, & Allison, 1995). The difficulty in accurately assessing gender differences in self-injurious behavior may arise from variation in how male and female self-injurers are identified, how they injure themselves, and whether or not they seek treatment (Alderman, 1997; Connors, 2000).

The relationship between self-injury and suicide is important but not clearly understood. Persons who engage in self-injury are more likely to consider or attempt suicide (Walsh & Rosen, 1988; Gardner & Cowdry, 1985; Hawton, Fagg, Simkin, Bale, & Bond, 2000). Nevertheless, in the vast majority of cases self-injury is used to alleviate distress temporarily rather than to signal the intention to end one’s life (Favazza, 1996; Rosen & Walsh, 1989; Tantam & Whittaker, 1993). Indeed, some scholars see it as a highly functional alternative to suicide (Alderman, 1997; Strong, 1998). Among clinical populations, self-injury is comorbid with borderline personality disorder, eating disorders, posttraumatic stress disorder, depression, anxiety disorders, and a history of abuse or trauma (Alderman, 1997; Connors, 2000; Conterio & Lader, 1998; Holmes & Nadelson, 2000; Sansone & Levitt, 2002; Yates, 2004). Indeed, some researchers have called for a new Diagnostic and Statistical Manual of Mental Disorders impulse- control disorder, deliberate self-harm syndrome, which would include self-injurious behavior (Favazza et al., 1989). Most, how-

ever, see self-injury as a manifestation of mental or emotional disorders or heterotypic manifestation of childhood trauma (Con- ners, 2000; Conterio & Lader, 1998; Favaro & Santonastaso, 2002; Strong, 1998; Yates, 2004).

Whether a trauma-linked developmental model will hold true for nonclinical populations is unclear. However, the proximal causes for those who engage in self-injurious behavior in both clinical and community settings are strongly linked to difficulties in regulating strong emotions and coping with stress (Favazza, 1996; Rosen & Walsh, 1989; Ross & Heath, 2002; Tantam & Whittaker, 1993). The cumulative impact of stressful life events that trigger self-injurious episodes, the subsequent shame associ- ated with the self-injurious response, and the secrecy surrounding the behavior create fertile ground for the development of margin- alized perceptions of the self. The Internet provides self-injurers with an anonymous venue for sharing actual and fabricated aspects of their true identity among a community of similar others. Be- cause young people with depressive symptomatology are much more likely to talk with strangers online and to disclose personal information than those without depressive symptomatology (Ybarra, Alexander, & Mitchell, 2005), self-injurers may be particularly receptive to using the Internet to locate virtual communities. Al- though such virtual communities may provide a much-needed source of support and cathartic sharing, the possibility that self- injury is communicable suggests that the Internet may also serve to spread or deepen the practice in adolescent populations.

Aim of the Present Studies

Two studies were designed to explore and document adolescent use of online message boards to share, solicit, and receive infor- mation and advice related to self-injurious behavior. Message boards are an electronic venue in which individuals (i.e., “posters”) register with the site under a chosen name and are allowed to post thoughts, ask questions, and respond to other posters. Unlike e-mail and IM, which permit private exchange between selected individuals, message boards and chat rooms (in which exchanges occur in real time) are entirely public: All postings are available to all members and, in many cases, nonmembers as well, although nonmembers cannot post material. When individuals log on to a message board, the content varies but typically includes informa- tion about the message board purpose and rules as well as direct links to “threads,” posts that have been made by users sorted by subject title. Anyone who logs on to a message board may read posts, but only those registered may make a post. Blogs, which are becoming increasingly common, are essentially public electronic diaries. Message board posters frequently link their blog site to their membership identity or the signature line in their posts.

The goal of the first study was to investigate the prevalence and nature of self-injury message boards and their users with the intention of better understanding the general subject categories most commonly raised for discussion. The goal of the second study was to explore the correlations between content areas raised for discussion. Both studies were intended to shed light on the role of message boards in spreading information about self-injurious prac- tices and influencing help-seeking behavior. Studies of the Internet are effectively studies of culture. Participants are free to construct their age and identity as they wish, and studies show that teen Internet users often construct themselves as somewhat older than

409SPECIAL SECTION: VIRTUAL CUTTING EDGE

they actually are (Gross, 2004). Both studies were approved by the University Internal Review Board and have abided by the stipu- lation that participant quotes would be paraphrased rather than exact quotes. All quotes used here comply with this requirement. Because the studies use observational methods, active consent was not required.

Study 1: Prevalence and Nature of Self-Injury Message Board Use

The first study was undertaken to describe: (a) the prevalence of self-injury message boards, (b) the number of self-injury message boards advertised jointly with message boards for another disorder or behavior (such as eating or bipolar disorders), (c) the ratio of active and inactive posters, (d) the age range of users, as revealed in their self-descriptions, (e) date of appearance on the Web, (f) the relationship between using message boards and blogs, and (g) the nature of and variation in the content of postings on active message boards.

Method

Identification and selection. For this study, there were two units of analysis: the message board and individual posts in a selected set of message boards. First, to identify the prevalence of self-injury message boards, five Internet search engines were used: Yahoo, Google, MSN, AOL, and Gurl.com. Terms searched included self-injury, self-harm, self- mutilation, and cutting. More than 400 self-injury message boards were identified using this method; of these, 10 were selected for in-depth content analysis. These boards represented the first 10 listed in each search engine, which are generally those Web sites with the most activity. The moderation level of each of these boards varied as well. Moderation level refers to the degree to which posters are actively monitored for potentially damaging content (such as sharing techniques for self-injury). For example, when a particularly graphic or triggering post is made to a high- or medium-level moderation board, it is blocked altogether (high moderation) or labeled as a potential trigger (medium moderation). We were most interested in message boards with medium and low moderation, because content would be less likely to be censored. Detailed membership and moderation level information for the 10 sites selected was recorded (see Table 1 for a

description of the sites; names are withheld to protect confidentiality). A total of 3,219 individual posts, examined over a 2-month period, were coded for themes of interest and used in the content analysis that follows.

In addition to individual-level posts, we were also interested in gathering descriptive information about message board characteristics. Although most message boards do not contain information about date of inception, cross-listing with other topics, and membership characteristics, one of the Internet service providers did permit access to this information. The iden- tity of this search engine is hidden to protect confidentiality. All 140 message boards identified through this engine were examined for dates of inception, number of members, and cross-listing with other conditions.

Measures for message board analysis. Depending on the analysis, the following information was obtained for each message board: date of establishment, mean self-reported user age, number of active and inactive members, percentage of posters with blogs, and colisting with message boards for other behaviors. Date of establishment was determined using information from the home page or, if this was absent, from the date of the first post. The mean range of self-described age was determined by selecting 50 members at random from the site introduction board, where basic member profiles are stored. The number of active and inactive posters was determined by comparing the number of members who have registered but never posted with that of members who have posted. Moderation level was determined by identifying whether there were any instructions to users about allowable and unallowable posts and whether potentially triggering posts (i.e., events that provoke self-injurious behavior) were labeled as such. Cross-listing with other topics refers to the extent to which a message board was advertised as a virtual meeting place for individuals interested in or affected by the intersection of two or more issues (e.g., self-injury and suicide). Cross-listing was measured by examining the formal message board title and description of each board. When a cross-listed board was identified, the cross-listed topic was noted and the message board was considered as a cross-listed board.

Measures for content analysis of postings. The content analysis of posts was conducted using a set of binary (present–absent) codes. These codes were drawn from a synthesis of the research cited previously, from interviews conducted with 15 self-injurers by Janis L. Whitlock, and from pilot message board observations that spanned a 2-week period. A team of three coders systematically reviewed posts and used the constant compar- ative method (Glaser & Strauss, 1967) to inductively monitor themes. Once the initial set of observations had been coded, thematically grouped clusters were identified (e.g., motivation). The 11 broad categories identified en-

Table 1 Membership Breakdown of Selected Self-Injury Message Boards

Site

Age (years)

% Female No. members % w/Blogs

Members w/posts (%)

Mod levelaRange M Never 2–10 �100

A 12–44 18.7 88.1 844 30 41.0 20.0 9.5 Low B 13–54 18.3 90.0 5,259 25 57.0 15.5 7.2 Med C 14–47 19.4 88.0 5,082 15 14.0 44.1 17.3 High D 12–37 17.6 74.0 NA 25 31.5 32.8 2.0 Low E 14–22 17.5 91.5 70 15 45.7 41.5 0 Low F 14–36 19.6 87.5 6,656 10 63.4 10.9 8.5 Med G 16–47 20.5 90.0 153 20 65.0 9.8 9.8 Low H 14–28 18.1 80.0 4,021 30 41.3 15.9 21.0 Med I 15–46 23.9 80.6 1,427 15 30.4 27.8 9.5 Med J 13–26 16.4 78.0 2,581 10 50.8 25.0 10.0 Mod

Note. Mod � moderation; NA � not available; Med � medium. a Moderation level was determined by the extent to which the message board moderator blocked or labeled potentially triggering posts. In high-moderation boards, potentially triggering or disruptive posts were blocked entirely. In medium-moderated message boards, triggering or otherwise disruptive posts were identified and labeled. Low-moderation message boards took no steps to identify or block posts of any sort.

410 WHITLOCK, POWERS, AND ECKENRODE

compassed virtually all message board exchanges and served, along with all specific themes they encompassed, as a coding tool (e.g., loneliness, anger, and dysphoria were all themes in the motivation category). Posts left uncoded were those with little capacity to illuminate understanding of self-injurious behavior, such as idle discussion about a current event or activity. The resulting 11 areas were as follows: (a) informal support and exchange, (b) motivation for self-injury, (c) concealment of self-injurious behavior (e.g., anxiety about exposure, methods for concealment of cuts and scars), (d) addiction language (e.g., days self-injury free, difficulty stopping), (e) formal help seeking and treatment, (f) sharing techniques, (g) links to other mental health or behavioral conditions known to be associ- ated with self-injurious behavior, (h) references to popular culture, (i) perceptions of non-self-injurers reactions to self-injurious behavior, (j) perception of self and behavior (e.g., self-worth, lovability, dissociation), and (k) venting or apologizing. These areas contained a total of 70 themes into which nearly all posts could be categorized.

Coding of posts. A content analysis was conducted on all original and follow-up responses to the original post in 10 message boards over a 2-month period. A time frame of 2 months was selected to ensure adequate breadth in content areas. A total of 3,219 posts were examined during this period. Both original posts and responses to posted messages were coded. Because not all posts contained content relating to the coding scheme, the total number of posts examined exceeds the total number of posts to which a code was assigned (2,942). (The criterion for leaving postings uncoded is described later). However, if posts contained more than one thematic reference, they were assigned multiple codes. Therefore, totals in the Results section can add up to more than 2,942.

Four coders were trained by Janis L. Whitlock and the study coordinator to observe and code message boards during this time period. This was accomplished by having each coder and the principal investigator code the same posts and discuss code assignment. Once training was completed, each coder logged on one to two times per week during the study period to identify and code new posts. To establish intercoder reliability, each coder team was given three randomly selected threads to code. Codes for each individual coder were then compared to establish agreement within each team. The teams were in full agreement about codes for 96% and 97% of the postings, respectively.

Results

Message board prevalence, dates of inception, and colisting with other topics. The search procedure described previously revealed 406 boards. As shown in Table 2, examination of the 140 boards available through the Internet service providing historical information shows an increase from 1998 to 2000 and then a generally stable trend over time, with peaks in 2000 and in 2002. Although caution is warranted when interpreting these data, be-

cause public familiarity and dedicated server space with message boards have also increased over time, these findings indicate that message boards as a cultural tool for self-injury grew dramatically between 1998 and 2000 and that interest in both establishing and participating in self-injury specific groups over the past 5 years has been sustained.

Examination of these same message boards for colisting with other topics reveals links with a number of topics known to be associated with self-injury in the literature. Table 3 shows the extent to which self-injury message boards were cross-listed with other topics. Although listed most often alone (56%), 44% of the message boards were cross-listed with one or more other topics. When linked to another topic, self-injury occurs most often with depression (32%) and eating disorders (17%). Although less fre- quent, it often occurs with message boards dedicated to discussions of dissociative identity and multiple personality disorders (10%); bipolar disorder (9.2%); sexual abuse (7%); obsessive– compulsive disorders (7%); addiction (2.8%); anxiety disorder (2.8%); lesbian, gay, bisexual, and transgender issues (2.8%); and autism (1.4%).

Message board membership analysis. The 10 most popular message boards were selected for membership analysis. Table 1 shows the averages for self-description: member age, age range, and percentage of females. It also shows the percentage of mem- bers with public blogs, the percentage of registered members who have never posted (potential viewers), percentage of posters with more than 100 posts, and moderation level. The average self- described age of members ranged from 16.4 to 23.9 years for each message board, although there was large variation in ages repre- sented and the mean tends to be negatively skewed. The majority of message boards had a mean stated age of 18 years; 80% of the members described themselves as being between the ages of 14 and 20; 31% of all members described themselves as being 15 or 16 years old. In all 10 boards examined, posters describing them- selves as female were more likely to be registered and to partici- pate actively. The number of members in each message board varied dramatically from 70 to 6,656. Membership data from one of the sites were not available because members post in any of a wide number of forums, not all of which relate to self-injury.

Table 3 Conditions and Behaviors With Which Self-Injury Message Boards Are Cross-Listed

Cross-listed condition/behavior Frequency (n � 140)

% of Total

Depression 45 32.0 Eating disorders 24 17.0 Dissociative identity disorder/multiple

personalities 14 10.0 Bipolar 13 9.2 Obsessive-compulsive disorder 10 7.1 Sexual abuse 10 7.1 Posttraumatic stress disorder 7 5.0 Anxiety 4 2.9 LGBTQ 2 1.4 Mood disorders 4 2.8 Addiction 4 2.8 Autism 2 1.4 No other disorder specified 78 56.0

Note. LGBTQ � lesbian, gay, bisexual, transgender, and questioning.

Table 2 Number of Self-Injury Message Boards From One Internet Service Established by Year

Year No. boards Total membership

1998 1 93 1999 7 949 2000 26 2,831 2001 25 703 2002 28 1,611 2003 19 952 2004 24 806 2005 38 1,698

Total 168 9,643

411SPECIAL SECTION: VIRTUAL CUTTING EDGE

The number of message board members with blogs ranged from 10% to 30%. Preliminary examination of blog content revealed few substantive differences from message board posts other than the fact that blogs often provided more detailed accounts of events and feelings. There was considerable variation in the extent to which members posted. In four of the boards, more than half of the membership had never posted a single comment. The majority of those who did participate posted 2 to 10 messages, although each message board had a group of high posters with more than 100 posts.

One notable characteristic of message board membership was that virtually all members have images associated with their online identity. Characters, such as “Emily the Strange,” digital images made by posters of scars and bleeding wounds, and other graphic and bloody icons were frequently included with posted text. Also, lines from songs, poetry, or books were included in signature lines and were often expressions of sentiment. One example is the lyric by the music artist Eminem: “Sometimes I even cut myself to see how much it bleeds. It’s like adrenaline, the pain is such a sudden rush for me.”

Message board content. A total of 3,219 posts were examined from six message boards over a 2-month period; 2,942 were assigned a code. Although posts could be assigned multiple codes, most were not. Tables 4 and 5 list the total number of times a post was assigned a particular content code and the proportion of all posts examined that fell into this content area. As shown in Table 4, the vast majority of all posts examined fell into 1 of 11 broad thematic categories. Several of the most common categories con- tained a variety of subthemes (see Table 5).

Providing informal support to other posters was the most com- mon type of exchange on the message boards, occurring in 28.3% of all posts (see Table 4). Comments such as “We’re glad you’ve come here” and “Just relax and try to breathe deeply and slowly” were very common. Although less frequent, exchanges around support were often linked with apologizing behavior (“I’m so sorry to lay this on you”) and confessions of self-loathing (“I hate myself for doing this”). Discussion of events that triggered a self-injurious episode occurred in 19.5% of the posts (see Table 4). Conflict with important others constituted the primary trigger (34.8% of all

trigger-related posts) followed by depression (24.8%) and stress (10.3%; see Table 5).

The next largest thematic category discussed involved conceal- ment of the practice and its effects (primarily scarring), accounting for 9.1% of all posts examined (see Table 4). Comments focused largely on anxiety about being discovered, how to manage scars, and the extent to which posters had to be dishonest to maintain secrecy (see Table 5).

Posts about the perceived addictiveness of self-injury were almost as common (8.9%) (see Table 4). Typical examples include “It just haunts me and I don’t think I’ll ever get away from it” and “I may try and quit but even if I succeed, I’ll always dream of razorblades and blood.” Observations starting with the phrase “I’ve been cutting free for [length of time]” accounted for almost half of posts coded with addiction elements. References to stable patterns of self-injury, tendency to minimize the problem, similar- ities to other drugs, multiple quit attempts, the need to self-injure more or more deeply because of increased tolerance for effects of self-injury, and relapse after quitting were all areas discussed within this category (see Table 5). Often this discussion co- occurred with comments regarding frustration at parents, caregiv- ers, and others, who, according to the posters, do not appreciate the addictive nature of the behavior.

Discussion of formal help seeking from physical or mental health professionals occurred in 7.1% of all posts (see Table 4). Attitudes toward and experiences with treatment were largely positive (44.1% of all help-seeking posts reflected positive atti- tudes toward therapy), and discussion of experiences related to

Table 4 Dominant Thematic Categories in Message Board Content

Category No. of category

occurrences

% Posts examined

(N � 3,219)a

Informal provision of support for others 913 28.3 Motivation/triggers 629 19.5 Concealment issues 292 9.1 Addiction elements 288 8.9 Formal help seeking, treatment 229 7.1 Requesting, sharing techniques 200 6.2 Link to other mental health conditions 153 4.7 References to popular culture 137 4.2 Interpretation of other’s perceptions 85 2.6 Perceptions of self 70 2.1 Venting or apologizing behavior 61 2.9 Uncategorized 277 8.6

a Individual posts could be assigned multiple codes; total percentages will not equal 100%.

Table 5 Breakdown of Most Common Themesa

Category/themes Total no. of occurrences

Percentage of primary category

Triggers Conflict with important others 212 34.8 Perceived depression 151 24.8 School or work stress 63 10.3 Loneliness 44 7.2 Sexual abuse/rape 22 3.6 Other 117 19.2

Concealment issues Anxiety about concealment 149 51.8 Managing scars 109 37.3 Acknowledgment of dishonesty 29 10.9

Addictive elements “I’ve been cutting free for . . .” 120 41.7 Stable pattern/cannot control urge 52 18.1 Minimize problem 28 9.7 Liken to other drugs 27 9.3 Multiple attempts to quit 27 9.4 Increased tolerance, need more over time 24 8.4 Relapse 10 3.4

Help seeking and treatment Positive attitudes toward therapy 101 44.1 In therapy 60 26.2 Negative attitude toward therapy 43 18.7 References to medications 25 10.9

a Provision of informal support, although the dominant theme, is not included in this table because specific categories of support were not tracked.

412 WHITLOCK, POWERS, AND ECKENRODE

therapy accounted for a sizable portion of formal help-seeking posts as well (26.2%). Negative attitudes and active discourage- ment from seeking therapy were also evident in 18.7% of the formal help-seeking posts. References to specific medications oc- curred in about 1 in 10 cases.

The sixth most common category of discussion for all age groups related to sharing details about techniques use to self- injury, accounting for 6.2% of all posts. These exchanges were generally either descriptions of specific self-injury techniques or requests for specific technique information. The following conver- sation exemplifies this type of exchange:

Poster 1: Does anyone know how to cut deep without having it sting and bleed too much?

Poster 2: I use box cutter blades. You have to pull the skin really tight and press the blade down really hard. You can also use a tourniquet to make it bleed more.

Poster 3: I’ve found that if you press your blade against the skin at the depth you want the cut to be and draw the blade really fast it doesn’t hurt and there is blood galore. Be careful, though, ‘cause you can go very deep without meaning to.

Poster 1: Okay, I’ll get a Stanley blade ‘cause I hear that it will cut right to the bone with no hassle. But I’ll be careful if I do use a tourniquet and I won’t cut that deep.

Mental health conditions empirically linked to self-injurious behavior, such as depression, eating disorders, suicidality, and sexual abuse, were cited as either linked with or as a trigger for self-injury in 4.7% of all posts (see Table 4). References to pop culture (e.g., music, movies, books, celebrities, and characters with special significance to posters) appeared in just over 4% of the posts. Discussion and interpretation of the perceptions of others came up in 2.6% of the posts. Recognition of the pain their behavior caused or might cause others, in conjunction with diffi- culty stopping a behavior once a pattern was formed, was often linked to the need for secrecy and feelings of shame. Similarly, perceptions of self, usually negative, were shared in just over 2% of all posts. Remaining posts were rants, aimless venting, usually frustration or anger, or apologies for sharing.

Study 2: Variations in the Use of Message Boards Among Self-Injurers

This study was intended to explore the correlations between content areas raised for discussion. It focuses on exchange in four broad areas: (a) help seeking and disclosure, (b) technique sharing, (c) comorbidity, and (d) attitudes toward self and other posters. These areas were selected to explore the role of message boards exchange in spreading information about self-injurious practices and influencing help-seeking behavior.

Method

Identification and sampling. The unit of analysis for this study was the individual self-injury message board poster. Individual posters were selected from five low- and one medium-moderation self-injury mes- sage boards included in the previous study (A, B, D, E, F, and G in Table 1).

Because the extent to which posters sought and shared self-injury techniques was of interest, low-moderation boards were selected as recruitment sites to ensure that the content of message posts would not be blocked. One medium-moderation board was selected to permit a balanced sample by age, because the low-moderation boards did not contain an adequate number of active posters for which age was available. Each site allowed for all posts from individual posters to be searched readily.

Individual posters were identified by looking through threads and iden- tifying individuals in the age brackets of interest with more than 50 posts. Then 20 in each of three self-described age groups (13–15, 16 –18, 19 –22 years) were selected at random, for a total sample size of 60. Once individual posters were identified, 50 posts from each individual were selected at random from posts made from July 2004 to January 2005. A 6-month window was chosen to ensure that each poster monitored would have adequate time to make multiple posts and engage in a variety of interactions.

Measures. As in Study 1, the content analysis of posts was conducted using a set of binary (present–absent) codes. These codes were similar to those used in Study 1 but were focused on six specific types of exchange: (a) soliciting and sharing techniques, (b) attitudes toward and disclosed use of formal support, (c) solicitation and provision of informal support, (d) disclosure to nontherapeutic others (e.g., family and friends), (e) disclosure of other mental health conditions, and (f) disclosed self-concept. A total of 17 measures were used to indicate each of these broad conceptual domains (Table 6). Although many of these measures are similar to those used in Study 1, Study 2 measures were designed to focus on a specific type of exchange and to permit individual-level analysis of correlations between variables. Self-represented demographics, including gender, age, and total number of posts, were also recorded.

Coding of posts. A total of 3,000 individual posts were coded (50 per each 60 individuals). The actual number of posts per individual ranged from 60 to several hundred. Because coders could query for all threads posted by individual posters, they randomly selected 50 by identifying the total number of pages and threads any one poster made and dividing by 50. The resulting number was used to systematically identify threads included in the analysis. For example, if a poster made 100 posts, coders coded every two posts. If the resulting interval inadvertently identified a redun- dant thread (one already coded but that surfaced later), the next thread was used. Posts were not coded when they did not contain content relevant to the study objectives. Six coders were assigned individual posters to track and code. To establish intercoder reliability, 40% of the observations (1,200 posts) were independently coded by three pairs of coders. Each pair of coders coded 400 posts in common. No two pairs coded the same posts. Agreement was assessed by calculating the proportion of posts each individual in the pair coded the same. The average intercoder agreement across all three pairs was 93% (range � 90 –96%).

Results

To assess the relationship between soliciting and sharing self- injurious practices and informal and formal help seeking, Spear- man correlations were conducted between all measures. Results are shown in Table 6.

The analysis revealed two trends in the correlations: one in which more positive exchanges were correlated and one in which negative exchanges were correlated. For example, offering infor- mal support was correlated with disclosing that someone knows, suggesting formal treatment, seeking advice on stopping, and disclosing a history of trauma. Similarly, seeking advice on stop- ping and harm reduction were correlated with each other and with seeking advice about disclosure and sharing positive remarks about oneself. Encouraging formal treatment of self-injurious be-

413SPECIAL SECTION: VIRTUAL CUTTING EDGE

T ab

le 6

S p ea

rm a n ’s

C o rr

el a ti

o n s

fo r

A ll

S tu

d y

V a ri

a b le

s (N

� 6 0

in d iv

id u a ls

; 3 ,0

0 0

p o st

M ea

su re

1 2

3 4

5 6

7 8

9 10

11 12

13 14

15 16

T ec

hn iq

ue s

1. S

ha re

te ch

ni qu

es —

2. R

eq ue

st te

ch ni

qu es

.2 31

—

In fo

rm al

su pp

or t

3. O

ff er

in fo

rm al

su pp

or t

.0 75

� .2

04 —

4. S

ee k

ad vi

ce on

st op

pi ng

.0 98

.0 88

.2 99

* —

5. S

ee k

ad vi

ce on

re du

ci ng

ha rm

� .0

42 .0

85 .1

08 .3

19 *

—

D is

cl os

ur e

at ti

tu de

6. S

ee k

ad vi

ce ab

ou t

di sc

lo su

re �

.0 84

� .2

18 .0

06 .2

28 .2

81 *

— 7.

E nc

ou ra

ge di

sc lo

su re

.1 03

.0 48

.1 87

.1 12

.0 09

.1 63

— 8.

D is

co ur

ag e

di sc

lo su

re .3

32 **

.0 09

.1 26

.3 14

* .2

43 .1

60 .1

34 —

F or

m al

su pp

or t

9. D

is cl

os e

tr ea

tm en

t .1

11 �

.1 45

.0 82

.1 50

.1 14

.0 02

� .0

88 .0

73 —

10 .

S ug

ge st

fo rm

al tr

ea tm

en t

� .0

25 �

.2 28

.5 29

** .1

29 �

.1 56

.1 28

.3 34

** .2

39 .1

08 —

D is

cl os

ur e

st at

11 .

D is

cl os

e th

at so

m eo

ne kn

ow s

.3 15

* �

.0 79

.2 89

* .1

91 .0

04 .0

56 .1

25 .1

97 .1

99 .1

69 —

A ss

oc ia

te d

co nd

it io

12 .

S ui

ci da

li ty

.1 10

� .1

76 .1

45 .1

49 .1

63 .0

31 �

.0 87

.1 67

.4 57

** �

.1 90

.2 99

* —

13 .

H is

to ry

of tr

au m

a .0

55 .1

05 �

.2 86

* .0

22 .0

08 .0

99 �

.0 82

.1 06

.1 57

� .0

24 �

.0 13

.0 05

— 14

. D

ia gn

os ed

di so

rd er

.1 64

.1 20

� .0

32 .1

08 .0

96 .0

82 �

.0 58

.1 20

.4 78

** .1

25 �

.0 48

.1 42

.2 34

— 15

. D

is pa

ra gi

ng co

m m

en ts

ab ou

t se

lf .1

85 .1

33 �

.0 68

.0 61

.1 07

.3 19

* �

.1 84

� .0

59 .0

92 .0

08 .0

70 .1

71 .1

39 .2

61 *

— 16

. P

os it

iv e

co m

m en

ts ab

ou t

se lf

.1 83

� .0

87 �

.1 27

.2 91

* .2

07 .2

37 .1

99 .3

01 *

.1 32

.1 82

.1 44

.1 54

.0 29

.1 03

.2 69

* —

17 .

D is

pa ra

gi ng

co m

m en

ts ab

ou t

ot he

rs .0

29 �

.0 10

� .1

44 .1

98 �

.0 31

� .1

14 .2

93 *

.2 37

� .0

03 .1

00 .0

66 .1

93 �

.1 02

� .1

43 �

.2 41

.0 98

—

* p

� .0

5. **

p �

.0 1.

414 WHITLOCK, POWERS, AND ECKENRODE

havior, offering informal support, and encouraging disclosure to others were interrelated as well.

A somewhat more negative cluster of exchange linked discus- sion of technique sharing with negative attitudes toward disclo- sure. However, discouraging disclosure was also associated both with positive views of self and with seeking advice on stopping. In addition, disclosing that one was in treatment was correlated with disclosing another diagnosed disorder and suicidality. Individuals who disclosed a history of trauma were likely to offer informal support, and those who disclosed suicide-related behaviors were more likely than others to disclose that someone knows about their self-injurious behavior.

Discussion

Our findings confirm that Internet message boards provide a powerful vehicle for bringing together self-injurious adolescents. Although the message boards examined for these studies may not be representative of all self-injury message boards, they do provide a snapshot of content and exchange common in those with high activity. We found that in the last 5 years hundreds of message boards specifically designed to provide a safe forum for self- injurious individuals have come into existence. Many of these are populated by individuals who identify themselves as females be- tween 14 and 20 years of age. Although the strong preponderance of females may not accurately reflect the gender breakdown of self-injurious behavior in the general population (Whitlock, Eck- enrode, & Silverman, in press), it may reflect the tendency for females to solicit more informal and formal help and social support compared with males (Fuhrer, Stansfield, Chemali, & Shipley, 1999; Saunders, Resnick, Hoberman, & Blum, 1994).

Once online, message board members are able to post or pas- sively observe a wide variety of anonymous exchanges. Just less than half of all message boards we investigated were cross-listed with conditions known to be comorbid with self-injurious behav- ior, such as depression, eating disorders, and suicide. Of all types of online dialogue, the giving and receiving of informal support and discussion of proximal life events that trigger self-injury are most common. Posters also share casual and sometimes very personal information related to the addictive qualities of their practice, their fears relating to disclosure, experiences with ther- apy, how they self-injure, and other related health concerns. Al- though our findings are generally consistent with existing literature on self-injurious behavior, the correlations documented among informal support, encouraging disclosure, and advising formal treatment suggest that online interactions may be providing self- injurers support and meaning outside the clinical setting.

What the self-injurious adolescents in our study appeared to do online is what most people who trust each other do in conversa- tion: exchange support, share personal stories about daily life events, and voice opinions and ideas. Because the anonymity of the Internet inspires the most personal and trusting of exchange between individuals with little or no previous relationship, online sharing may encourage greater and more truthful disclosures (McKenna & Bargh, 2000), especially among self-injurers, many of whom suffer from depressive symptomatology (Ybarra et al., 2005). For adolescents, this support may be particularly valuable, because healthy social and emotional development hinges on their ability to establish caring, meaningful relationships, to find accep-

tance and belonging in social groups, and to establish interpersonal intimacy (Reis & Shaver, 1988; Sullivan, 1953). These develop- mental tasks can be especially difficult for young people struggling with intense shame, isolation, and distress, particularly when the source and outcome of these feelings must be kept hidden. The assurance of online anonymity may contribute to identity construc- tion by providing opportunities for adolescents with marginalized or nonmainstream proclivities to experiment with different social roles and selves (McKenna & Bargh, 2000; Turkle, 1995).

The less positive side of our findings suggests that participation in self-injury message boards may also expose vulnerable adoles- cents to a subculture in which self-injury is normalized and en- couraged. For example, issues related to concealment of self- injurious behavior, identified as a dominant theme here, may make self-injury message boards particularly potent agents of self-injury, because sharing techniques and motives can take place anony- mously. In light of evidence here and elsewhere that self-injurious behavior may possess addictive qualities (see Yates, 2004, for review), the adolescent drive to belong and the satisfaction that comes with associating with a community of similar others may inadvertently feed a fundamentally self-destructive behavior for some participants. The correlation documented here between shar- ing injurious techniques and discouraging disclosure lends support to this possibility.

Although not impossible for individual self-injurers to have gathered before the advent of the Internet, easy access to a virtual subculture of like-minded others may reinforce the behavior for a much larger number of youth. The tendency for self-injurious behavior to follow epidemic-like patterns in institutional settings such as hospitals and detention facilities (Matthews, 1968; Ross & McKay, 1979; Rosen & Walsh, 1989; Taiminen, Kallio- Soukainen, Nokso-Koivisto, Kaljonen, & Helenius, 1998) suggests that the behavior may be socially contagious in other settings and, therefore, through the Internet as well. As Brumberg (1992) has argued for eating disorders, discussion of techniques and the perceived benefits of self-injury may add potentially lethal behav- iors to the repertoire of established self-injurers, not yet committed message board members, and even nonparticipating message board viewers who are exploring identity options. Indeed, some message boards contain links to pro-self-injury Web sites where Internet users can purchase self-injury paraphernalia such as bracelets or clothing that signify self-injury status and cutting clubs have been rumored to be a growing form of friendship ritual (Booth, 2004). Some self-injury Web sites host forums specifically dedicated to sharing new self-injury techniques. Discussion of technique shar- ing, triggers, negative attitudes toward formal and informal help seeking, and the pleasures and pains of self-injury addiction may influence behavioral choices outside of the virtual realm that are later brought back, shared, and used to ensure support and mem- bership. It may also make some youth targets for individuals who falsely pose as supporters to accomplish other, less benevolent aims. For vulnerable adolescents, the difficulty of ending a strat- egy for coping with distress (self-injury) and leaving a needed source of support (individual or collective members of the virtual community) may stifle the desire to find alternate ways of coping with stress. Moreover, the low-sense of self-worth common among self-injurers may expose them to damaging online relationships.

This study supports the findings of other scholars of Internet and development (Suzuki & Calzo, 2004) by suggesting that electronic

415SPECIAL SECTION: VIRTUAL CUTTING EDGE

forums provide a rich data source for studying issues pertinent to marginalized subgroups of the adolescent population that are hard to identify and reach. Observation of interaction as it unfolds, rather than in retrospective self-reports, eliminates biases relating to self-report and recall and allows study of actual transactions between individuals. Moreover, because self-injury is typically a private, secretive behavior, the Internet provides a unique opportunity to study exchange between members of a group rarely assembled outside of a clinical setting. As such, it may provide a valuable means of accessing information and perspectives useful in clinical settings. Indeed, self-injury message boards may provide a vehicle for admin- istering a Web-based intervention to reach self-injurers.

Our methods have limits. It was not possible to observe all 406 message boards identified in this study. The 10 selected for anal- ysis, although the most active at the time of the study, may not be generalizable to all self-injury message boards. Message boards, like all communities, possess their own culture and character and are governed by subtle and overt norms and mores. Rules posted and enforced by the moderator is one example of this. How these affect participant self-selection into specific message board com- munities and the content of their exchange is unclear. Moreover, the absence of demographic data on individual Internet users prohibits many important and useful analyses.

In conclusion, this research has several implications for those interested in the nexus between adolescents, the Internet, and self-injury. Like other social environments (e.g., schools, families, or neighborhoods), the Internet plays a powerful role in shaping opportunities for adaptive and maladaptive social interaction. What occurs in the virtual world, however, is largely invisible to adults and treatment providers who have a vested interest in adolescents and self-injurious behavior. Although it is neither possible nor, perhaps, desirable to monitor all adolescent Internet use, particularly as youth age and become increasingly indepen- dent, it is very important for adults to know something about what adolescents, particularly vulnerable adolescents, encounter in the virtual communities they inhabit. The data presented here are exploratory and would be strengthened by tracking a larger num- ber of individual posters over a longer period of time to prospec- tively capture variability over time within individual and within and between age groups. Additionally, future research in this area should seek to verify the accuracy of what adolescent users share online, to better understand the unique role the Internet may play in affecting off-line behavior, and to clarify the role played by Internet use in spreading or deepening self-injurious practices among adolescents.

References

Alderman, T. (1997). The scarred soul: Understanding and ending self- inflicted violence. Oakland, CA: New Harbinger.

Arnett, J. J. (1999). Adolescent storm and stress, reconsidered. American Psychologist, 54, 317–326.

Baumeister, R. F., & Leary, M. R. (1995). The need to belong: Desire for interpersonal attachments as a fundamental human motivation. Psycho- logical Bulletin, 117, 497–529.

Becker, H. J. (2000). Who’s wired and who’s not: Children’s access to and use of computer technology. Future of Children, 10, 44 –75.

Bishop, J. A., & Inderbitzen, H. M. (1995). Peer acceptance of friendship: An investigation of the relation to self-esteem. Journal of Early Ado- lescence, 15, 476 – 489.

Booth, S. (2004, February). Cutting clubs: What’s the latest and most shocking new “friendship” ritual? Teen People looks at how a growing number of kids are bonding with their peers by slicing themselves with razor blades. Teen People, 7, 98.

Briere, J., & Gil, E. (1998). Self-mutilation in clinical and general popu- lation samples: Prevalence, correlates, and functions. American Journal of Orthopsychiatry, 68, 609 – 620.

Brodsky, B. S., Cloitre, M., & Dulit, R. A. (1995). Relationship of dissociation to self-mutilation and childhood abuse in borderline per- sonality disorder. American Journal of Psychiatry, 152, 1788 –1792.

Brumberg, J. J. (1992). From psychiatric syndrome to “communicable” disease: The case of anorexia nervosa. In C. Rosenberg & J. Golden (Eds.), Framing disease (pp. 134 –154). Piscataway, NJ: Rutgers Uni- versity Press.

Claes, L., Vandereycken, W., & Vertommen, H. (2003). Eating-disordered patients with and without self-injurious behaviours: A comparison of psy- chopathological features. European Eating Disorders Review, 11, 379 –396.

Compas, B. E. (1987). Coping with stress during childhood and adoles- cence. Psychological Bulletin, 101, 393– 403.

Connors, R. (2000). Self-injury: Psychotherapy with people who engage in self-inflicted violence. Northvale, NJ: Jason Aronson.

Conterio, K., & Lader, W. (1998). Bodily harm: The breakthrough healing program for self injurers. New York: Hyperion Press.

Deiter, P. J., Nicholls, S. S., & Pearlman, L. A. (2000). Self-injury and self capacities: Assisting an individual in crisis. Journal of Clinical Psychol- ogy, 56, 1173–1191.

Dubo, E. D., Zanarini, M. C., Lewis, R. E., & Williams, A. A. (1997). Childhood antecedents of self-destructiveness in borderline personality disorder. Canadian Journal of Psychiatry, 42, 63– 69.

Dulit, R. A., Fyer, M. R., Leon, A. C., Brodsky, B. S., & Frances, A. J. (1994). Clinical correlates of self-mutilation in borderline personality disorder. American Journal of Psychiatry, 151, 1305–1311.

Favaro, A., & Santonastaso, P. (2002). The spectrum of self-injurious behavior in eating disorders. Eating Disorders, 10, 215–225.

Favazza, A. R. (1992). Repetitive self-mutilation. Psychiatric Annals, 22, 60 – 63.

Favazza, A. R. (1996). Bodies under siege: Self-mutilation and body modification in culture and psychiatry (2nd ed.). Baltimore, MD: Johns Hopkins University Press.

Favazza, A. R. (1998). The coming age of self-mutilation. Journal of Nervous and Mental Disease, 186 (5), 259 –268.

Favazza, A. R. (1999). Self-mutilation. In D. G. Jacobs (Ed.), The Harvard Medical School guide to suicide assessment and intervention (pp. 125– 145). San Francisco: Jossey-Bass.

Favazza, A. R., & Conterio, K. (1989). Female habitual self-mutilators. Acta Psychiatrica Scandinavica, 79, 283–289.

Favazza, A. R., DeRosear, L., & Conterio, K. (1989). Self-mutilation and eating disorders. Suicide & Life-Threatening Behavior, 19, 352–361.

Fuhrer, R., Stansfield, S. A., Chemali, J., & Shipley, M. J. (1999). Gender, social relations and mental health: Prospective findings from an occu- pational cohort. Social Science & Medicine, 48, 77– 87.

Galley, M. (2003, December 3). Student self-harm: Silent school crisis. Education Week, p. 2.

Gardner, D. L., & Cowdry, R. W. (1985). Suicidal and parasuicidal behavior in borderline personality disorder. Psychiatric Clinics of North America, 8, 389 – 403.

Glaser, B. G., & Strauss, A. L. (1967). The discovery of grounded theory. Strategies for qualitative research. Chicago: Aldine.

Gratz, K. L., Conrad, S. D., & Roemer, L. (2002). Risk factors for deliberate self-harm among college students. American Journal of Or- thopsychiatry, 72, 128 –140.

Gross, E. F. (2004). Adolescent Internet use: What we expect, what teens report. Journal of Applied Developmental Psychology, 25, 633– 649.

416 WHITLOCK, POWERS, AND ECKENRODE

Gross, E. F., Juvonen, J., & Gable, S. L. (2002). Internet use and well-being in adolescence. Journal of Social Issues, 58, 75–90.

Hartup, W. W. (1996). The company they keep: Friendships and their developmental significance. Child Development, 67, 1–13.

Hawton, K., Fagg, J., Simkin, S., Bale, E., & Bond, A. (2000). Deliberate self-harm in Oxford, 1985–1995. Journal of Adolescence, 23, 47–55.

Heitner, E. (2002). The relationship between use of the Internet and social development. Unpublished doctoral dissertation, Pace University.

Holmes, A., & Nadelson, C. A. (Eds.). (2000). Cutting the pain away: Understanding self mutilation. Philadelphia: Chelsea House.

Kokaliari, E. (2005). Deliberate self-injury: An investigation of the prev- alence and psychosocial meanings in a non-clinical female college population. Unpublished doctoral dissertation, Smith College, School for Social Work.

Kraut, R., Kiesler, S., Boneva, B., Cummings, J., Helgeson, V., & Crawford, A. (2002). Internet paradox revisited. Journal of Social Issues, 51, 49 –74.

Kraut, R., Patterson, M., Lundmark, V., Kiesler, S., Mukopadhyay, T., & Scherlis, W. (1998). Internet paradox: A social technology that reduces social involvement and psychological well-being? American Psycholo- gist, 53, 1017–1031.

Lenhart, A., Madden, M., & Hitlin, P. (2005, July 27). Teens and technol- ogy: Youth are leading the transition to a fully wired and mobile nation. Retrieved October 13, 2005, from http://www.pewInternet.org/pdfs/ PIP_Teens_Tech_July2005web.pdf

Lenhart, A., Rainie, L., & Lewis, O. (2001). Teenage life online: The rise of the instant-message generation and the Internet’s impact on friend- ships and family relationships. Washington, DC: Pew Internet & Amer- ican Life Project.

Maczewski, M. (2002). Exploring identities through the Internet: Youth experiences online. Child & Youth Care Forum, 31, 111–129.

Martin, G., Rozanes, P., Pearce, C., & Allison, S. (1995). Adolescent suicide, depression and family dysfunction. Acta Psychiatrica Scandi- navica, 92, 336 –344.

Matthews, P. C. (1968). Epidemic self-injury in an adolescent unit. Inter- national Journal of Social Psychiatry, 14, 125–133.

McKenna, K. Y. A., & Bargh, J. A. (2000). Plan 9 from Cyberspace: The implications of the Internet for personality and social psychology. Per- sonality and Social Psychology Review, 4, 57–75.

McKenna, K. Y. A., & Green, A. S. (2002). Virtual group dynamics. Group Dynamics: Theory, Research and Practice, 6, 116 –127.

McKenna, K. Y. A., Green, A., & Gleason, M. (2002). Relationship formation on the Internet: What’s the big attraction? Journal of Social Issues, 58, 9 –31.

Merten, D. E. (1996). Visibility and vulnerability: Responses to rejection by nonaggressive junior high school boys. Journal of Early Adolescence, 16, 5–26.

Moffit, T. E. (1993). Adolescence-limited and life-course-persistent anti- social behavior: A developmental taxonomy. Psychological Review, 100, 674 –701.

Muehlenkamp, J. J., & Gutierrez, P. M. (2004). An investigation of differences between self-injurious behavior and suicide attempts in a sample of adolescents. Suicide & Life-Threatening Behavior, 34, 12–24.

Nie, N. H., & Erbring, L. (2000). Internet and society: A preliminary report. Stanford, CA: Stanford Institute for the Quantitative Study of Society.

Nixon, M. K., Cloutier, P. F., & Aggarwal, S. (2002). Affect regulation and addictive aspects of repetitive self-injury in hospitalized adolescents. Journal of the American Academy of Child & Adolescent Psychiatry, 41, 1333–1341.

Petersen, A. C., Kennedy, R. E., & Sullivan, P. (1991). Coping with adolescence. In M. E. Colton & S. Gore (Eds.), Adolescent stress (pp. 93–110). New York: Aldine.

Reis, H. T., & Shaver, P. (1988). Intimacy as an interpersonal process. In S. Duck (Ed.), Handbook of personal relationships: Theory, research, and interventions (pp. 367–389). Chichester, UK: Wiley.

Roberts, D. F., Foehr, U. G., & Rideout, V. (2005). Generation M: Media in the lives of 8 –18 year-olds. Washington, DC: Henry J. Kaiser Family Foundation.

Rosen, P. M., & Walsh, B. W. (1989). Patterns of contagion in self mutilation epidemics. American Journal of Psychiatry, 146, 656 – 658.

Ross, R. R., & McKay, H. B. (1979). Self mutilation. Lexington, MA: Lexington Books.

Ross, S., & Heath, N. (2002). A study of the frequency of self-mutilation in a community sample of adolescents. Journal of Youth and Adoles- cence, 31, 66 –77.

Ross, S., & Heath, N. L. (2003). Two models of adolescent self-mutilation. Suicide & Life-Threatening Behavior, 33, 277–287.

Sansone, R. A., & Levitt, J. L. (2002). Self-harm behaviors among those with eating disorders: An overview. Eating Disorders, 10, 205–213.

Saunders, S. N., Resnick, M. D., Hoberman, H. M., & Blum, R. W. (1994). Formal help seeking behavior of adolescents identifying themselves as having mental health problems. Journal of American Academy of Child and Adolescent Psychiatry, 33, 718 –728.

Strong, M. (1998). A bright red scream: Self-mutilation and the language of pain. New York: Viking.

Subrahmanyam, K., Greenfield, P. M., & Tynes, B. (2004). Constructing sexuality and identity in an online teen chat room. Journal of Applied Developmental Psychology, 25, 651– 666.

Sullivan, H. S. (1953). The interpersonal theory of psychiatry. New York: Norton.

Suzuki, L. K., & Calzo, J. P. (2004). The search for peer advice in cyberspace: An examination of online teen bulletin boards about health and sexuality. Applied Developmental Psychology, 25, 685– 698.

Taiminen, T. J., Kallio-Soukainen, K., Nokso-Kovisto, H., Kaljonen, A., & Helenius, H. (1998). Contagion of deliberate self-harm among adoles- cent inpatients. Journal of the American Academy of Child & Adolescent Psychiatry, 37, 211–217.

Tantam, D., & Whittaker, J. (1993). Self-wounding and personality disor- der. In P. Tyrer & G. Stein (Eds.), Personality disorder reviewed (pp. 191–224). London: American Psychiatric Press.

Turkle, S. (1995). Life on the screen: Identity in the age of the Internet. New York: Simon & Schuster.

Tynes, B., Reynolds, L., & Greenfield, P. M. (2004). Adolescence, race, and ethnicity on the Internet: A comparison of discourse in monitored vs. unmonitored chat rooms. Journal of Applied Developmental Psychology, 25, 667– 684.

Walsh, B. W., & Rosen, P. M. (1988). Self-mutilation: Theory, research and treatment. New York: Guilford Press.

Wartella, E., Caplovitz, A. G., & Lee, J. H. (2004). From baby Einstein to leapfrog, from doom to The Sims, from instant messaging to Internet chat rooms: Public interest in the role of interactive media in children’s lives. Social Policy Report, 18, 3–16.

Welsh, P. (2004). Look students’ scars point to emotional pain. USA Today, p. 11a.

Whitlock, J. L., Eckenrode, J. J., & Silverman, D. (in press). Self-injurious behavior in a college population. Pediatrics.

Yates, T. M. (2004). The developmental psychopathology of self-injurious behavior: Compensatory regulation in posttraumatic adaptation. Clinical Psychological Review, 24, 35–74.

Ybarra, M. L., Alexander, C., & Mitchell, K. J. (2005). Depressive symp- tomatology, youth Internet use, and online interactions: A national survey. Journal of Adolescent Health, 36, 9 –18.

Young People and Self Harm: A National Inquiry. (2004). What do we already know? Prevalence, risk factors, and models of intervention. Retrieved June 29, 2004, from www.selfharmuk.org

Received April 1, 2005 Revision received December 15, 2005

Accepted December 16, 2005 �

417SPECIAL SECTION: VIRTUAL CUTTING EDGE