The Virtual Cutting Edge The Internet and Adolescent SelfInjury Janis L - PDF document

increaseindisclosuresofself-injurytonationalchildren’shelplinesoverthe5yearsbeforethestudy,notinga65%increaseinthelast2years(YoungPeopleandSelfHarm:ANationalInquiry,Thereasonsforthisapparentincreaseareunclear.Althoughgeneralawarenessofself-injurymayleadtoincreasedwillingnesstodisclosethebehavior,itmayalsoreflectatrueincreaseinincidence.Socialcontagionhasbeenidentifiedbysome(Yates,2004;Rosen&Walsh,1989),becauseself-injuryfollowsepidemic-likepatternsininstitutionalsettingssuchashospitalsanddetentionfacilities(Matthews,1968;Ross&McKay,1979;Taiminen,Kallio-Soukainen,Nokso-Koivisto,Kaljonen,&Hele-nius,1998).Thepossibilitythatself-injuryiscommunicablemayreflectapatternsimilartowhatBrumberg(1992)arguedwithregardtothespreadofanorexianervosainthe1980s,whenheightenedculturalvisibilitythroughthemassmediarenderedanorexianervosaanavailableemotionaloutletforindividualswithreceptivepredispositions.Althoughnosingleself-injurerprofilehasemerged,thereisgeneralconsensusthatself-injuryismostcommonamongadoles-cents.Self-injuriousbehaviormayparallelotherproblembehav-iors,whichbeginduringearlyadolescence,peakduringmid-tolateadolescence,andthendeclineinadulthood(Briere&Gil,1998;Favazza,1999;Favazza&Conterio,1989).However,thereissomeevidencethatself-injuriousbehaviormayfollowtwodistinctpatternssimilartothelifecourse-persistentandadolescence-limitedtrajectoriesevidentinantisocialbehavior(Moffit,1993).Oneofthesebeginsinearlychildhoodandpersiststhroughadulthood.Theotherfollowsatypicaladolescence-limitedcoursethatemergesinearlyadolescenceanddeclinesinlateadolescenceorearlyadulthood(Dubo,Zanarini,Lewis,&Wil-liams,1997;Nixon,Cloutier,&Aggarwal,2002).Althoughsomeresearchfindsfemalesmorelikelytoinjurethemselvesthanmales(Conterio&Lader,1998;Favazza,1999),otherstudiessuggestthatthegendergapmaybenarrower(Briere&Gil,1998;Deiteretal.,2000;Dulit,Fyer,Leon,Brodsky,&Frances,1994;Galley,2003;Martin,Rozanes,Pearce,&Allison,1995).Thedifficultyinaccuratelyassessinggenderdifferencesinself-injuriousbehaviormayarisefromvariationinhowmaleandfemaleself-injurersareidentified,howtheyinjurethemselves,andwhetherornottheyseektreatment(Alderman,1997;Connors,2000).Therelationshipbetweenself-injuryandsuicideisimportantbutnotclearlyunderstood.Personswhoengageinself-injuryaremorelikelytoconsiderorattemptsuicide(Walsh&Rosen,1988;Gardner&Cowdry,1985;Hawton,Fagg,Simkin,Bale,&Bond,2000).Nevertheless,inthevastmajorityofcasesself-injuryisusedtoalleviatedistresstemporarilyratherthantosignaltheintentiontoendone’slife(Favazza,1996;Rosen&Walsh,1989;Tantam&Whittaker,1993).Indeed,somescholarsseeitasahighlyfunctionalalternativetosuicide(Alderman,1997;Strong,1998).Amongclinicalpopulations,self-injuryiscomorbidwithborderlinepersonalitydisorder,eatingdisorders,posttraumaticstressdisorder,depression,anxietydisorders,andahistoryofabuseortrauma(Alderman,1997;Connors,2000;Conterio&Lader,1998;Holmes&Nadelson,2000;Sansone&Levitt,2002;Yates,2004).Indeed,someresearchershavecalledforanewDiagnosticandStatisticalManualofMentalDisorderscontroldisorder,deliberateself-harmsyndrome,whichwouldincludeself-injuriousbehavior(Favazzaetal.,1989).Most,how-ever,seeself-injuryasamanifestationofmentaloremotionaldisordersorheterotypicmanifestationofchildhoodtrauma(Con-ners,2000;Conterio&Lader,1998;Favaro&Santonastaso,2002;Strong,1998;Yates,2004).Whetheratrauma-linkeddevelopmentalmodelwillholdtruefornonclinicalpopulationsisunclear.However,theproximalcausesforthosewhoengageinself-injuriousbehaviorinbothclinicalandcommunitysettingsarestronglylinkedtodifficultiesinregulatingstrongemotionsandcopingwithstress(Favazza,1996;Rosen&Walsh,1989;Ross&Heath,2002;Tantam&Whittaker,1993).Thecumulativeimpactofstressfullifeeventsthattriggerself-injuriousepisodes,thesubsequentshameassoci-atedwiththeself-injuriousresponse,andthesecrecysurroundingthebehaviorcreatefertilegroundforthedevelopmentofmargin-alizedperceptionsoftheself.TheInternetprovidesself-injurerswithananonymousvenueforsharingactualandfabricatedaspectsoftheirtrueidentityamongacommunityofsimilarothers.Be-causeyoungpeoplewithdepressivesymptomatologyaremuchmorelikelytotalkwithstrangersonlineandtodisclosepersonalinformationthanthosewithoutdepressivesymptomatology(Ybarra,Alexander,&Mitchell,2005),self-injurersmaybeparticularlyreceptivetousingtheInternettolocatevirtualcommunities.Al-thoughsuchvirtualcommunitiesmayprovideamuch-neededsourceofsupportandcatharticsharing,thepossibilitythatself-injuryiscommunicablesuggeststhattheInternetmayalsoservetospreadordeepenthepracticeinadolescentpopulations.AimofthePresentStudiesTwostudiesweredesignedtoexploreanddocumentadolescentuseofonlinemessageboardstoshare,solicit,andreceiveinfor-mationandadvicerelatedtoself-injuriousbehavior.Messageboardsareanelectronicvenueinwhichindividuals(i.e.,“posters”)registerwiththesiteunderachosennameandareallowedtopostthoughts,askquestions,andrespondtootherposters.Unlikee-mailandIM,whichpermitprivateexchangebetweenselectedindividuals,messageboardsandchatrooms(inwhichexchangesoccurinrealtime)areentirelypublic:Allpostingsareavailabletoallmembersand,inmanycases,nonmembersaswell,althoughnonmemberscannotpostmaterial.Whenindividualslogontoamessageboard,thecontentvariesbuttypicallyincludesinforma-tionaboutthemessageboardpurposeandrulesaswellasdirectlinksto“threads,”poststhathavebeenmadebyuserssortedbysubjecttitle.Anyonewhologsontoamessageboardmayreadposts,butonlythoseregisteredmaymakeapost.Blogs,whicharebecomingincreasinglycommon,areessentiallypublicelectronicdiaries.Messageboardpostersfrequentlylinktheirblogsitetotheirmembershipidentityorthesignaturelineintheirposts.Thegoalofthefirststudywastoinvestigatetheprevalenceandnatureofself-injurymessageboardsandtheiruserswiththeintentionofbetterunderstandingthegeneralsubjectcategoriesmostcommonlyraisedfordiscussion.Thegoalofthesecondstudywastoexplorethecorrelationsbetweencontentareasraisedfordiscussion.Bothstudieswereintendedtoshedlightontheroleofmessageboardsinspreadinginformationaboutself-injuriousprac-ticesandinfluencinghelp-seekingbehavior.StudiesoftheInternetareeffectivelystudiesofculture.Participantsarefreetoconstructtheirageandidentityastheywish,andstudiesshowthatteenInternetusersoftenconstructthemselvesassomewhatolderthanVIRTUALCUTTINGEDGE compassedvirtuallyallmessageboardexchangesandserved,alongwithallspecificthemestheyencompassed,asacodingtool(e.g.,loneliness,anger,anddysphoriawereallthemesinthemotivationcategory).Postsleftuncodedwerethosewithlittlecapacitytoilluminateunderstandingofself-injuriousbehavior,suchasidlediscussionaboutacurrenteventoractivity.Theresulting11areaswereasfollows:(a)informalsupportandexchange,(b)motivationforself-injury,(c)concealmentofself-injuriousbehavior(e.g.,anxietyaboutexposure,methodsforconcealmentofcutsandscars),(d)addictionlanguage(e.g.,daysself-injuryfree,difficultystopping),(e)formalhelpseekingandtreatment,(f)sharingtechniques,(g)linkstoothermentalhealthorbehavioralconditionsknowntobeassoci-atedwithself-injuriousbehavior,(h)referencestopopularculture,(i)perceptionsofnon-self-injurersreactionstoself-injuriousbehavior,(j)perceptionofselfandbehavior(e.g.,self-worth,lovability,dissociation),and(k)ventingorapologizing.Theseareascontainedatotalof70themesintowhichnearlyallpostscouldbecategorized.Codingofposts.Acontentanalysiswasconductedonalloriginalandfollow-upresponsestotheoriginalpostin10messageboardsovera2-monthperiod.Atimeframeof2monthswasselectedtoensureadequatebreadthincontentareas.Atotalof3,219postswereexaminedduringthisperiod.Bothoriginalpostsandresponsestopostedmessageswerecoded.Becausenotallpostscontainedcontentrelatingtothecodingscheme,thetotalnumberofpostsexaminedexceedsthetotalnumberofpoststowhichacodewasassigned(2,942).(Thecriterionforleavingpostingsuncodedisdescribedlater).However,ifpostscontainedmorethanonethematicreference,theywereassignedmultiplecodes.Therefore,totalsintheResultssectioncanadduptomorethan2,942.FourcodersweretrainedbyJanisL.Whitlockandthestudycoordinatortoobserveandcodemessageboardsduringthistimeperiod.Thiswasaccomplishedbyhavingeachcoderandtheprincipalinvestigatorcodethesamepostsanddiscusscodeassignment.Oncetrainingwascompleted,eachcoderloggedononetotwotimesperweekduringthestudyperiodtoidentifyandcodenewposts.Toestablishintercoderreliability,eachcoderteamwasgiventhreerandomlyselectedthreadstocode.Codesforeachindividualcoderwerethencomparedtoestablishagreementwithineachteam.Theteamswereinfullagreementaboutcodesfor96%and97%ofthepostings,respectively.Messageboardprevalence,datesofinception,andcolistingwithothertopics.Thesearchproceduredescribedpreviouslyrevealed406boards.AsshowninTable2,examinationofthe140boardsavailablethroughtheInternetserviceprovidinghistoricalinformationshowsanincreasefrom1998to2000andthenagenerallystabletrendovertime,withpeaksin2000andin2002.Althoughcautioniswarrantedwheninterpretingthesedata,be-causepublicfamiliarityanddedicatedserverspacewithmessageboardshavealsoincreasedovertime,thesefindingsindicatethatmessageboardsasaculturaltoolforself-injurygrewdramaticallybetween1998and2000andthatinterestinbothestablishingandparticipatinginself-injuryspecificgroupsoverthepast5yearshasbeensustained.Examinationofthesesamemessageboardsforcolistingwithothertopicsrevealslinkswithanumberoftopicsknowntobeassociatedwithself-injuryintheliterature.Table3showstheextenttowhichself-injurymessageboardswerecross-listedwithothertopics.Althoughlistedmostoftenalone(56%),44%ofthemessageboardswerecross-listedwithoneormoreothertopics.Whenlinkedtoanothertopic,self-injuryoccursmostoftenwithdepression(32%)andeatingdisorders(17%).Althoughlessfre-quent,itoftenoccurswithmessageboardsdedicatedtodiscussionsofdissociativeidentityandmultiplepersonalitydisorders(10%);bipolardisorder(9.2%);sexualabuse(7%);obsessive–compulsivedisorders(7%);addiction(2.8%);anxietydisorder(2.8%);lesbian,gay,bisexual,andtransgenderissues(2.8%);andautism(1.4%).Messageboardmembershipanalysis.The10mostpopularmessageboardswereselectedformembershipanalysis.Table1showstheaveragesforself-description:memberage,agerange,andpercentageoffemales.Italsoshowsthepercentageofmem-berswithpublicblogs,thepercentageofregisteredmemberswhohaveneverposted(potentialviewers),percentageofposterswithmorethan100posts,andmoderationlevel.Theaverageself-describedageofmembersrangedfrom16.4to23.9yearsforeachmessageboard,althoughtherewaslargevariationinagesrepre-sentedandthemeantendstobenegativelyskewed.Themajorityofmessageboardshadameanstatedageof18years;80%ofthemembersdescribedthemselvesasbeingbetweentheagesof14and20;31%ofallmembersdescribedthemselvesasbeing15or16yearsold.Inall10boardsexamined,postersdescribingthem-selvesasfemaleweremorelikelytoberegisteredandtopartici-pateactively.Thenumberofmembersineachmessageboardvarieddramaticallyfrom70to6,656.Membershipdatafromoneofthesiteswerenotavailablebecausememberspostinanyofawidenumberofforums,notallofwhichrelatetoself-injury.Table3ConditionsandBehaviorsWithWhichSelf-InjuryMessageBoardsAreCross-Listed Cross-listedcondition/behavior%ofDepression4532.0Eatingdisorders2417.0Dissociativeidentitydisorder/multiplepersonalities1410.0Bipolar139.2Obsessive-compulsivedisorder107.1Sexualabuse107.1Posttraumaticstressdisorder75.0Anxiety42.9LGBTQ21.4Mooddisorders42.8Addiction42.8Autism21.4Nootherdisorderspecified7856.0lesbian,gay,bisexual,transgender,andquestioning.Table2NumberofSelf-InjuryMessageBoardsFromOneInternetServiceEstablishedbyYear YearNo.boardsTotalmembership1998193199979492000262,8312001257032002281,6112003199522004248062005381,698Total1689,643VIRTUALCUTTINGEDGE therapyaccountedforasizableportionofformalhelp-seekingpostsaswell(26.2%).Negativeattitudesandactivediscourage-mentfromseekingtherapywerealsoevidentin18.7%oftheformalhelp-seekingposts.Referencestospecificmedicationsoc-curredinabout1in10cases.Thesixthmostcommoncategoryofdiscussionforallagegroupsrelatedtosharingdetailsabouttechniquesusetoself-injury,accountingfor6.2%ofallposts.Theseexchangesweregenerallyeitherdescriptionsofspecificself-injurytechniquesorrequestsforspecifictechniqueinformation.Thefollowingconver-sationexemplifiesthistypeofexchange:Poster1:Doesanyoneknowhowtocutdeepwithouthavingitstingandbleedtoomuch?Poster2:Iuseboxcutterblades.Youhavetopulltheskinreallytightandpressthebladedownreallyhard.Youcanalsouseatourniquettomakeitbleedmore.Poster3:I’vefoundthatifyoupressyourbladeagainsttheskinatthedepthyouwantthecuttobeanddrawthebladereallyfastitdoesn’thurtandthereisbloodgalore.Becareful,though,‘causeyoucangoverydeepwithoutmeaningto.Poster1:Okay,I’llgetaStanleyblade‘causeIhearthatitwillcutrighttothebonewithnohassle.ButI’llbecarefulifIdouseatourniquetandIwon’tcutthatdeep.Mentalhealthconditionsempiricallylinkedtoself-injuriousbehavior,suchasdepression,eatingdisorders,suicidality,andsexualabuse,werecitedaseitherlinkedwithorasatriggerforself-injuryin4.7%ofallposts(seeTable4).Referencestopopculture(e.g.,music,movies,books,celebrities,andcharacterswithspecialsignificancetoposters)appearedinjustover4%oftheposts.Discussionandinterpretationoftheperceptionsofotherscameupin2.6%oftheposts.Recognitionofthepaintheirbehaviorcausedormightcauseothers,inconjunctionwithdiffi-cultystoppingabehavioronceapatternwasformed,wasoftenlinkedtotheneedforsecrecyandfeelingsofshame.Similarly,perceptionsofself,usuallynegative,weresharedinjustover2%ofallposts.Remainingpostswererants,aimlessventing,usuallyfrustrationoranger,orapologiesforsharing.Study2:VariationsintheUseofMessageBoardsAmongThisstudywasintendedtoexplorethecorrelationsbetweencontentareasraisedfordiscussion.Itfocusesonexchangeinfourbroadareas:(a)helpseekinganddisclosure,(b)techniquesharing,(c)comorbidity,and(d)attitudestowardselfandotherposters.Theseareaswereselectedtoexploretheroleofmessageboardsexchangeinspreadinginformationaboutself-injuriouspracticesandinfluencinghelp-seekingbehavior.Identificationandsampling.Theunitofanalysisforthisstudywastheindividualself-injurymessageboardposter.Individualposterswereselectedfromfivelow-andonemedium-moderationself-injurymes-sageboardsincludedinthepreviousstudy(A,B,D,E,F,andGinTable1).Becausetheextenttowhichposterssoughtandsharedself-injurytechniqueswasofinterest,low-moderationboardswereselectedasrecruitmentsitestoensurethatthecontentofmessagepostswouldnotbeblocked.Onemedium-moderationboardwasselectedtopermitabalancedsamplebyage,becausethelow-moderationboardsdidnotcontainanadequatenumberofactivepostersforwhichagewasavailable.Eachsiteallowedforallpostsfromindividualposterstobesearchedreadily.Individualposterswereidentifiedbylookingthroughthreadsandiden-tifyingindividualsintheagebracketsofinterestwithmorethan50posts.Then20ineachofthreeself-describedagegroups(13–15,16–18,19–22years)wereselectedatrandom,foratotalsamplesizeof60.Onceindividualposterswereidentified,50postsfromeachindividualwereselectedatrandomfrompostsmadefromJuly2004toJanuary2005.A6-monthwindowwaschosentoensurethateachpostermonitoredwouldhaveadequatetimetomakemultiplepostsandengageinavarietyofAsinStudy1,thecontentanalysisofpostswasconductedusingasetofbinary(present–absent)codes.ThesecodesweresimilartothoseusedinStudy1butwerefocusedonsixspecifictypesofexchange:(a)solicitingandsharingtechniques,(b)attitudestowardanddiscloseduseofformalsupport,(c)solicitationandprovisionofinformalsupport,(d)disclosuretonontherapeuticothers(e.g.,familyandfriends),(e)disclosureofothermentalhealthconditions,and(f)disclosedself-concept.Atotalof17measureswereusedtoindicateeachofthesebroadconceptualdomains(Table6).AlthoughmanyofthesemeasuresaresimilartothoseusedinStudy1,Study2measuresweredesignedtofocusonaspecifictypeofexchangeandtopermitindividual-levelanalysisofcorrelationsbetweenvariables.Self-representeddemographics,includinggender,age,andtotalnumberofposts,werealsorecorded.Codingofposts.Atotalof3,000individualpostswerecoded(50pereach60individuals).Theactualnumberofpostsperindividualrangedfrom60toseveralhundred.Becausecoderscouldqueryforallthreadspostedbyindividualposters,theyrandomlyselected50byidentifyingthetotalnumberofpagesandthreadsanyonepostermadeanddividingby50.Theresultingnumberwasusedtosystematicallyidentifythreadsincludedintheanalysis.Forexample,ifapostermade100posts,coderscodedeverytwoposts.Iftheresultingintervalinadvertentlyidentifiedaredun-dantthread(onealreadycodedbutthatsurfacedlater),thenextthreadwasused.Postswerenotcodedwhentheydidnotcontaincontentrelevanttothestudyobjectives.Sixcoderswereassignedindividualposterstotrackandcode.Toestablishintercoderreliability,40%oftheobservations(1,200posts)wereindependentlycodedbythreepairsofcoders.Eachpairofcoderscoded400postsincommon.Notwopairscodedthesameposts.Agreementwasassessedbycalculatingtheproportionofpostseachindividualinthepaircodedthesame.Theaverageintercoderagreementacrossallthreepairswas93%(range90–96%).Toassesstherelationshipbetweensolicitingandsharingself-injuriouspracticesandinformalandformalhelpseeking,Spear-mancorrelationswereconductedbetweenallmeasures.ResultsareshowninTable6.Theanalysisrevealedtwotrendsinthecorrelations:oneinwhichmorepositiveexchangeswerecorrelatedandoneinwhichnegativeexchangeswerecorrelated.Forexample,offeringinfor-malsupportwascorrelatedwithdisclosingthatsomeoneknows,suggestingformaltreatment,seekingadviceonstopping,anddisclosingahistoryoftrauma.Similarly,seekingadviceonstop-pingandharmreductionwerecorrelatedwitheachotherandwithseekingadviceaboutdisclosureandsharingpositiveremarksaboutoneself.Encouragingformaltreatmentofself-injuriousbe-VIRTUALCUTTINGEDGE havior,offeringinformalsupport,andencouragingdisclosuretootherswereinterrelatedaswell.Asomewhatmorenegativeclusterofexchangelinkeddiscus-sionoftechniquesharingwithnegativeattitudestowarddisclo-sure.However,discouragingdisclosurewasalsoassociatedbothwithpositiveviewsofselfandwithseekingadviceonstopping.Inaddition,disclosingthatonewasintreatmentwascorrelatedwithdisclosinganotherdiagnoseddisorderandsuicidality.Individualswhodisclosedahistoryoftraumawerelikelytoofferinformalsupport,andthosewhodisclosedsuicide-relatedbehaviorsweremorelikelythanotherstodisclosethatsomeoneknowsabouttheirself-injuriousbehavior.OurfindingsconfirmthatInternetmessageboardsprovideapowerfulvehicleforbringingtogetherself-injuriousadolescents.Althoughthemessageboardsexaminedforthesestudiesmaynotberepresentativeofallself-injurymessageboards,theydoprovideasnapshotofcontentandexchangecommoninthosewithhighactivity.Wefoundthatinthelast5yearshundredsofmessageboardsspecificallydesignedtoprovideasafeforumforself-injuriousindividualshavecomeintoexistence.Manyofthesearepopulatedbyindividualswhoidentifythemselvesasfemalesbe-tween14and20yearsofage.Althoughthestrongpreponderanceoffemalesmaynotaccuratelyreflectthegenderbreakdownofself-injuriousbehaviorinthegeneralpopulation(Whitlock,Eck-enrode,&Silverman,inpress),itmayreflectthetendencyforfemalestosolicitmoreinformalandformalhelpandsocialsupportcomparedwithmales(Fuhrer,Stansfield,Chemali,&Shipley,1999;Saunders,Resnick,Hoberman,&Blum,1994).Onceonline,messageboardmembersareabletopostorpas-sivelyobserveawidevarietyofanonymousexchanges.Justlessthanhalfofallmessageboardsweinvestigatedwerecross-listedwithconditionsknowntobecomorbidwithself-injuriousbehav-ior,suchasdepression,eatingdisorders,andsuicide.Ofalltypesofonlinedialogue,thegivingandreceivingofinformalsupportanddiscussionofproximallifeeventsthattriggerself-injuryaremostcommon.Postersalsosharecasualandsometimesverypersonalinformationrelatedtotheaddictivequalitiesoftheirpractice,theirfearsrelatingtodisclosure,experienceswithther-apy,howtheyself-injure,andotherrelatedhealthconcerns.Al-thoughourfindingsaregenerallyconsistentwithexistingliteratureonself-injuriousbehavior,thecorrelationsdocumentedamonginformalsupport,encouragingdisclosure,andadvisingformaltreatmentsuggestthatonlineinteractionsmaybeprovidingself-injurerssupportandmeaningoutsidetheclinicalsetting.Whattheself-injuriousadolescentsinourstudyappearedtodoonlineiswhatmostpeoplewhotrusteachotherdoinconversa-tion:exchangesupport,sharepersonalstoriesaboutdailylifeevents,andvoiceopinionsandideas.BecausetheanonymityoftheInternetinspiresthemostpersonalandtrustingofexchangebetweenindividualswithlittleornopreviousrelationship,onlinesharingmayencouragegreaterandmoretruthfuldisclosures(McKenna&Bargh,2000),especiallyamongself-injurers,manyofwhomsufferfromdepressivesymptomatology(Ybarraetal.,2005).Foradolescents,thissupportmaybeparticularlyvaluable,becausehealthysocialandemotionaldevelopmenthingesontheirabilitytoestablishcaring,meaningfulrelationships,tofindaccep-tanceandbelonginginsocialgroups,andtoestablishinterpersonalintimacy(Reis&Shaver,1988;Sullivan,1953).Thesedevelop-mentaltaskscanbeespeciallydifficultforyoungpeoplestrugglingwithintenseshame,isolation,anddistress,particularlywhenthesourceandoutcomeofthesefeelingsmustbekepthidden.Theassuranceofonlineanonymitymaycontributetoidentityconstruc-tionbyprovidingopportunitiesforadolescentswithmarginalizedornonmainstreamproclivitiestoexperimentwithdifferentsocialrolesandselves(McKenna&Bargh,2000;Turkle,1995).Thelesspositivesideofourfindingssuggeststhatparticipationinself-injurymessageboardsmayalsoexposevulnerableadoles-centstoasubcultureinwhichself-injuryisnormalizedanden-couraged.Forexample,issuesrelatedtoconcealmentofself-injuriousbehavior,identifiedasadominantthemehere,maymakeself-injurymessageboardsparticularlypotentagentsofself-injury,becausesharingtechniquesandmotivescantakeplaceanony-mously.Inlightofevidencehereandelsewherethatself-injuriousbehaviormaypossessaddictivequalities(seeYates,2004,forreview),theadolescentdrivetobelongandthesatisfactionthatcomeswithassociatingwithacommunityofsimilarothersmayinadvertentlyfeedafundamentallyself-destructivebehaviorforsomeparticipants.Thecorrelationdocumentedherebetweenshar-inginjurioustechniquesanddiscouragingdisclosurelendssupporttothispossibility.Althoughnotimpossibleforindividualself-injurerstohavegatheredbeforetheadventoftheInternet,easyaccesstoavirtualsubcultureoflike-mindedothersmayreinforcethebehaviorforamuchlargernumberofyouth.Thetendencyforself-injuriousbehaviortofollowepidemic-likepatternsininstitutionalsettingssuchashospitalsanddetentionfacilities(Matthews,1968;Ross&McKay,1979;Rosen&Walsh,1989;Taiminen,Kallio-Soukainen,Nokso-Koivisto,Kaljonen,&Helenius,1998)suggeststhatthebehaviormaybesociallycontagiousinothersettingsand,therefore,throughtheInternetaswell.AsBrumberg(1992)hasarguedforeatingdisorders,discussionoftechniquesandtheperceivedbenefitsofself-injurymayaddpotentiallylethalbehav-iorstotherepertoireofestablishedself-injurers,notyetcommittedmessageboardmembers,andevennonparticipatingmessageboardviewerswhoareexploringidentityoptions.Indeed,somemessageboardscontainlinkstopro-self-injuryWebsiteswhereInternetuserscanpurchaseself-injuryparaphernaliasuchasbraceletsorclothingthatsignifyself-injurystatusandcuttingclubshavebeenrumoredtobeagrowingformoffriendshipritual(Booth,2004).Someself-injuryWebsiteshostforumsspecificallydedicatedtosharingnewself-injurytechniques.Discussionoftechniqueshar-ing,triggers,negativeattitudestowardformalandinformalhelpseeking,andthepleasuresandpainsofself-injuryaddictionmayinfluencebehavioralchoicesoutsideofthevirtualrealmthatarelaterbroughtback,shared,andusedtoensuresupportandmem-bership.Itmayalsomakesomeyouthtargetsforindividualswhofalselyposeassupporterstoaccomplishother,lessbenevolentaims.Forvulnerableadolescents,thedifficultyofendingastrat-egyforcopingwithdistress(self-injury)andleavinganeededsourceofsupport(individualorcollectivemembersofthevirtualcommunity)maystiflethedesiretofindalternatewaysofcopingwithstress.Moreover,thelow-senseofself-worthcommonamongself-injurersmayexposethemtodamagingonlinerelationships.ThisstudysupportsthefindingsofotherscholarsofInternetanddevelopment(Suzuki&Calzo,2004)bysuggestingthatelectronicVIRTUALCUTTINGEDGE Gross,E.F.,Juvonen,J.,&Gable,S.L.(2002).Internetuseandwell-beinginadolescence.JournalofSocialIssues,58,Hartup,W.W.(1996).Thecompanytheykeep:Friendshipsandtheirdevelopmentalsignificance.ChildDevelopment,67,Hawton,K.,Fagg,J.,Simkin,S.,Bale,E.,&Bond,A.(2000).Deliberateself-harminOxford,1985–1995.JournalofAdolescence,23,Heitner,E.(2002).TherelationshipbetweenuseoftheInternetandsocialUnpublisheddoctoraldissertation,PaceUniversity.Holmes,A.,&Nadelson,C.A.(Eds.).(2000).Cuttingthepainaway:Understandingselfmutilation.Philadelphia:ChelseaHouse.Kokaliari,E.(2005).Deliberateself-injury:Aninvestigationoftheprev-alenceandpsychosocialmeaningsinanon-clinicalfemalecollegeUnpublisheddoctoraldissertation,SmithCollege,SchoolforSocialWork.Kraut,R.,Kiesler,S.,Boneva,B.,Cummings,J.,Helgeson,V.,&Crawford,A.(2002).Internetparadoxrevisited.JournalofSocialIssues,51,49–74.Kraut,R.,Patterson,M.,Lundmark,V.,Kiesler,S.,Mukopadhyay,T.,&Scherlis,W.(1998).Internetparadox:Asocialtechnologythatreducessocialinvolvementandpsychologicalwell-being?AmericanPsycholo-gist,53,Lenhart,A.,Madden,M.,&Hitlin,P.(2005,July27).Teensandtechnol-ogy:Youthareleadingthetransitiontoafullywiredandmobilenation.RetrievedOctober13,2005,fromhttp://www.pewInternet.org/pdfs/Lenhart,A.,Rainie,L.,&Lewis,O.(2001).Teenagelifeonline:Theriseoftheinstant-messagegenerationandtheInternet’simpactonfriend-shipsandfamilyrelationships.Washington,DC:PewInternet&Amer-icanLifeProject.Maczewski,M.(2002).ExploringidentitiesthroughtheInternet:Youthexperiencesonline.Child&YouthCareForum,31,Martin,G.,Rozanes,P.,Pearce,C.,&Allison,S.(1995).Adolescentsuicide,depressionandfamilydysfunction.ActaPsychiatricaScandi-navica,92,336–344.Matthews,P.C.(1968).Epidemicself-injuryinanadolescentunit.nationalJournalofSocialPsychiatry,14,McKenna,K.Y.A.,&Bargh,J.A.(2000).Plan9fromCyberspace:TheimplicationsoftheInternetforpersonalityandsocialpsychology.sonalityandSocialPsychologyReview,4,McKenna,K.Y.A.,&Green,A.S.(2002).Virtualgroupdynamics.GroupDynamics:Theory,ResearchandPractice,6,116–127.McKenna,K.Y.A.,Green,A.,&Gleason,M.(2002).RelationshipformationontheInternet:What’sthebigattraction?JournalofSocialIssues,58,9–31.Merten,D.E.(1996).Visibilityandvulnerability:Responsestorejectionbynonaggressivejuniorhighschoolboys.JournalofEarlyAdolescence,Moffit,T.E.(1993).Adolescence-limitedandlife-course-persistentanti-socialbehavior:Adevelopmentaltaxonomy.PsychologicalReview,674–701.Muehlenkamp,J.J.,&Gutierrez,P.M.(2004).Aninvestigationofdifferencesbetweenself-injuriousbehaviorandsuicideattemptsinasampleofadolescents.Suicide&Life-ThreateningBehavior,34,Nie,N.H.,&Erbring,L.(2000).Internetandsociety:Apreliminaryreport.Stanford,CA:StanfordInstitutefortheQuantitativeStudyofSociety.Nixon,M.K.,Cloutier,P.F.,&Aggarwal,S.(2002).Affectregulationandaddictiveaspectsofrepetitiveself-injuryinhospitalizedadolescents.JournaloftheAmericanAcademyofChild&AdolescentPsychiatry,41,Petersen,A.C.,Kennedy,R.E.,&Sullivan,P.(1991).Copingwithadolescence.InM.E.Colton&S.Gore(Eds.),Adolescentstress93–110).NewYork:Aldine.Reis,H.T.,&Shaver,P.(1988).Intimacyasaninterpersonalprocess.InS.Duck(Ed.),Handbookofpersonalrelationships:Theory,research,andinterventions(pp.367–389).Chichester,UK:Wiley.Roberts,D.F.,Foehr,U.G.,&Rideout,V.(2005).GenerationM:Mediainthelivesof8–18year-olds.Washington,DC:HenryJ.KaiserFamilyRosen,P.M.,&Walsh,B.W.(1989).Patternsofcontagioninselfmutilationepidemics.AmericanJournalofPsychiatry,146,656–658.Ross,R.R.,&McKay,H.B.(1979).Selfmutilation.Lexington,MA:LexingtonBooks.Ross,S.,&Heath,N.(2002).Astudyofthefrequencyofself-mutilationinacommunitysampleofadolescents.JournalofYouthandAdoles-cence,31,66–77.Ross,S.,&Heath,N.L.(2003).Twomodelsofadolescentself-mutilation.Suicide&Life-ThreateningBehavior,33,Sansone,R.A.,&Levitt,J.L.(2002).Self-harmbehaviorsamongthosewitheatingdisorders:Anoverview.EatingDisorders,10,Saunders,S.N.,Resnick,M.D.,Hoberman,H.M.,&Blum,R.W.(1994).Formalhelpseekingbehaviorofadolescentsidentifyingthemselvesashavingmentalhealthproblems.JournalofAmericanAcademyofChildandAdolescentPsychiatry,33,718–728.Strong,M.(1998).Abrightredscream:Self-mutilationandthelanguageofpain.NewYork:Viking.Subrahmanyam,K.,Greenfield,P.M.,&Tynes,B.(2004).Constructingsexualityandidentityinanonlineteenchatroom.JournalofAppliedDevelopmentalPsychology,25,651–666.Sullivan,H.S.(1953).Theinterpersonaltheoryofpsychiatry.NewYork:Suzuki,L.K.,&Calzo,J.P.(2004).Thesearchforpeeradviceincyberspace:Anexaminationofonlineteenbulletinboardsabouthealthandsexuality.AppliedDevelopmentalPsychology,25,685–698.Taiminen,T.J.,Kallio-Soukainen,K.,Nokso-Kovisto,H.,Kaljonen,A.,&Helenius,H.(1998).Contagionofdeliberateself-harmamongadoles-centinpatients.JournaloftheAmericanAcademyofChild&AdolescentPsychiatry,37,Tantam,D.,&Whittaker,J.(1993).Self-woundingandpersonalitydisor-der.InP.Tyrer&G.Stein(Eds.),Personalitydisorderreviewed191–224).London:AmericanPsychiatricPress.Turkle,S.(1995).Lifeonthescreen:IdentityintheageoftheInternet.NewYork:Simon&Schuster.Tynes,B.,Reynolds,L.,&Greenfield,P.M.(2004).Adolescence,race,andethnicityontheInternet:Acomparisonofdiscourseinmonitoredvs.unmonitoredchatrooms.JournalofAppliedDevelopmentalPsychology,667–684.Walsh,B.W.,&Rosen,P.M.(1988).Self-mutilation:Theory,researchandtreatment.NewYork:GuilfordPress.Wartella,E.,Caplovitz,A.G.,&Lee,J.H.(2004).FrombabyEinsteintoleapfrog,fromdoomtoTheSims,frominstantmessagingtoInternetchatrooms:Publicinterestintheroleofinteractivemediainchildren’sSocialPolicyReport,18,Welsh,P.(2004).Lookstudents’scarspointtoemotionalpain.,p.11a.Whitlock,J.L.,Eckenrode,J.J.,&Silverman,D.(inpress).Self-injuriousbehaviorinacollegepopulation.Yates,T.M.(2004).Thedevelopmentalpsychopathologyofself-injuriousbehavior:Compensatoryregulationinposttraumaticadaptation.PsychologicalReview,24,Ybarra,M.L.,Alexander,C.,&Mitchell,K.J.(2005).Depressivesymp-tomatology,youthInternetuse,andonlineinteractions:AnationalJournalofAdolescentHealth,36,9–18.YoungPeopleandSelfHarm:ANationalInquiry.(2004).Whatdowealreadyknow?Prevalence,riskfactors,andmodelsofintervention.RetrievedJune29,2004,fromwww.selfharmuk.orgReceivedApril1,2005RevisionreceivedDecember15,2005AcceptedDecember16,2005VIRTUALCUTTINGEDGE