- Email: [email protected]
Helpful or Harmful? An Examination of Viewers' Responses to Nonsuicidal Self-Injury Videos on YouTube
Journal of Adolescent Health 51 (2012) 380 –385
www.jahonline.org Original article
Helpful or Harmful? An Examination of Viewers’ Responses to Nonsuicidal Self-Injury Videos on YouTube Stephen P. Lewis, Ph.D.a,*, Nancy L. Heath, Ph.D.b, Michael J. Sornberger, M.A.b, and Alexis E. Arbuthnott, B.A.H.a a b
Department of Psychology, University of Guelph, Guelph, Ontario, Canada Department of Educational and Counselling Psychology, McGill University, Quebec, Canada
Article history: Received August 11, 2011; Accepted January 18, 2012 Keywords: Self-injury; Self-harm; YouTube; Internet; Adolescent; Youth
A B S T R A C T
Purpose: To examine viewers’ comment responses to nonsuicidal self-injury (NSSI) YouTube videos to determine the potential risks (e.g., NSSI continuation) and benefits (e.g., recoveryoriented social support) of the videos. Methods: Viewers’ comments from the 100 most-viewed NSSI videos on YouTube were examined using two coding rubrics, one for the global nature of comments and one for recoveryoriented themes. Both rubrics were developed using an inductive (bottom-up) approach and had high coding inter-rater reliability (exceeding .80 in all cases). For the global nature of comments, 869 randomly selected comments were evaluated using the rubric, which included 8 coding categories and 22 subcategories. For the examination of recovery-oriented themes, self-disclosure comments (n ⫽ 377) were evaluated for nature of recovery statements. Results: Results revealed that the most frequent comments were self-disclosure comments in which individuals shared their own NSSI experiences (38.39%), followed by feedback for the video uploader, including admiration of the video quality (21.95%) or message (17.01%), and admiration for the uploader (15.40%) or encouragement to the video uploader (11.15%). Evaluation of the common self-disclosure comments for recovery-oriented content revealed that the majority did not mention recovery at all (42.89%) and indicated that they were still self-injuring (34.00%). Positive recovery statements were uncommon. Conclusions: Results suggest that viewers’ responses to videos may maintain the behavior (by sharing their own self-injury experiences) and rarely encourage or mention recovery. It is evident that sharing their own experience online is a strong motivator for viewers of NSSI YouTube videos. 䉷 2012 Society for Adolescent Health and Medicine. All rights reserved.
“I don’t know how it helps you to watch stuff like this ... makes things a whole lot worse for me.” “Wow, great video, and great job on all these SI videos. I’ve been in recovery from SI, and haven’t cut in 8 months. I came to YouTube tonight because I was feeling triggered, and your vids helped. I’m not going to cut today. Thanks.”
* Address correspondence to: Stephen P. Lewis, Ph.D., Department of Psychology, University of Guelph, Guelph, Ontario, Canada, N1G 2W1. E-mail address: [email protected] (S.P. Lewis).
IMPLICATIONS AND CONTRIBUTION
Self-injury videos on YouTube are prevalent; however, little is known about viewers’ responses to these videos, that is, whether they maintain self-injury or encourage recovery. This study provided the first empirical evidence that viewers’ responses to videos may maintain self-injury (by sharing their own selfinjury experiences) rather than encourage recovery.
The presentation of nonsuicidal self-injury (NSSI) on the Internet has received increasing empirical attention, as some material may normalize or perhaps maintain the behavior [1– 4]. Recently, the content of NSSI videos on YouTube was examined [3]. Results from this study received a very strong media response. Within a week of publication, it was featured in more than 300 printed news articles worldwide, in addition to radio and television segments [5]. Although this study brought awareness about NSSI videos and highlighted the content of the videos, much of the media attention focused on how viewers respond to the videos. Examining the video content alone is insufficient to understand the potential impact of
1054-139X/$ - see front matter 䉷 2012 Society for Adolescent Health and Medicine. All rights reserved. doi:10.1016/j.jadohealth.2012.01.013
S.P. Lewis et al. / Journal of Adolescent Health 51 (2012) 380 –385
the videos. It is important to investigate viewer responses to these videos, which may represent a better index of their impact. Consistent with concerns expressed by researchers, the manner by which viewers respond to these videos may help understand if NSSI is maintained. The examination of responses to videos may also have clinical implications, if the maintenance of NSSI indicated in the responses relates to aspects of NSSI recovery. Understanding how viewers respond to these videos through video comments represents the next step in this line of research. The current study examined the nature of viewer responses to 100 highly viewed NSSI videos on YouTube. NSSI refers to purposeful acts (e.g., cutting, carving, burning) in which individuals damage their own body tissue for nonsuicidal reasons (e.g., to regulate affect, punish oneself, communicate a need to others) [6]. Adolescents and young adults report the highest NSSI rates [7] and use the Internet more than any other age-group—primarily for social networking purposes [8,9], including online video sharing [9,10]. Moreover, youth who self-injure may engage in more online activity compared with youth who do not self-injure [11]. Specifically, the Internet may be used as a means of communication and as a way to receive acceptance, social support, and validation [2,4,12,13,14]. Despite these benefits (of the Internet), NSSI content on the Internet may be harmful. Researchers caution that some NSSI material may maintain NSSI for some youth and young adults [1– 4]. That is, online NSSI representations may focus on enacting the behavior and aspects of distress rather than hopeful messages about NSSI recovery; thus, repeated access to these messages may strengthen the attitude that little can be done to stop NSSI. Graphic NSSI material may also trigger NSSI urges for some individuals who self-injure [2,3]; in some cases, the presentation of NSSI online may glorify or sensationalize the behavior [3,14]. YouTube is the third most popular Web site worldwide and is frequently used by young people [10,15]. In a recent study, 100 highly viewed NSSI videos on YouTube were examined [3]. These videos were drawn from the over 5,000 NSSI videos available through YouTube at that time. Specifically, 50 videos with no live person (noncharacter videos) and 50 videos with at least one live person (i.e., character videos) were examined. Overall, these 100 videos had been viewed in excess of 2 million times and were favorably rated. Videos in this study primarily had informational and/or melancholic/hopeless messages, although approximately one-fourth were coded as hopeful. Many contained NSSI imagery. In particular, almost all videos with no live character had graphic NSSI photography. Although less common in videos with a live character, visual depictions of NSSI were still portrayed. Specifically, 14 of 50 character videos depicted live NSSI enactments. Given the number of NSSI videos on YouTube (i.e., ⬎5,000) and the concerns identified regarding video impact (i.e., NSSI continuation, triggering content), understanding how youth respond to these videos represents an important next step in this line of research. Beyond permitting uploading and viewing videos, YouTube enables users to respond to videos in the form of video comments. By examining these comment responses by viewers, it may be possible to learn more about the potential risks (e.g., NSSI continuation) and benefits (e.g., recovery-oriented social support) associated with certain NSSI videos. The current study built on previous work [3] and examined 869 randomly selected com-
381
ments from the top 100 NSSI videos on YouTube described in a previous study [3]. Methods Video and comment selection In December 2009, a search was conducted for the key words “self-injury” and “self-harm” using YouTube’s search function. Results were sorted in descending order by number of views. Videos were categorized into two groups: character videos, where the video focused on one or more live individuals, and noncharacter videos, which did not focus on individual people. The 50 most-viewed videos of each group were included in the sample of videos (n ⫽ 100). That is, the top 50 character videos were identified, as were the top 50 noncharacter videos. In total, there were 22,311 comments (mean ⫽ 232.41; standard deviation ⫽ 659.82) posted as responses to these videos. For each video view, there were approximately 107 video comments made. A random sampling of up to 10 user comments was collected for each of the 100 videos. If a video had fewer than 10 comments, all comments were included. If a video had more than 10, a random number generator was used to choose 10 comments randomly selected to enhance the representativeness of the sample. The total number of comments per video was tremendously variable, with the majority (62%) videos having 100 or fewer comments, many having between 100 and 500 (28%) comments, and a small portion (10%) having ⬎500. Replies to existing comments and comments left by the video’s creator were excluded from the sample; these exclusions were made to focus on direct reactions to the videos by other members of the YouTube community. In total, 869 comments were collected; 402 comments were included from character videos, and 467 from noncharacter videos. The researchers consulted their respective institutional ethics review boards and were informed that an ethics approval was not needed as the data used (i.e., YouTube publicly posted comments) are public domain. Data collection and coding procedure Two coding rubrics were developed using an approach similar to that used in other YouTube research [3]. The first rubric focused on the nature of comments posted in response to YouTube videos. To develop this first rubric, three independent coders used an inductive (bottom-up) approach in which comments not used in the final data set were repeatedly read to identify salient themes. The comments used to develop this rubric were drawn from the 100 target videos, but were not the 10 randomly selected comments that were in the sample of comments used in our final analysis. The comments used for the development of the rubric were similar in content to those used in the final analysis. To determine how comments could be coded, the coders independently and carefully read over the comments used to develop the rubric. Specifically, comments were read and potential codes emerged and were noted until saturation (i.e., until no new codes were identified). This led to the development of a preliminary set of codes. From here, codes were discussed among the coders and authors, and a final coding rubric was determined. In all, there were eight coding categories, each of which had several subcategories (n ⫽ 22); please refer to Table 1 for a list of comment categories and corresponding subcategories. Table 2 has example
382
S.P. Lewis et al. / Journal of Adolescent Health 51 (2012) 380 –385
Table 1 General coding rubric for the global characteristics of comments Comment categories
Code descriptions
1. Self-disclosure
a. Viewer shares personal experiences related to NSSI. b. Viewer shares personal experiences about the reason for engaging in NSSI. c. Comment contains an indication that the viewer has a comorbid mental illness (e.g., depression, anorexia, borderline personality disorder). a. Viewer admires the video quality (i.e., technical or visual aspects of the video). b. Viewer validates the video’s message. c. Viewer expresses personal validation, acceptance, or admiration toward the uploader. d. Viewer thanks the uploader. e. Viewer expresses pride toward the uploader. f. Viewer gives hostile, angry, or negative feedback. g. Viewer provides general words of encouragement. a. Viewer corrects information from the video. b. Viewer comments on factual information. c. Viewer asks a factual question. a. Viewer offers to help the uploader. b. Viewer offers to exchange messages with the uploader through YouTube. c. Viewer offers to connect with the uploader outside of YouTube, and provides information (e.g., personal e-mail address) to make this connection possible. d. Viewer asks for help. e. Viewer encourages the uploader to seek help. Viewer gives information or education on NSSI. Comment provides some indication that the viewer does not engage in NSSI. Comment has some explicit mention of the video possibly triggering NSSI thoughts or behaviors. The comment does not make sense or is irrelevant.
2. Feedback directed toward the person who uploaded the video
3. Factual information
4. Help-related comments
5. NSSI method and strategies 6. Viewer does not engage in NSSI 7. Triggers 8. Indecipherable NSSI ⫽ nonsuicidal self-injury.
comments for each general code. Coding categories were not mutually exclusive; thus, some video comments could have more than one category coded. After determining the final rubric, four independent coders coded another 100 comments not used in the final data set; coding inter-rater reliability was high and exceeded .80 in all cases. Because coding reliability was so high, coding was then split between two coders, who each coded one half of the comments to the top 100 NSSI videos on YouTube (n ⫽ 869). This approach is consistent with past research [3]. The next round of coding involved the development of a second rubric to further examine the most-frequent comment category determined in the previously mentioned first analysis, namely, selfdisclosure with emphasis on recovery. The decision to undertake a second rubric focusing on self-disclosure comments was made post hoc in light of the overwhelming predominance of this
category in the first analysis. The comments used to develop this recovery rubric comprised the self-disclosure comments determined from the first analysis. Similar to the previous coding rubric, an inductive approach was used to identify different recovery-oriented themes, which continued until no new themes emerged. Coding categories were then discussed, and a final coding rubric was determined. Resultant themes included wanting to recover, having made previous attempts to recover, presently seeking help or treatment to recover, mention that the commenter has recovered from NSSI, mention that the commenter has not yet recovered from NSSI, linking recovery to the video itself, and no mention of recovery. Codes were not mutually exclusive. Next, four independent coders were trained to reach .80 inter-rater reliability using 30 comments. Final coding
Table 2 General examples of comments posted on NSSI videos Category
Comment
1. Self-disclosure
“I feel ashamed when I cut I have scars and fresh cuts that I have done recently all up and down my wrist and arm. It’s hard for me to talk about it face to face with someone.” “omg i started crying i love it you are the best video maker you rock”
2. Feedback directed toward the person who uploaded the video 3. Factual information
4. Help-related comments
5. NSSI method and strategies
6. Viewer does not engage in NSSI
7. Triggers 8. Indecipherable
“I appreciated the documentary though it does make out that almost the only cause of self-injury is abuse, whereas this is not the case. Many self-injurers have been abused, but there are also many others that haven’t. But other than that, good movie” “Have you thought about admitting yourself? I mean to a mental hospital? Just so you can get serious medical help. If you don’t have a lot of other options. Hey guys if you self-harm it is really important that you seek assistance.” “I used to and still do scrub my cuts with really hot water and soap.it feels like your arm is on fire and they bleed again. but if you keep them under warm water it’ll stop it from bleeding and the cuts are less red. they fade easier. well at least for me that works” “um, hi im not a cutter . . . i haddent even heard about it before, until i friend told me he does. im worried for my friend... im not sure what to really do or say, can i get some advice. i want him to kno im on his side . . . im like there for him and stuff . . . but i just really would like to help him out or something . . . help?” “As a former self-injurer . . . this was quite a bit triggering . . . But I’m glad you made it and made it well.” “hello”
S.P. Lewis et al. / Journal of Adolescent Health 51 (2012) 380 –385
383
Theme subcategories
Shares personal experiences
38.43
Shares reasons for NSSI Comordities/mental illness Admiration of video quality Admiration of video message Personal validation Thanks for uploading video Pride in uploader Hostility Words of encouragement Correction Comment Question Offering to help Offering to message through YouTube Offering to connect outside YouTube Asking for help Encouraging uploader to seek help Information/education on NSSI Indication commenter does not NSSI Mentrion of triggering Comment is indecipherable
5.41 7.83 21.98 17.03 15.42 9.67 0.35 7.71 11.16 7.94 12.77 9.09 2.65 0.92 0.69 2.65 2.42 1.38 2.19 1.73 4.26 0
5
10
15 20 25 Comment percentages
30
35
40
Figure 1. Global characteristics for nonsuicidal self-injury YouTube video comments (n ⫽ 869) presented in percentages. Please note comment codes were not mutually exclusive.
reliability exceeded .80, and the coding rubric was split between two coders, who each coded one half of the self-disclosure comments (n ⫽ 377). Results General comments The content of comments posted about NSSI videos on YouTube varied (Figure 1). A few comments were indecipherable and could not be coded (4.25%). All comments in this study were examined as a function of video features; specifically, we examined whether there were differences in the types of comments made to videos based on the number of views per video (i.e., viewership), the rating associated with each video (as YouTube permits users to rate videos on a scale of 1–5), and type of video (i.e., character vs. noncharacter). However, no significant findings emerged. The most frequent comments were self-disclosure comments in which individuals shared their own NSSI experiences (38.39%). A small portion of self-disclosure comments indicated why the individual had self-injured (5.41%) and that the individual had a mental illness (7.83%). Following self-disclosure comments were those with feedback for the video uploader. Within this category, comments expressed admiration of video quality (21.95%), admiration for the video’s message (17.01%), and validation and admiration for the individual who uploaded the video (15.40%). Almost as common were comments offering words of encouragement to the video uploader (11.15%) or thanking the uploader for sharing his/her NSSI story (9.66%). In some cases, hostile comments were made about NSSI and/or the uploader (6.55%). Comments expressing pride toward the uploader (e.g., “this was very powerful. good job, i’m proud of you”) were rare (⬍1%).
Other comments were factual in nature and included those that asked questions to the YouTube community (8.97%), corrected NSSI facts that were presented in videos (7.93%), or presented NSSI facts themselves (7.93%). A small number of comments pertained to help seeking and provision. An equal number of comments asked for help (2.64%) or offered help to the uploader (2.64%). Related to this, a small number of comments encouraged the uploader to seek help (2.41%), and some offered to communicate with the uploader through YouTube or e-mail (1.61%). Comments about sharing NSSI methods and strategies (e.g., wound cleaning, concealing) were not common. Specifically, only 1.38% of comments fell in this category. Comments that mentioned triggering effects were also infrequent, with only a few comments explicitly mentioning that the video content was triggering and evoked NSSI urges (1.72%). Finally, comments explicitly indicating that the individuals posting the comment had not engaged in NSSI were uncommon and occurred just 2.18% of the time.
Self-disclosure comments To better understand the most common comment type, namely, those sharing personal NSSI experiences, a second coding rubric (described previously) was developed to identify features of NSSI recovery (Figure 2). Overall, the majority of comments in which people shared their own NSSI experiences did not mention recovery (42.89%); rather, they indicated that they had not recovered and were still self-injuring (34.00%). Much less common were comments indicating that the commenter had tried to seek treatment in the past or that he/she was presently seeking treatment (9.56%). Even less common were comments posted by individuals who indicated they had recovered (7.11%) or who expressed a desire to recover (3.78%).
S.P. Lewis et al. / Journal of Adolescent Health 51 (2012) 380 –385
Recovery categories
384
Viewer expresses desire to recover
4.51
Viewer has made previous efforts to recover
5.04
Viewer is presently seeking treatment to recover from NSSI
6.37
Viewer has recovered
8.49
Viewer has not yet recovered
40.58
Video helped viewer stop NSSI
1.33
Viewer believes video may help others stop NSSI
0.27
No mention of recovery
51.19
0
10
20
30
40
50
60
Comment Percentages Figure 2. Recovery characteristics for comments containing self-disclosure (n ⫽ 377) presented in percentages. Please note comment codes were not mutually exclusive.
Discussion Response comments to popular NSSI videos on YouTube primarily involve individuals sharing their NSSI experiences. A significant number of individuals who communicate about NSSI online do so by sharing their own NSSI story. Research in this area suggests that individuals may self-disclose NSSI online to receive validation and acceptance for these experiences [12] as well as to get peer support from others who self-injure [4]. These shared experiences, along with the nature of some of the other comments, which offer admiration and/or validation for the video uploader, the message of the video, or the uploader’s NSSI experience, may serve to maintain NSSI in some cases. These findings mirror concerns expressed by previous researchers examining NSSI material online [3,4], particularly the potential impact of NSSI videos [3]. Interestingly, only a small portion of comments mentioned that the videos triggered NSSI for the viewer; this is in contrast to widespread clinical assumption. This is consistent, however, with recent work examining NSSI Web sites, which were developed by those who self-injure and shared in NSSI e-communities [2]. These Web sites contained detailed descriptions of the authors’ NSSI experiences, and many had graphic NSSI images. The authors of this study found that although Web site authors reported being triggered by viewing NSSI images on other sites, these reports were infrequent (⬍10%). The extent to which individuals who access NSSI images online are triggered merits further empirical attention. Indeed, some research indicates that some individuals are triggered [2]; however, other research suggests that not all individuals are triggered by NSSI material [16]. A closer look at the shared NSSI experience comments indicated that only a small portion mentioned recovery. Most comments had no mention of recovery and/or indicated that individuals (who made the comment) were still self-injuring. The lack of recovery-based themes in many of the comments may also contribute to NSSI continuation if hopeful and recovery-oriented comments about NSSI are not presented. At the same time, research indicates that NSSI videos posted by youth and young adults are frequently viewed,
favorably rated, and elicit many video comments [3]. Results from the current study add to this and indicate that these comments comprise numerous NSSI disclosures. This coincides with other work indicating that youth and young adults who discuss NSSI online receive validation and may engage in this form of communication to be heard by, and connect with, others who share similar experiences [2,4,12]. Thus, to effectively reach youth and young adults who self-injure and who go online to communicate about it, the use of online video sharing and other modes of social networking merits consideration. For instance, an important step in the provision of helpful NSSI resources and support may be the use of online videos and other e-content to emphasize hope and recovery. However, this may need to occur in conjunction with allowing these individuals to feel heard and validated about having experienced NSSI. These approaches align with recent work indicating that youth prefer receiving help for NSSI through the Internet [17], and with findings indicating that YouTube may be a preferred communication medium for some youth who self-injure [3]. The nature of comments made to these videos (e.g., admiration for the video uploader) suggests that viewers may, in some cases, identify with some video uploaders and/or the NSSI experiences portrayed in the video. Thus, when providing youth and young adults with hopeful and recovery-based messages about NSSI (e.g., sharing NSSI recovery stories), the message may have more relevance and meaning if it comes from same-age peers who have not just experienced NSSI, but who have overcome it. Indeed, a large body of research indicates that the more similar the source of a message to the audience, the more effective the message [18]. Accordingly, this approach may be important not just on YouTube but also on other Web sites, as young adults communicate about NSSI across a variety of e-forums and are often actively seeking validation and acceptance [2,4,12]. Limitations Unfortunately, demographic data from those who commented on NSSI videos were unavailable. We surmise that these
S.P. Lewis et al. / Journal of Adolescent Health 51 (2012) 380 –385
individuals are largely youth and young adults, as indicated by research on Internet use among youth [9,10] and findings that these age-groups chiefly post these videos [3], but this cannot be confirmed. We also note that we selected the top 50 character videos and top 50 noncharacter videos to yield a total of 100. Accordingly, ⬎100 videos were used to generate an equal number of videos per group. Future research should examine the percentage of character and noncharacter videos in the top 100 videos overall to obtain an index of video popularity by type. Although not all video comments were examined, comments were randomly selected for each video to guard against selection bias, and the authors having read through all comments are confident that the selected comments were representative of the array of posted comments. Nevertheless, not all individuals who view NSSI videos post comments; thus, future work should examine how individuals who do not upload comments respond to NSSI videos. Examining video comments represents one way to assess video impact, but there may be other ways to assess this as well. For example, it is still unclear how videos impact individuals when they are first viewed (i.e., initial, in-the-moment impact) and when they are subsequently viewed (i.e., the long-term effects of videos). Although the coding rubric used in this study had good reliability, it arguably reflects the authors’ researchpractitioner perspective, and youth who communicate about NSSI on YouTube may not agree with some of the themes coded. Future work may be needed to understand how youth involved in NSSI communication on YouTube perceive these videos and their comments more directly. Implications Consistent with recent recommendations from researchers and clinicians [2– 4], health and mental professionals working with adolescents who self-injure need to assess the Internet behavior of these youth. Guidelines for assessing the nature and scope of online activity regarding NSSI have been offered and may be useful in clinical contexts [4,19]. Caregivers and school professionals can also benefit from learning about the potential impact of NSSI videos. To this end, open and informed discussions about NSSI, its representation online, and the potential risks involved are important. From a public health perspective, when adolescents and young adults use YouTube and other online forums to access NSSI material, it will be important to provide these youth with helpful resources in the corresponding search results. We are presently investigating these initiatives. Finally, future research is needed to identify ways to effectively reach youth who communicate about NSSI through the Internet. Given the salience of YouTube and other social networks among youth and young adults [8,9,10], efforts aimed to reach youth through these venues may be particularly relevant. It may be important for hopeful messages to come from youth who have experienced, but recovered from, NSSI, as this may resonate more with youth who presently struggle with NSSI. It will also be important to offer youth a safe and validating forum to share their experiences online.
385
Summary The representation of NSSI on the Internet has emerged as an important health issue for today’s youth and emerging adults [2– 4,13]. In particular, there has been concern that the content of NSSI videos on YouTube may maintain NSSI [3,14]. Findings from the current study provide support for these concerns and suggest that this may occur not just through video content but also through video comments. At the same time, the current findings suggest that the use of e-videos and other online media with hopeful messages from same-age peers may be needed to effectively reach individuals who self-injure and who go online to discuss it. Findings also indicate that youth may require a safe place to share their NSSI experiences with others. Thus, results from this study highlight an opportunity for researchers and clinicians to effectively reach youth who self-injure and who communicate about it through the Internet. References [1] Duggan JM, Heath NL, Lewis SP, Baxter A. An examination of the scope and nature of online activities among youth engaging in non-suicidal selfinjury. School Ment Health (in press). [2] Lewis SP, Baker TG. The possible risks of self-injury websites: A content analysis. Arch Suicide Res 2011;15:390 – 6. [3] Lewis SP, Heath NL, St Denis JM, Noble R. The scope of nonsuicidal selfinjury on youtube. Pediatrics 2011;127:e552–7. [4] Whitlock J, Eckenrode J, Silverman D. Self-injurious behaviors in a college population. Pediatrics 2006;117:1939 – 48. [5] Google News. Health section. Available at: http://news.google.com/news/ section?pz⫽1&cf⫽all&ned⫽ca&topic⫽m. Accessed February 25, 2011. [6] Nock MK, Favazza AR. Nonsuicidal self-injury: Definition and classification. In: Nock MK, ed. Understanding Nonsuicidal Self-Injury: Origins, Assessment, and Treatment. Washington, DC: American Psychological Association, 2009:3–18. [7] Rodham K, Hawton K. Epidemiology and phenomenology of nonsuicidal self-injury. In: Nock MK, ed. Understanding Nonsuicidal Self-Injury: Origins, Assessment, and Treatment. Washington, DC: American Psychological Association, 2009:37– 62. [8] Lenhart A. Teens and social media: An overview. Washington, DC: Pew Internet and American Life, 2009. [9] Rideout V, Foehr UG, Roberts DF. Generation M2: Media in the lives of 8- to 18-year-olds. Menlo Park, CA: Kaiser Family Foundation, 2010. [10] Nielson Wire. June 2011: Top U.S. online destinations for video. Available at: http://blog.nielsen.com/nielsenwire/online_mobile/june-2011-top-u-s-onlinedestinations-for-video/. Accessed July 27, 2010. [11] Ybarra ML, Mitchell KJ. Prevalence and frequency of internet harassment instigation: Implications for adolescent health. J Adolesc Health 2007;41: 189 –95. [12] Lewis SP, Rosenrot SR, Messner MA. Seeking validation in unlikely places: The nature of online questions about non-suicidal selfinjury. Arch Suicide Res (In press). [13] Rodham K, Gavin J, Miles M. I hear, I listen and I care: A qualitative investigation into the function of a self-harm message board. Suicide Life Threat Behav 2007;37:422–30. [14] Whitlock JL, Purington A, Gershkovich M. Influence of the media on self injurious behavior. In: Nock MK, ed. Understanding Non-Suicidal SelfInjury: Current Science and Practice. Washington, DC: American Psychological Association, 2009:139 –56. [15] Alexa. Google.com’s worldwide traffic rank. Available at: http://www. alexa.com/siteinfo/google.com. Accessed July 17, 2011. [16] Murray CD, Fox J. Do internet self-harm discussion groups alleviate or exacerbate self-harming behaviour? Adv Mental Health 2006;5:225–33. [17] Hasking P, Martin G, Berger E. How do we help those who self-injure? An examination of what young people think. Paper presented at: Chicago, Illinois: International Society for the Study of Self-Injury Conference, 2010. [18] Pornpitakpan C. The persuasiveness of source credibility: Critical review of five decades of research. J Appl Soc Psychol 2004;32:243– 81. [19] Lewis SP, Heath NL, Michal NJ, Duggan JM. Non-suicidal self-injury, youth and the Internet: What mental health professionals need to know. Child Adol Psych Mental Health. (In press).
www.jahonline.org Original article
Helpful or Harmful? An Examination of Viewers’ Responses to Nonsuicidal Self-Injury Videos on YouTube Stephen P. Lewis, Ph.D.a,*, Nancy L. Heath, Ph.D.b, Michael J. Sornberger, M.A.b, and Alexis E. Arbuthnott, B.A.H.a a b
Department of Psychology, University of Guelph, Guelph, Ontario, Canada Department of Educational and Counselling Psychology, McGill University, Quebec, Canada
Article history: Received August 11, 2011; Accepted January 18, 2012 Keywords: Self-injury; Self-harm; YouTube; Internet; Adolescent; Youth
A B S T R A C T
Purpose: To examine viewers’ comment responses to nonsuicidal self-injury (NSSI) YouTube videos to determine the potential risks (e.g., NSSI continuation) and benefits (e.g., recoveryoriented social support) of the videos. Methods: Viewers’ comments from the 100 most-viewed NSSI videos on YouTube were examined using two coding rubrics, one for the global nature of comments and one for recoveryoriented themes. Both rubrics were developed using an inductive (bottom-up) approach and had high coding inter-rater reliability (exceeding .80 in all cases). For the global nature of comments, 869 randomly selected comments were evaluated using the rubric, which included 8 coding categories and 22 subcategories. For the examination of recovery-oriented themes, self-disclosure comments (n ⫽ 377) were evaluated for nature of recovery statements. Results: Results revealed that the most frequent comments were self-disclosure comments in which individuals shared their own NSSI experiences (38.39%), followed by feedback for the video uploader, including admiration of the video quality (21.95%) or message (17.01%), and admiration for the uploader (15.40%) or encouragement to the video uploader (11.15%). Evaluation of the common self-disclosure comments for recovery-oriented content revealed that the majority did not mention recovery at all (42.89%) and indicated that they were still self-injuring (34.00%). Positive recovery statements were uncommon. Conclusions: Results suggest that viewers’ responses to videos may maintain the behavior (by sharing their own self-injury experiences) and rarely encourage or mention recovery. It is evident that sharing their own experience online is a strong motivator for viewers of NSSI YouTube videos. 䉷 2012 Society for Adolescent Health and Medicine. All rights reserved.
“I don’t know how it helps you to watch stuff like this ... makes things a whole lot worse for me.” “Wow, great video, and great job on all these SI videos. I’ve been in recovery from SI, and haven’t cut in 8 months. I came to YouTube tonight because I was feeling triggered, and your vids helped. I’m not going to cut today. Thanks.”
* Address correspondence to: Stephen P. Lewis, Ph.D., Department of Psychology, University of Guelph, Guelph, Ontario, Canada, N1G 2W1. E-mail address: [email protected] (S.P. Lewis).
IMPLICATIONS AND CONTRIBUTION
Self-injury videos on YouTube are prevalent; however, little is known about viewers’ responses to these videos, that is, whether they maintain self-injury or encourage recovery. This study provided the first empirical evidence that viewers’ responses to videos may maintain self-injury (by sharing their own selfinjury experiences) rather than encourage recovery.
The presentation of nonsuicidal self-injury (NSSI) on the Internet has received increasing empirical attention, as some material may normalize or perhaps maintain the behavior [1– 4]. Recently, the content of NSSI videos on YouTube was examined [3]. Results from this study received a very strong media response. Within a week of publication, it was featured in more than 300 printed news articles worldwide, in addition to radio and television segments [5]. Although this study brought awareness about NSSI videos and highlighted the content of the videos, much of the media attention focused on how viewers respond to the videos. Examining the video content alone is insufficient to understand the potential impact of
1054-139X/$ - see front matter 䉷 2012 Society for Adolescent Health and Medicine. All rights reserved. doi:10.1016/j.jadohealth.2012.01.013
S.P. Lewis et al. / Journal of Adolescent Health 51 (2012) 380 –385
the videos. It is important to investigate viewer responses to these videos, which may represent a better index of their impact. Consistent with concerns expressed by researchers, the manner by which viewers respond to these videos may help understand if NSSI is maintained. The examination of responses to videos may also have clinical implications, if the maintenance of NSSI indicated in the responses relates to aspects of NSSI recovery. Understanding how viewers respond to these videos through video comments represents the next step in this line of research. The current study examined the nature of viewer responses to 100 highly viewed NSSI videos on YouTube. NSSI refers to purposeful acts (e.g., cutting, carving, burning) in which individuals damage their own body tissue for nonsuicidal reasons (e.g., to regulate affect, punish oneself, communicate a need to others) [6]. Adolescents and young adults report the highest NSSI rates [7] and use the Internet more than any other age-group—primarily for social networking purposes [8,9], including online video sharing [9,10]. Moreover, youth who self-injure may engage in more online activity compared with youth who do not self-injure [11]. Specifically, the Internet may be used as a means of communication and as a way to receive acceptance, social support, and validation [2,4,12,13,14]. Despite these benefits (of the Internet), NSSI content on the Internet may be harmful. Researchers caution that some NSSI material may maintain NSSI for some youth and young adults [1– 4]. That is, online NSSI representations may focus on enacting the behavior and aspects of distress rather than hopeful messages about NSSI recovery; thus, repeated access to these messages may strengthen the attitude that little can be done to stop NSSI. Graphic NSSI material may also trigger NSSI urges for some individuals who self-injure [2,3]; in some cases, the presentation of NSSI online may glorify or sensationalize the behavior [3,14]. YouTube is the third most popular Web site worldwide and is frequently used by young people [10,15]. In a recent study, 100 highly viewed NSSI videos on YouTube were examined [3]. These videos were drawn from the over 5,000 NSSI videos available through YouTube at that time. Specifically, 50 videos with no live person (noncharacter videos) and 50 videos with at least one live person (i.e., character videos) were examined. Overall, these 100 videos had been viewed in excess of 2 million times and were favorably rated. Videos in this study primarily had informational and/or melancholic/hopeless messages, although approximately one-fourth were coded as hopeful. Many contained NSSI imagery. In particular, almost all videos with no live character had graphic NSSI photography. Although less common in videos with a live character, visual depictions of NSSI were still portrayed. Specifically, 14 of 50 character videos depicted live NSSI enactments. Given the number of NSSI videos on YouTube (i.e., ⬎5,000) and the concerns identified regarding video impact (i.e., NSSI continuation, triggering content), understanding how youth respond to these videos represents an important next step in this line of research. Beyond permitting uploading and viewing videos, YouTube enables users to respond to videos in the form of video comments. By examining these comment responses by viewers, it may be possible to learn more about the potential risks (e.g., NSSI continuation) and benefits (e.g., recovery-oriented social support) associated with certain NSSI videos. The current study built on previous work [3] and examined 869 randomly selected com-
381
ments from the top 100 NSSI videos on YouTube described in a previous study [3]. Methods Video and comment selection In December 2009, a search was conducted for the key words “self-injury” and “self-harm” using YouTube’s search function. Results were sorted in descending order by number of views. Videos were categorized into two groups: character videos, where the video focused on one or more live individuals, and noncharacter videos, which did not focus on individual people. The 50 most-viewed videos of each group were included in the sample of videos (n ⫽ 100). That is, the top 50 character videos were identified, as were the top 50 noncharacter videos. In total, there were 22,311 comments (mean ⫽ 232.41; standard deviation ⫽ 659.82) posted as responses to these videos. For each video view, there were approximately 107 video comments made. A random sampling of up to 10 user comments was collected for each of the 100 videos. If a video had fewer than 10 comments, all comments were included. If a video had more than 10, a random number generator was used to choose 10 comments randomly selected to enhance the representativeness of the sample. The total number of comments per video was tremendously variable, with the majority (62%) videos having 100 or fewer comments, many having between 100 and 500 (28%) comments, and a small portion (10%) having ⬎500. Replies to existing comments and comments left by the video’s creator were excluded from the sample; these exclusions were made to focus on direct reactions to the videos by other members of the YouTube community. In total, 869 comments were collected; 402 comments were included from character videos, and 467 from noncharacter videos. The researchers consulted their respective institutional ethics review boards and were informed that an ethics approval was not needed as the data used (i.e., YouTube publicly posted comments) are public domain. Data collection and coding procedure Two coding rubrics were developed using an approach similar to that used in other YouTube research [3]. The first rubric focused on the nature of comments posted in response to YouTube videos. To develop this first rubric, three independent coders used an inductive (bottom-up) approach in which comments not used in the final data set were repeatedly read to identify salient themes. The comments used to develop this rubric were drawn from the 100 target videos, but were not the 10 randomly selected comments that were in the sample of comments used in our final analysis. The comments used for the development of the rubric were similar in content to those used in the final analysis. To determine how comments could be coded, the coders independently and carefully read over the comments used to develop the rubric. Specifically, comments were read and potential codes emerged and were noted until saturation (i.e., until no new codes were identified). This led to the development of a preliminary set of codes. From here, codes were discussed among the coders and authors, and a final coding rubric was determined. In all, there were eight coding categories, each of which had several subcategories (n ⫽ 22); please refer to Table 1 for a list of comment categories and corresponding subcategories. Table 2 has example
382
S.P. Lewis et al. / Journal of Adolescent Health 51 (2012) 380 –385
Table 1 General coding rubric for the global characteristics of comments Comment categories
Code descriptions
1. Self-disclosure
a. Viewer shares personal experiences related to NSSI. b. Viewer shares personal experiences about the reason for engaging in NSSI. c. Comment contains an indication that the viewer has a comorbid mental illness (e.g., depression, anorexia, borderline personality disorder). a. Viewer admires the video quality (i.e., technical or visual aspects of the video). b. Viewer validates the video’s message. c. Viewer expresses personal validation, acceptance, or admiration toward the uploader. d. Viewer thanks the uploader. e. Viewer expresses pride toward the uploader. f. Viewer gives hostile, angry, or negative feedback. g. Viewer provides general words of encouragement. a. Viewer corrects information from the video. b. Viewer comments on factual information. c. Viewer asks a factual question. a. Viewer offers to help the uploader. b. Viewer offers to exchange messages with the uploader through YouTube. c. Viewer offers to connect with the uploader outside of YouTube, and provides information (e.g., personal e-mail address) to make this connection possible. d. Viewer asks for help. e. Viewer encourages the uploader to seek help. Viewer gives information or education on NSSI. Comment provides some indication that the viewer does not engage in NSSI. Comment has some explicit mention of the video possibly triggering NSSI thoughts or behaviors. The comment does not make sense or is irrelevant.
2. Feedback directed toward the person who uploaded the video
3. Factual information
4. Help-related comments
5. NSSI method and strategies 6. Viewer does not engage in NSSI 7. Triggers 8. Indecipherable NSSI ⫽ nonsuicidal self-injury.
comments for each general code. Coding categories were not mutually exclusive; thus, some video comments could have more than one category coded. After determining the final rubric, four independent coders coded another 100 comments not used in the final data set; coding inter-rater reliability was high and exceeded .80 in all cases. Because coding reliability was so high, coding was then split between two coders, who each coded one half of the comments to the top 100 NSSI videos on YouTube (n ⫽ 869). This approach is consistent with past research [3]. The next round of coding involved the development of a second rubric to further examine the most-frequent comment category determined in the previously mentioned first analysis, namely, selfdisclosure with emphasis on recovery. The decision to undertake a second rubric focusing on self-disclosure comments was made post hoc in light of the overwhelming predominance of this
category in the first analysis. The comments used to develop this recovery rubric comprised the self-disclosure comments determined from the first analysis. Similar to the previous coding rubric, an inductive approach was used to identify different recovery-oriented themes, which continued until no new themes emerged. Coding categories were then discussed, and a final coding rubric was determined. Resultant themes included wanting to recover, having made previous attempts to recover, presently seeking help or treatment to recover, mention that the commenter has recovered from NSSI, mention that the commenter has not yet recovered from NSSI, linking recovery to the video itself, and no mention of recovery. Codes were not mutually exclusive. Next, four independent coders were trained to reach .80 inter-rater reliability using 30 comments. Final coding
Table 2 General examples of comments posted on NSSI videos Category
Comment
1. Self-disclosure
“I feel ashamed when I cut I have scars and fresh cuts that I have done recently all up and down my wrist and arm. It’s hard for me to talk about it face to face with someone.” “omg i started crying i love it you are the best video maker you rock”
2. Feedback directed toward the person who uploaded the video 3. Factual information
4. Help-related comments
5. NSSI method and strategies
6. Viewer does not engage in NSSI
7. Triggers 8. Indecipherable
“I appreciated the documentary though it does make out that almost the only cause of self-injury is abuse, whereas this is not the case. Many self-injurers have been abused, but there are also many others that haven’t. But other than that, good movie” “Have you thought about admitting yourself? I mean to a mental hospital? Just so you can get serious medical help. If you don’t have a lot of other options. Hey guys if you self-harm it is really important that you seek assistance.” “I used to and still do scrub my cuts with really hot water and soap.it feels like your arm is on fire and they bleed again. but if you keep them under warm water it’ll stop it from bleeding and the cuts are less red. they fade easier. well at least for me that works” “um, hi im not a cutter . . . i haddent even heard about it before, until i friend told me he does. im worried for my friend... im not sure what to really do or say, can i get some advice. i want him to kno im on his side . . . im like there for him and stuff . . . but i just really would like to help him out or something . . . help?” “As a former self-injurer . . . this was quite a bit triggering . . . But I’m glad you made it and made it well.” “hello”
S.P. Lewis et al. / Journal of Adolescent Health 51 (2012) 380 –385
383
Theme subcategories
Shares personal experiences
38.43
Shares reasons for NSSI Comordities/mental illness Admiration of video quality Admiration of video message Personal validation Thanks for uploading video Pride in uploader Hostility Words of encouragement Correction Comment Question Offering to help Offering to message through YouTube Offering to connect outside YouTube Asking for help Encouraging uploader to seek help Information/education on NSSI Indication commenter does not NSSI Mentrion of triggering Comment is indecipherable
5.41 7.83 21.98 17.03 15.42 9.67 0.35 7.71 11.16 7.94 12.77 9.09 2.65 0.92 0.69 2.65 2.42 1.38 2.19 1.73 4.26 0
5
10
15 20 25 Comment percentages
30
35
40
Figure 1. Global characteristics for nonsuicidal self-injury YouTube video comments (n ⫽ 869) presented in percentages. Please note comment codes were not mutually exclusive.
reliability exceeded .80, and the coding rubric was split between two coders, who each coded one half of the self-disclosure comments (n ⫽ 377). Results General comments The content of comments posted about NSSI videos on YouTube varied (Figure 1). A few comments were indecipherable and could not be coded (4.25%). All comments in this study were examined as a function of video features; specifically, we examined whether there were differences in the types of comments made to videos based on the number of views per video (i.e., viewership), the rating associated with each video (as YouTube permits users to rate videos on a scale of 1–5), and type of video (i.e., character vs. noncharacter). However, no significant findings emerged. The most frequent comments were self-disclosure comments in which individuals shared their own NSSI experiences (38.39%). A small portion of self-disclosure comments indicated why the individual had self-injured (5.41%) and that the individual had a mental illness (7.83%). Following self-disclosure comments were those with feedback for the video uploader. Within this category, comments expressed admiration of video quality (21.95%), admiration for the video’s message (17.01%), and validation and admiration for the individual who uploaded the video (15.40%). Almost as common were comments offering words of encouragement to the video uploader (11.15%) or thanking the uploader for sharing his/her NSSI story (9.66%). In some cases, hostile comments were made about NSSI and/or the uploader (6.55%). Comments expressing pride toward the uploader (e.g., “this was very powerful. good job, i’m proud of you”) were rare (⬍1%).
Other comments were factual in nature and included those that asked questions to the YouTube community (8.97%), corrected NSSI facts that were presented in videos (7.93%), or presented NSSI facts themselves (7.93%). A small number of comments pertained to help seeking and provision. An equal number of comments asked for help (2.64%) or offered help to the uploader (2.64%). Related to this, a small number of comments encouraged the uploader to seek help (2.41%), and some offered to communicate with the uploader through YouTube or e-mail (1.61%). Comments about sharing NSSI methods and strategies (e.g., wound cleaning, concealing) were not common. Specifically, only 1.38% of comments fell in this category. Comments that mentioned triggering effects were also infrequent, with only a few comments explicitly mentioning that the video content was triggering and evoked NSSI urges (1.72%). Finally, comments explicitly indicating that the individuals posting the comment had not engaged in NSSI were uncommon and occurred just 2.18% of the time.
Self-disclosure comments To better understand the most common comment type, namely, those sharing personal NSSI experiences, a second coding rubric (described previously) was developed to identify features of NSSI recovery (Figure 2). Overall, the majority of comments in which people shared their own NSSI experiences did not mention recovery (42.89%); rather, they indicated that they had not recovered and were still self-injuring (34.00%). Much less common were comments indicating that the commenter had tried to seek treatment in the past or that he/she was presently seeking treatment (9.56%). Even less common were comments posted by individuals who indicated they had recovered (7.11%) or who expressed a desire to recover (3.78%).
S.P. Lewis et al. / Journal of Adolescent Health 51 (2012) 380 –385
Recovery categories
384
Viewer expresses desire to recover
4.51
Viewer has made previous efforts to recover
5.04
Viewer is presently seeking treatment to recover from NSSI
6.37
Viewer has recovered
8.49
Viewer has not yet recovered
40.58
Video helped viewer stop NSSI
1.33
Viewer believes video may help others stop NSSI
0.27
No mention of recovery
51.19
0
10
20
30
40
50
60
Comment Percentages Figure 2. Recovery characteristics for comments containing self-disclosure (n ⫽ 377) presented in percentages. Please note comment codes were not mutually exclusive.
Discussion Response comments to popular NSSI videos on YouTube primarily involve individuals sharing their NSSI experiences. A significant number of individuals who communicate about NSSI online do so by sharing their own NSSI story. Research in this area suggests that individuals may self-disclose NSSI online to receive validation and acceptance for these experiences [12] as well as to get peer support from others who self-injure [4]. These shared experiences, along with the nature of some of the other comments, which offer admiration and/or validation for the video uploader, the message of the video, or the uploader’s NSSI experience, may serve to maintain NSSI in some cases. These findings mirror concerns expressed by previous researchers examining NSSI material online [3,4], particularly the potential impact of NSSI videos [3]. Interestingly, only a small portion of comments mentioned that the videos triggered NSSI for the viewer; this is in contrast to widespread clinical assumption. This is consistent, however, with recent work examining NSSI Web sites, which were developed by those who self-injure and shared in NSSI e-communities [2]. These Web sites contained detailed descriptions of the authors’ NSSI experiences, and many had graphic NSSI images. The authors of this study found that although Web site authors reported being triggered by viewing NSSI images on other sites, these reports were infrequent (⬍10%). The extent to which individuals who access NSSI images online are triggered merits further empirical attention. Indeed, some research indicates that some individuals are triggered [2]; however, other research suggests that not all individuals are triggered by NSSI material [16]. A closer look at the shared NSSI experience comments indicated that only a small portion mentioned recovery. Most comments had no mention of recovery and/or indicated that individuals (who made the comment) were still self-injuring. The lack of recovery-based themes in many of the comments may also contribute to NSSI continuation if hopeful and recovery-oriented comments about NSSI are not presented. At the same time, research indicates that NSSI videos posted by youth and young adults are frequently viewed,
favorably rated, and elicit many video comments [3]. Results from the current study add to this and indicate that these comments comprise numerous NSSI disclosures. This coincides with other work indicating that youth and young adults who discuss NSSI online receive validation and may engage in this form of communication to be heard by, and connect with, others who share similar experiences [2,4,12]. Thus, to effectively reach youth and young adults who self-injure and who go online to communicate about it, the use of online video sharing and other modes of social networking merits consideration. For instance, an important step in the provision of helpful NSSI resources and support may be the use of online videos and other e-content to emphasize hope and recovery. However, this may need to occur in conjunction with allowing these individuals to feel heard and validated about having experienced NSSI. These approaches align with recent work indicating that youth prefer receiving help for NSSI through the Internet [17], and with findings indicating that YouTube may be a preferred communication medium for some youth who self-injure [3]. The nature of comments made to these videos (e.g., admiration for the video uploader) suggests that viewers may, in some cases, identify with some video uploaders and/or the NSSI experiences portrayed in the video. Thus, when providing youth and young adults with hopeful and recovery-based messages about NSSI (e.g., sharing NSSI recovery stories), the message may have more relevance and meaning if it comes from same-age peers who have not just experienced NSSI, but who have overcome it. Indeed, a large body of research indicates that the more similar the source of a message to the audience, the more effective the message [18]. Accordingly, this approach may be important not just on YouTube but also on other Web sites, as young adults communicate about NSSI across a variety of e-forums and are often actively seeking validation and acceptance [2,4,12]. Limitations Unfortunately, demographic data from those who commented on NSSI videos were unavailable. We surmise that these
S.P. Lewis et al. / Journal of Adolescent Health 51 (2012) 380 –385
individuals are largely youth and young adults, as indicated by research on Internet use among youth [9,10] and findings that these age-groups chiefly post these videos [3], but this cannot be confirmed. We also note that we selected the top 50 character videos and top 50 noncharacter videos to yield a total of 100. Accordingly, ⬎100 videos were used to generate an equal number of videos per group. Future research should examine the percentage of character and noncharacter videos in the top 100 videos overall to obtain an index of video popularity by type. Although not all video comments were examined, comments were randomly selected for each video to guard against selection bias, and the authors having read through all comments are confident that the selected comments were representative of the array of posted comments. Nevertheless, not all individuals who view NSSI videos post comments; thus, future work should examine how individuals who do not upload comments respond to NSSI videos. Examining video comments represents one way to assess video impact, but there may be other ways to assess this as well. For example, it is still unclear how videos impact individuals when they are first viewed (i.e., initial, in-the-moment impact) and when they are subsequently viewed (i.e., the long-term effects of videos). Although the coding rubric used in this study had good reliability, it arguably reflects the authors’ researchpractitioner perspective, and youth who communicate about NSSI on YouTube may not agree with some of the themes coded. Future work may be needed to understand how youth involved in NSSI communication on YouTube perceive these videos and their comments more directly. Implications Consistent with recent recommendations from researchers and clinicians [2– 4], health and mental professionals working with adolescents who self-injure need to assess the Internet behavior of these youth. Guidelines for assessing the nature and scope of online activity regarding NSSI have been offered and may be useful in clinical contexts [4,19]. Caregivers and school professionals can also benefit from learning about the potential impact of NSSI videos. To this end, open and informed discussions about NSSI, its representation online, and the potential risks involved are important. From a public health perspective, when adolescents and young adults use YouTube and other online forums to access NSSI material, it will be important to provide these youth with helpful resources in the corresponding search results. We are presently investigating these initiatives. Finally, future research is needed to identify ways to effectively reach youth who communicate about NSSI through the Internet. Given the salience of YouTube and other social networks among youth and young adults [8,9,10], efforts aimed to reach youth through these venues may be particularly relevant. It may be important for hopeful messages to come from youth who have experienced, but recovered from, NSSI, as this may resonate more with youth who presently struggle with NSSI. It will also be important to offer youth a safe and validating forum to share their experiences online.
385
Summary The representation of NSSI on the Internet has emerged as an important health issue for today’s youth and emerging adults [2– 4,13]. In particular, there has been concern that the content of NSSI videos on YouTube may maintain NSSI [3,14]. Findings from the current study provide support for these concerns and suggest that this may occur not just through video content but also through video comments. At the same time, the current findings suggest that the use of e-videos and other online media with hopeful messages from same-age peers may be needed to effectively reach individuals who self-injure and who go online to discuss it. Findings also indicate that youth may require a safe place to share their NSSI experiences with others. Thus, results from this study highlight an opportunity for researchers and clinicians to effectively reach youth who self-injure and who communicate about it through the Internet. References [1] Duggan JM, Heath NL, Lewis SP, Baxter A. An examination of the scope and nature of online activities among youth engaging in non-suicidal selfinjury. School Ment Health (in press). [2] Lewis SP, Baker TG. The possible risks of self-injury websites: A content analysis. Arch Suicide Res 2011;15:390 – 6. [3] Lewis SP, Heath NL, St Denis JM, Noble R. The scope of nonsuicidal selfinjury on youtube. Pediatrics 2011;127:e552–7. [4] Whitlock J, Eckenrode J, Silverman D. Self-injurious behaviors in a college population. Pediatrics 2006;117:1939 – 48. [5] Google News. Health section. Available at: http://news.google.com/news/ section?pz⫽1&cf⫽all&ned⫽ca&topic⫽m. Accessed February 25, 2011. [6] Nock MK, Favazza AR. Nonsuicidal self-injury: Definition and classification. In: Nock MK, ed. Understanding Nonsuicidal Self-Injury: Origins, Assessment, and Treatment. Washington, DC: American Psychological Association, 2009:3–18. [7] Rodham K, Hawton K. Epidemiology and phenomenology of nonsuicidal self-injury. In: Nock MK, ed. Understanding Nonsuicidal Self-Injury: Origins, Assessment, and Treatment. Washington, DC: American Psychological Association, 2009:37– 62. [8] Lenhart A. Teens and social media: An overview. Washington, DC: Pew Internet and American Life, 2009. [9] Rideout V, Foehr UG, Roberts DF. Generation M2: Media in the lives of 8- to 18-year-olds. Menlo Park, CA: Kaiser Family Foundation, 2010. [10] Nielson Wire. June 2011: Top U.S. online destinations for video. Available at: http://blog.nielsen.com/nielsenwire/online_mobile/june-2011-top-u-s-onlinedestinations-for-video/. Accessed July 27, 2010. [11] Ybarra ML, Mitchell KJ. Prevalence and frequency of internet harassment instigation: Implications for adolescent health. J Adolesc Health 2007;41: 189 –95. [12] Lewis SP, Rosenrot SR, Messner MA. Seeking validation in unlikely places: The nature of online questions about non-suicidal selfinjury. Arch Suicide Res (In press). [13] Rodham K, Gavin J, Miles M. I hear, I listen and I care: A qualitative investigation into the function of a self-harm message board. Suicide Life Threat Behav 2007;37:422–30. [14] Whitlock JL, Purington A, Gershkovich M. Influence of the media on self injurious behavior. In: Nock MK, ed. Understanding Non-Suicidal SelfInjury: Current Science and Practice. Washington, DC: American Psychological Association, 2009:139 –56. [15] Alexa. Google.com’s worldwide traffic rank. Available at: http://www. alexa.com/siteinfo/google.com. Accessed July 17, 2011. [16] Murray CD, Fox J. Do internet self-harm discussion groups alleviate or exacerbate self-harming behaviour? Adv Mental Health 2006;5:225–33. [17] Hasking P, Martin G, Berger E. How do we help those who self-injure? An examination of what young people think. Paper presented at: Chicago, Illinois: International Society for the Study of Self-Injury Conference, 2010. [18] Pornpitakpan C. The persuasiveness of source credibility: Critical review of five decades of research. J Appl Soc Psychol 2004;32:243– 81. [19] Lewis SP, Heath NL, Michal NJ, Duggan JM. Non-suicidal self-injury, youth and the Internet: What mental health professionals need to know. Child Adol Psych Mental Health. (In press).