Jefferson Journal of Psychiatry Jefferson Journal of Psychiatry

Volume 16 Issue 1 Article 6

January 2001

Personality Traits in Juvenile Delinquents: Associations with Peer Personality Traits in Juvenile Delinquents: Associations with Peer

and Family Relations and Family Relations

Jeffrey J. Wilson M.D. Columbia University Department of Psychiatry

Sharon E. Willams Ph.D. Stanford University Department of Psychiatry

Evan Garner M.D. Stanford University

Elaine Duxbury B.A. Department of the Youth Authority, Sacramento, CA

Hans Steiner Dr. Med. Univ. Stanford University Department of Psychiatry Follow this and additional works at: https://jdc.jefferson.edu/jeffjpsychiatry

Part of the Psychiatry Commons

Let us know how access to this document benefits you

Recommended Citation Recommended Citation Wilson, Jeffrey J. M.D.; Willams, Sharon E. Ph.D.; Garner, Evan M.D.; Duxbury, Elaine B.A.; and Steiner, Hans Dr. Med. Univ. (2001) "Personality Traits in Juvenile Delinquents: Associations with Peer and Family Relations," Jefferson Journal of Psychiatry: Vol. 16 : Iss. 1 , Article 6. DOI: https://doi.org/10.29046/JJP.016.1.005 Available at: https://jdc.jefferson.edu/jeffjpsychiatry/vol16/iss1/6

This Article is brought to you for free and open access by the Jefferson Digital Commons. The Jefferson Digital Commons is a service of Thomas Jefferson University's Center for Teaching and Learning (CTL). The Commons is a showcase for Jefferson books and journals, peer-reviewed scholarly publications, unique historical collections from the University archives, and teaching tools. The Jefferson Digital Commons allows researchers and interested readers anywhere in the world to learn about and keep up to date with Jefferson scholarship. This article has been accepted for inclusion in Jefferson Journal of Psychiatry by an authorized administrator of the Jefferson Digital Commons. For more information, please contact: JeffersonDigitalCommons@jefferson.edu.

Personality Traits in Juvenile Delinquents:Associations with Peer and Family Relations

Jeffrey J. Wilson, M.D. ,1 Sharon E. Williams, Ph.D.,2Evan Garner , M.D. ,3 Elaine Duxbury, B.A.,4 and

Dr. Med. Univ. Hans Steiner"

Abs tract

Objective: To establishfam ily and peercorrelatesqf'personality traits shown to bepredictiveqff uture criminal recidivism.

Method: 137 incarceratedboys aged 13-18 (x = 16 ± 1.2), 35%black, 21% Hispanic, 13%white, and 1% other completed the Weinberger Adjustment I IIveiltory (WAI) and a psychosocialhistory obtained by a social worker. Records were summarized using two raters who assignednumerical ratings to dimensions qf'family and peer relations.

Results: As expected, observer ratings qff amily and peer relationships were correlated withthe personality characteristics qf distress, denial and restraint as measured by the WAI.

Conclusion: Family and peer relations are associated with certain personality traits that arepredictive qf criminal recidivism in delinquents. This studyfurth er expands the knowledge baseregarding the social and interpersonalcorrelatesofindividual traits predicting criminal recidivism.

Ii\TTRODUCTION

Few issu es perplex modern soc ie ty more th an delinquen cy. The cost of delin ­qu cnt crime is pervasive, including th e devastating pri ce th e victims pay, the pub lica nd approp riate fear for safety in many U .S. cit ies, the pri ce th e socie ty pays to carefor thes e incarcerated youth, and th e toll th at a crime path ex torts from theindividua l delinquent. Worldwide, increa sing a tten t ion is bei ng paid to t hc develop­ment of conduc t problem s a nd delinquen cy. As a result of pu blic interest andfunding, conside rable ga ins have been mad e in th e scien t ific underst a nd ing ofdelinquen cy, particul arly related to th e classification a nd ep idemiology of conduc tproblems a mong youth (I) .

Crim e a mo ng youth is a complex ph enomenon. Wh ile young peop le under 25

IJe fTrey J . Wil son , M.D. Clinical Research Fellow in Ado lescent Substance Abus e at ColumbiaUniversi ty, Dep artm en t of Psychi at ry, Division on Subs tance Abuse.2Sha ron E. Williams, Ph.D. is Assista nt Professor of Psychia try at Stanford Univers ity,Dep artmen t of Psychia t ry, Division of C hild and Ado lescen t Psychi a t ry.3Eva n Garner, M.D. Residen t in Psychi at ry a t St an ford Univers ity.lElaine Duxbury, B.A. is Chief of Resea rch at th e Departm ent of the Youth Authority,Sacramento, CA.5Dr. Med. Uni v. Hans St einer is Professor of Psychia t ry and Dir ector of Tra ini ng a t StanfordUnivers ity, Depa rtment of Psychia try, Division of C hild and Ado lescent Psychia t ry.

32

PERSONALITY TRAITS IN J UVENILE DELINQ UEl\JTS 33

com mit half of nonfatal violent crime, th ey also face th e high est risk of assa ult (2) .The relationship between youth violen ce a nd exposure to viole nce du ring childhoodis strongly suggested by th e risk enge nde red for conduc t dis order when a child isabused (3). In fact , witnessing violen ce has been shown to incr ease antisocialbehavior in high risk, urban boys (4). The perpetrators of th ese violen t crime s ar every oft en victims th em selves, and mental health workers are increasingly aware ofth e psychopathology of crim inal beh avior and it s ant eced ents.

Aich orn 's study of Wayward You th in 1935 (5) elegant ly describes neurot icunderpinnings of delinquent beh avior for a group of youth. In 1946, Bowlby 'slandmark study of 44 thieves (6) postulates enviro nme nta l a ntecede nts to de linquentbeh avior from the perspective of object relations. Theoreticall y, fami ly pr act ices arethought to be vital in learning relat ed to behavioral regula tion. The development ofse lf-regula t ion a ppears to reduce th e need for exte rnal cont ro l a nd promote broadersocializa t ion outside th e family. As cite d by Weinberger (21) , both psych oan alyt ic (7)and social cognitive th eories (8) propose that se lf-regula tory controls develop pr i­marily within th e conte x t of th ese ea rly family relationships. Some aut hors qu estionth e relative contribu tions of th ese parental and peer environmental infl uences.Recently,judith Rich H arris ( 1998) (9) concludes th at th ere is littl e evidence for a nyparental nu rturing influ en ce on juvenile beh avior.

Despite th ese qu esti ons, the lit erature as well as clinica l wisdom sugges ts apowerful role for both parental a nd peer influen ces on delinquent be havior. Attach­ment diffi culties are implicated in th e et iology of disruptive beh avioral d isorders. Fo rexa mple, dis organized attachme nt status a t 12 months has been shown pr ospectivelyto pr edi ct high er ag gr essive problem s a t age 5 (10) . Moreover, family fac to rs such asparental rejection, lack of involv em ent, a nd low levels of monitoring of th e childpr edi ct th e lat er development of conduc t disorder a nd delinquen cy ( II). In fact ,Pat terson , DeBarysh e, a nd Ramsey ( 1989) show th at fa mily interacti ons and parentpr acti ces account for 30%-40% of th e variance in antisocia l beh avior ( 12).

Peer influ en ces a re a lso involved in th e beh avioral express ion of psychologicaltraits suc h as aggr essiven ess. For exa mple, Elli ott e t a l ( 1985) have shown th atdelinquen cy is pr edi ct ed by deviant peer relation s (13). "Peer pressur e" is a commo nterm used by schools a nd mental health pr ofession al s a like. Peer influ en ces ma y notbe comple te ly isolated from family fac to rs, however. For example, cons ide r DiLallae t al (1988) : whil e th ey find a significant relationship between lax punishment andhigh aggressiven ess, th ey also relate parent ratings of youth aggress iveness to bo thhigh turmoil in th e home, and youths' positive opinions of delinquen t peers (14) .

While this relation ship between parental and peer rela t ions is likely to bebidirecti on al , th ere is a conside rable body of knowled ge sugges ti ng th at th e initialindividual development of internal controls is relat ed to pa renting. Ainsworth (1989)sugges ts th at sec ure a ttachme nt facilitates th e pr ocess of successful individuation byproviding a sec ure base from whi ch to explore th e world (15). Fa mil ial a ttachmentsa nd later familial relation ships are import ant in th e develop ment of both extra­familial relationships a nd individual coping styles (16) . These results provide evi-

34 J EFFERSON J O R;,\AL OF PSYCHIATRY

den ce th at a sec ure parental base fac ilita tes th e developmen t of peer rela tio nshi ps,possibly by fost ering th e development of internal contro ls such as self-r estraint.

There appears to be a relationship between family environm en t a nd indi vidu alpsychological development. Positive and negative family cha rac te rist ics correlatewith mature and immature defenses , respectively, in adolescent fem al es ( 17). T hesedata imply that th e context in whi ch person ality developm en t occurs is int ima telyinvolved with th e development of int ernal psychological cons t ruc ts , i.e., de fensem ech anisms. More gene ra lly, secure paren tal a ttachments and peer relation sh ipsalso appear to facilitat e th e development of coping reso urces , with th e former beingrelated to grea ter resourcefulness ( 18). Amon g non-clinical popu lat ions, it is clea rth at fa milial influ en ces cont inue into ado lesce nce and provid e suppor t to thedeveloping ad olescent by weaken ing peer influen ces on pot entiall y dangerou s beh av­ior s such as substa nce abuse (19).

The developmental th em e of adolescence ge ne ra lly involves th e incr ea singinflu en ce of peers with de cr easing influ en ce of fa m ilia l relationsh ips. T he rela t ion­ship between th ese socia l sphe res is complica ted a nd difficult to separat e, particu­larly cons ide ring th e internali zation of early family influ en ces an d th eir manifesta­tion in personality development. Peer a nd family rela tio ns arc in terrelat ed in acomplex fashion , a nd both a re st ro ngly suggested to be int imately relat ed to t hedevelopment of conduc t, substance a buse a nd a nt isocia l personality disorders. Im­proved underst anding of delinquen cy re quires investi gation of int eracti on s betweensoc ia l relations (fa mily and peer) a nd person ality trait s suc h as se lf-re straint.

In order to evalua te th e parental rol e in th e develop ment of sel f-regula tion,Feldman & Weinberger (199 4) investigated self-res t ra int as a media tor of parenta linflu en ce on delinquent behavior in boys (20) . In a longitudinal study of 8 1 boys a ndth eir famili es (from a non-clinical population), initially in 6t h a nd later in lOt " gr ad e,th ey found th at parenting practi ces are asso cia te d with delinquen cy only indirectl ythrough th ei r asso cia tion with se lf-rest rain t. The first measure of self-r estraint , insixth-grade, predi ct ed delinquent beh avior more reli abl y tha n parent ing in lOt"grade. Their work further sugges ts th at by the end of childhood, a relatively stableinternal const ruc t of se lf-rest raint ca n be measured , which has predi ctive valueduring lat er development.

The measure used in this study, th e Weinberger Adj us tmen t Inventory, hasparticul ar utility for th e ass essm ent of socioe mo tiona l regul at ion among delinq uentyouth. This inventory mea sures two higher order person ality trait s, dis tress andrestraint. These broad traits ca n be placed within th e five factor model of pers onality,and encompass 4 (agr eeablen ess, conscie n tious ness, neuroticism a nd nega tive affec­tivit y) of th ese 5 dimen sions (21). Abnormal levels of distress and re s traint hav e beenrelated to va rious pr oblems, including: delinquent beh aviors, postt raumatic stressdisorder a mong delinquents (28), a nd alcoho l use amo ng colleg e stude nts (48) .Distress and restraint a re related to seve ra l dimen sions of th e Minnesota Multipha­sic Personality Inventory (MMPI), bu t may be a more favorabl e personal ity measurefor th e delinquent population (22) . Among delinquent youth, th ese tra its a rc linkedto delinquent beh avior cro ss-sec t ionally, retrosp ecti vely and prospecti vely (23,24,25).

PERSONALITY T RAITS IN J UVENILE DELINQUEl\'TS 35

These measures of di stress a nd restrain t have been sugges ted to inform psych oth er­ape u t ic a pproaches to delinquents a nd to deli neate fou r types of delinquen cy basedon high and low scores of restraint a nd distress (27,29). T he pro files described(Reactive, Suppre ssor, Repressor and Low Rea ctive) di scriminat e fu ture cr im inalrecidivism a mong incarcerat ed delinquents (25,50 ).

The a ffec tive dimen sion , di stress, is defin ed as individu al s' tend en cies to feeldissatisfied with th emselves and th eir subjective percept ion s of t he ir abilit ies toac hieve desired outco mes. In th is measure, four sub types are op erationally defin ed:pr on eness to a nxie ty, dep ression , low self-esteem, a nd low well-being. Examples ofth ese it ems follow, respecti vely: "I sp e nd a lot of t ime t hin king about things th atmight go wron g," "I often fee l th a t nobody really cares about me th e way 1want th emto," "I really don 't like myse lf very much ," and "I enj oy most of t he t hings I do duringth e week. " (T his last it em is rated inversely.)

Self-restraint refers to tenden cies across th e lifespan to inh ibit imm edi at e,se lf-focuse d desires in th e int erest of promoting lon g-t erm goals and positiverelation s wit h others. Se lf-res t ra int is a lso ope ra tio na lly d ivided in to four sub­dimen sion s: suppress ion of aggression, impulse con trol, conside ra t ion of ot hers, a ndres po ns ibility. Exampl es of th ese items follow, resp ectively: "When someone sta r ts afight wit h me, I fight back ," "I sto p a nd think things through before 1 ac t, " "I thinkabou t other people's fee lings before I do something t hey may not like, " and "I mak esure 1 stay out of trouble. " (The firs t of th ose it ems is rated inve rsely). Ext ernali zin gbeh avior, as measured by th e Child Behavior C hecklist, is strongly re lat ed to lowse lf-re st ra in t (26).

This study seek s to expa nd on pr eviou s st udy of the relation shi p betweencond uc t problems in youth and th eir soc ia l co rre la tes by exam ining a highlydisturbed gro up of you th in an incarce rat ed se tt ing . The object ive of th is study is toestablish family a nd peer co rre la tes of person ality trait s, which are pr edi ctive offu tu re cr imi na l recidi vism. It is hypoth esized th a t pe rso na lity t raits that are predi c­tive of cr im ina l recid ivism will be associated wit h avers ive fam ily relat ionships aswell as poor peer re lat ions .

METHODS

Subject Selection

A sample of 137 ad olescent boys from 13-1 8 years of age (x = 16, 0 = 1.2; 43%Whit e, 35% Black, 2 1% Hi spanic, a nd 1% othe r) was selec ted from inca rcera tedyou th offe nde rs a t th e Californ ia Youth Authority 's O .H. C lose Schoo l in St ockt on,CA. More th a n hal f of th e offe nders a re incarcerated for violen t offenses, and th eme dian number of prior offe nses is 4. Su bjects com ple ted a narrat ive of th eircom mitt ing offense, Weinberger Adjustment In ventory scored by a licensed psychol­ogist, a nd a n eva luation by a psychi a tric soc ia l worker. Records were summarized by2 ind ependent rat ers trained by a nd under th e supervis ion of a boa rd-cer t ifiedresearch C hild & Adol escent Psychiatrist a nd a licensed research psychologist.

36

M easures

JEFFERSON JOURNAL O F PSYCHIATRY

The Weinberger Adjustment Inventory (30) is an 84 it em self-r ep ort qu esti on­naire measuring two main d ime nsions, distress (29 ite ms tapping four dimen sions:anxie ty, dep ression , low well bein g, a nd se lf-es te em; a lpha = .8 1) and restraint (30items in 4 dimen sion s: impulse control, suppress ion of aggression, respon sibil ity, a ndconside ra t ion of othe rs; alpha = .86), as well as measur es of defen siven ess a ndden ial. This sca le has been shown to hav e reli abl e psychom et ric properti es and todisplay conve rge nt, di scriminant a nd predi ctive va lidity (20,29) .

Family an d peer rela tion s were rat ed by 2 ra te rs blind to th e WAI resul ts. Onerater eva lua ted family re la t ions, whi le a second eva luated peer relations. Family an dpeer relations were scored from criminological records of th e ward 's commi ttingoffe nse an d fro m th e eva luat ion by the psych iat ric social work er. Family relationswere rated for th e pr esen ce of pa re ntal figures (no, maybe, yes), st re ng th ofa ttachme nt (1-10), obj ective health of th e family ( 1- 10), wa rd 's percep tion offamily's health ( 1- 10), a nd the clear pr esence of ab use (yes, no). Peer relation s werera ted in four inde pe ndent categories : Normal, gang, delinquent , and drug-relat ed(no, maybe, yes) .

Statist ics

A se ries of Spearman corre la t ion coefficien ts were performed in order toeva lua te th e associations between : I. Fami ly re lations and personality cha racteris­tics, and 2. Peer relations and personality cha rac te rist ics.

RESULTS

Superordinate Personality D imensions and Social Relations

I . Family relations and personality . 38.0 % of our subjects report ed a n absen t orincon sistent ly present mother figure, and 79.6% rep ort ed a n abse nt orinconsist ently pr esent fath er figure. 78.8% of th e subjects reported some kindof abuse. The mos t robust fami ly relations variab le is child ab use, whic hpos itive ly corre la te s wit h distress (+.227, p = .008), a nd deni al (+ .224, p =.008) . (In an effort to kee p the dire ct ions of th e family variables uniform , wewill hereafter describ e ab use in th e opposit e direction, i.e., non abusiveparenting.) Presence or absence of parental figures , obj ective or subject ivemeasures of fami ly health, or self-reports of a tt ach me nt to parent al figuresare not significantly correlat ed with measures of distress . Co nt ra ry to ourhypotheses, fami ly relations measures are not associated with measures ofres t ra in t, except perhaps for a trend indi cating a rela tionship between t hesubscale of respons ibility and nonabusive parenti ng.

2. Peer relations and personality. "Norma l" peer relations arc signifi cantly associ­a ted with th e measure of restraint (+ .189, p = .027). Gang-r elat ed , del in-

PERSONALITY TRAITS IN JUVENILE DELINQ UENTS

TABLE 1.

37

Relationships Between Family, Peer and Personali ty Trait SubscalesAmong In ca rcerated Delinquents

SocialPersonality Trait s

Relations Dist ress Rest rain t

Family Al~X DEI' SEST LWB SAGR IMPC CONS RESI'

Mom+ -.027 - .047 -.050 .0 19 .059 .0 11 .062 .116l'op+ .099 .027 .040 - .001 - .074 -.052 .023 - .003Health Fam .187** - .070 -. 154* - .177** .092 -.065 .073 .120Non abusive

Parent ing -.076 - .227* ** - .188** -.215*** .089 .056 - .026 .159*Attach. . 176** - .067 - .117 -.131 .134 - .0 19 .121 .110Subj Fam .092 - .088 -.1 39 -.187** .092 -.0 18 .066 .136

Peer ANX DEI' SEST LWB SAGR IMPC CO NS RESP

Norm -.083 - .092 - .051 0.090 .107 .207** .160 .154*Gang -.1 32 - .044 -.043 - .002 -. 129 .053 - .164** - .062Delinq .068 .006 - .073 .115 - .02 1 .042 - .182** - .046Drugs - .001 .077 -.020 .220*** - .039 -.040 - .085 - .032

*p -s 0.10, **p s 0.05, ***p S 0.01Distress Sub scal esANX = Anxi et y, DEI' = Depression , SEST = Low Self-Es teem , and LWB = Low Well Being;Restraint Sub sca1esSAGR = Suppression of Aggression, 1MPC = Impulse Co ntro l, CONS = Co nsideration ofOther s, and RESP = Responsibility.

qu ent or drug-related peer rel ations are not significan t ly correlated withpersonality measures, nor are any family relations variabl es.

In summary, distress among incarcerated delinquent s is ass ocia te d with famil yrelationships, while restraint is associated with peer rela tionships.

Personality Dimension Subscales and Social R elations

I. Family relations and personality . Correlat ions between fam ily variabl es a nddistress and rest raint subscales of the WAI are summa rized in T abl e I .No nabus ive parenting is negatively associated with depression, low sclf­esteem, a nd low well being (LWB) . LWB and anxiet y subscales a re a lsoassociated with objective family healt h, although in opposit e directions: i.c. ,anxiet y (ANX) positively a nd LWB negatively. In a sim ila r fashion, ANX alsoassociat es posit ively wit h attachment ratings, whil e LWB negatively associ­ates with subjective ratings of fami ly health.

2. Peer relations and personality. See Table I. More normal peer rel ationsh ipsrelate to th e subscal e of impulse control (r = .2 1, P = .02). Conside ra tion ofothers negatively corre la tes with gang- and delinquent-ori ented peer rela-

38 J EFFERSON J OUR1"lAL OF PSYCHIATRY

TABLE 2.

Spearman Correlations between Family and Peer Relationsamong Incarcerated Delinquents

Peer Rela tions

Fa mil y Rel a ti on s Normal Gang Delinqu ent Drugs

Hca lth Fam .200* -.065 -.1 43 -. 182No n-Abus ive Pa renting .24 1** .023 - .143 - .161Attachmcnt .19 1* - .20 1* .13 1 .008Subj Family .142 -. 147 .109 - .031

t ionships (r = - .16, P = .05; r = - .18, P = .03; resp ). T he on ly distresssubscal e whi ch relates to peer relations is LWB . LWB is posi tively correlat edwith drug-abusing peers (r = .22, p = .0 I).

In summary, distress subscales are associa te d wit h fa mi ly relat ionsh ips, whil erestraint subscales a re ass ocia te d more with peer relation ships; th e only exce pt ionbeing a n associa t ion between LWB a nd drug-abusin g peers.

Family and Peer Relations

The indep endent ratings of peer and fa mily rela tio nsh ip chara ct eristics a lsoshowed some cong rue nce . Correla tions between observer ra t ings of family relationsa nd peer relation s a re summarized in T abl e 2. ormal peer relations positivelyassocia te with ratings of a ttachment st re ng th, family health a nd nonabusive paren t­ing (.19, .20, and .24, resp ). Ga ng rela tion s negatively corre la te with a ttachmen t(- .20), and drug-r elated peer rel ation s negatively correlat e with obs erver ra t ings offamily health (- .18). A trend toward a signifi cant relation sh ip bet ween drug-r elatedrelation s and child abuse is also re po rte d: r = .16, P = .059.

DIS C USSION

In consistent with pr eviously reported data in non- clinical populat ions as well asour hypothesis , restraint is not significa n t ly associa te d with family relationshipsamong this oft en-negl ect ed delinquent population (31). There usu ally is a bidir ec­tional relationship between an individual and his/her socia l relations which begins inthe home (i.e., youth ca n affect parents as parents ca n affec t youth, Patterson, 1995)(35). Prospective study of youth subjecte d to child abuse has found th at there is a nincr eas ed risk of disruptive behavioral problems including delinquen cy (36) . It maybe th at in this highly disturbed population th e in t ra psychic remnan t of familyenviro nment is found in more emo tional person ality t raits such as distress (seebelow). The lack of a relation ship between restraint a nd personali ty in our study mayvery likely represent the lack of parenting in this dom ain, i.e. , th e development ofse lf-rest rain t through effec t ive, a utho ritaria n paren ting.

PERSONALITY TRAITS IN j UVEl\'1LE DELINQ UEl\'TS 39

Restraint and its subscales prima rily rel at e to peer relationships, apparentlyhaving more of an impact on more recent soc ia lizat ion. Restrain t is positivelyassociated with ratings of normal peer relations. In pa rticul a r, th e su bscale ofrestraint , impulse contro l, was strongly correlate d with "normal" peer relat ions,support ing pr evious reports of a relationship between impulsivity a nd peer rej ection(32) . As might be expected, ga ng and delinquent peer relations a re fou nd to benegatively associa te d to th e WAI subscale of rest raint , "co nside ra tion of others."Good socia liza t ion, as reflect ed in "normal peer re la t ions," is associated with higherlevels of desirabl e person ality characte ris tics in delinquents. A diminished ab ility tocons ide r the needs of othe rs may, sim ila rly, reflect poor socia lization. This isconsis te n t with previous rep orts th at relat e indi vidu al psych ological characteristics(suc h as poor impulse cont rol) to peer relation probl em s a mo ng yout h with disrup­tive beh avior disorders (32,33,34) .

The pattern of th e relationships between social relations a nd per son al ity isinteresting. While restraint primarily relates to peer cha ra cte ris tics, distress and itssubscales primarily rel at e to family cha rac te rist ics . These dat a suggest th a t th efamily enviro nme nt leaves an impact on th ese delinquent youth emotionally. Dis­tress, a person ality trait also previou sly shown to be related to criminal historicalfactors (24), was most st ro ng ly associated with a hist ory of abuse, th e most robustfamily variable. C hild abuse has been associa te d with multiple psychologica l sequel­lae, including emo tional distress (37) . Deni al was negatively associated with childabuse. A possibl e explanation for thi s may be th at deni al requi res at least a modicumof positive family characte rist ics to be present. Severe a buse may lead to conditionswhere denial is no longer adaptive. In fact, study of abuse d child re n has ind ica tedthat denial of abuse is rather un common (38). As has been previously sugges ted byAraujo and St ein er, denial may be an effec tive defen se onl y up to a ce rtain level ofenviro nme nta l st ress (39) .

No other sign ificant associat ions were not ed between t he \VAI measure ofdistress a nd family or peer relations, including pr esen ce or abse nce of ma tern al orpaternal figures. This is a n in teresting finding (or ra ther, the lack thereof). In th ege ne ra l population, th e absence of a parental figure, wh ether by abandonment ,divor ce, or death is a pot entially devast ating in sult. Eve n in milder cases of divorce,some distress is measurabl e (40) . One might a lso expec t th at yout h involved withdrug-abusing, delinquent , or gang-related peers might a lso ex pe rie nce high er levelsof distress . The lack of distress, coupled with th e lack of deni al a nd th e difficul tfamily enviro nme nts of this group of youth, suggest s a n alte rnat ive patt ern of cop ingfor th ese delinquent youth.

While over time this distress is likely to be attenua ted, it is reason abl e to expecttha t eve n this rudimentary measure of family fun cti oning (i.e., presen ce or absen ceof maternal or paternal figu re) would be associa te d with some measurable emotionaldistress. Yet with th ese youth, it is no t as socia te d with a ny significan t dist ress orsubscale th ereof. One hypothesi s is th at th e toxic parental figures (e .g., severelyabusive or neglectful parents) are not mi ssed , either becau se of th eir toxicity orbecause of th eir negl ectful parenting. A sec ond hypothesis is th at th ere is something

40 JEFFERSON JOURJ~AL OF PSYCHIATRY

(or some thing mi ssing) wit hin th e youth which prevents emotional attachment (e .g.,a developing psych opath).

In our study, objective ra t ings of attachme nt a re posit ively associated witha nxie ty, and negatively associa te d with ga ng relationships (see tables 2, and 3, resp ).This may indicate th at anxie ty serves a different func tion for these youth , and maybe a sign of insecure a ttachme nt, as re por te d by Bowlby (I 994) (6). Without amodi cum of positive fa mily influence, the ex pected emotional rea ction of a nxietymay not develop , a nd only th e most distu rb ed peer relations ma y be possibl e.Conside r ing thi s deficit , ga ng-re la te d peer relations may indica te a misguideda t te mpt by th e youth on some level to compe nsate for a rela t ive lack of supportiverelat ionships.

Regarding th e ove ra ll lack of any significan t associat ions between distress andfam ily or peer re lat ions, it is likely th at th e severity of the abusive and neglectfulhom e situa tions, coupled with th e a ttendant trauma ta of t he cr imina l lifestyl e,cre ate a ce iling effec t th at masks va riables which might be otherwise expect ed to besignifica nt in a non-in carcerated population (see int roduct ion). Ou r explora to rydesign is limited by our measures of fami ly func t ioning , based as th ey are on obse rverrat ings of cr imi na l narratives, which may not be sensitive enoug h to discriminat ebetween th ese pr esumabl y inten sely affec ted va riab les. More sensitive family mea­sures might be used in th e future to assess th e relationship be twe e n fami ly relationsa nd person ality traits in this highl y traumat ized pop ula t ion.

Dist ress subscales are associa ted wit h va rio us fam ily relations variabl es a nd onepeer relation s variable . ANX is noted to have a posit ive, significant associa tion wit hboth obj ecti ve ratings of family health (Healt h Fam) and attachment (Att ach ).Anxiet y is usu ally conside re d to be an aversive trait, and one would not expect thata nx iety would be positively associat ed with bett er fami ly fun ctioning. On th e othe rhand, mos t of th ese fam ilies are not aveI-age, a nd many of these families a re abusiveor othe rwise dysfun ct ional. In th is case, elevated a nxi ety would be a normal reacti onto an abnormal si tua t ion, an d th e relative lack of a nxi e ty is likely to be an indi cationof anx ious insecure attachme nt.

An addit iona l cons ide ra tio n is that better fa m ily funct ion ing might in somefas hion be th e psych osocial basis for an a lt ernative to crimina l behavior; i.e ., th efa mily might provide the youth with another pe rsp ective on his or her delinquentbehavior, a persp ective th at creates internal conflict a nd subsequent anxiet y. Thisinternal conflict ma y be first ste ps on a pathway out of delinqu en cy. These consid­era t ions are not necessarily mutuall y excl usive, an d may actua lly be int errelat ed .Clearly, assumptions about causality a re beyond the power of th is st udy, but not ingthis initiall y counte ri nt uitive, positive associat ion yields in terest ing hypotheses forfu ture work.

As might be ex pec ted, low well bein g (LWB ) nega tively associa tes with obj ectiv era tings of family health and subject ive family health, as well as nonabusive par ent­ing . A particularly in te re st ing associat ion is between LWB and drug-r elat ed peerre la tions . One pri or s tudy of non-clinical popul at ions has suggeste d a pathway frome mo tional distress to ado lescen t d rug use via the choice of substance-using pee rs

PERSO NALITY TRAITS IN J UVENILE DELINQUENTS 4 1

(4 1). Further study might add res s th e qu estion of th e relation shi p between LWB a nddrug-related peer relation s. Man y fac tors may con t r ibute to th is relationship includ­ing th e effec ts of drug abuse on th e individual , se lf-me d icat ion of unpleas antemo tional states, and/or un satisfyin g peer relation s.

Ne ithe r restraint nor any of its subscales a re found to be sign ifican tly assoc iatedwith subs tance-usi ng peer rela t ions in our study. Restrain t has been pr eviouslyre po r te d in non -clinical population s to be pr edi cti ve of subs tance abuse (42) , as hasbehavioral se lf-regula tion (BSR) (43). BSR has been proposed as a potential factor inth e choice of delinquent a nd substance-a busing peers, and as a prod roma l factor forsubsta nce abuse a mo ng popula tion s at high risk for substance ab use. In ou r study,ge ne ra lly low restraint (the mean restraint level in our st udy is 98, which is th e 20t h

percentile for norms based on non-clinical populat ion s) (30) may have cre a te d abasement effec t, whi ch masks th e influence of rest raint on the choice of substanceabusin g peers.

While, again, co rre lat ion does not infer ca usa lity, this association yield s so meinteresting hypotheses relating person al we ll being (or the lack th ereof) , anddrug-relat ed peer relations . It may be that in some fash ion , a sense of LWB mightcont r ibu te to th e cho ice of subs tance -us ing peers, a nd potentiall y substance abuse.Peer pr essures a re clearly th e greates t proximal corre la te of adoles ce n t subst an ceuse (44). In add it ion, subst ance use (a nd particul arly misuse) may cont r ibu te to aperson al se nse of LWB , which might maintain a vicious cycle of LWB lead ing tosubstance use, leading to a wor sened person al LWB , and so on . T hese hypo th esesmi ght be ex plore d with further study of this relation shi p between this pa rticular typeof distress a nd peer choice s, as we ll as subs tance use.

The weak correla tions between family environment and our personality mea­sures a re worthy of furth er cons iderat ion. The relations are overwhelmingly nega­tive, with th e most robu st variable being nonabus ive paren ting, i.e., th e a bse nce ofa buse. Con sid ering th at th e most helpful fa mil y m easu re is the lack of abuse, andth at the re was no positi ve effect on eithe r restra int or di stress if eit he r parent waspresent , speaks very little for resilien cy fac tors in th is environment. In fact , it pointsto th e subj ects ' lack of socia l support. Mor eover, a t tachm ent in a more healthy famil yincreases a nx ie ty in our highl y disturbed population. O nly 2 signifi cant corre la tions(out of 24 total ) a ppeare d to be helpful to th e emotio nal state of the deli nq uen ts inour study, a nd th ese were a mo ng th e weakest. While not initiall y hypothesized, th isrelative lack of signifi cant relation ships between fami ly a nd thes e personalit y trait sis reason abl e cons ide ring th e ex te n t to whic h th ese youth have been neglected a nd/o rabused.

The lit erature suppor ts th e premi se that th e fam ilia l env ironment is to someex te nt internalized, in multiple non -clinical stud ies a nd in at least on e severelydelinquent sa m ple . Ru chkin et a l (1998) report substant ia l re lat ionships betweenpa renting practi ces and personality dimen sion s a mong their sample of incarcera teddelinquents in No rt he rn Ru ssi a (45). While Ru chkin et al fou nd a rela t ionsh ip, th ege ne ralizability to our sa m ple is qu estionabl e, pa rticul arly consideri ng that theyexclude families wi th less th an 2 parents. The m ajority of our subjects in our sample

42 J EFFERSON JOUR1'JAL OF PSYCHI ATRY

had one parent figure, an d this pe rso n was oft en abus ive. For th ere to be a st rongerrel ation ship, parents have to at least be pr esent , a nd maybe eve n help ful. Thedim ini sh ed rel ation ship bet ween fa mil y va ria bles a nd pe rsonality traits may actu all yindi cate the lack of fun cti on al paren ting.

There is some relat ionship betwee n family a nd peer rela t ions. In our s tudy, peerrel ation s a re corre la te d with fa mily re la tions, and healthy fami ly variables a re moreassocia te d with normal peer rel ation s. No rmal peer re lationships are posi tivelyas socia ted with fa mily health , a ttachment and nonabu sive pa ren ting. In addi t ion,ga ng-types of peer rel ation s a re associated nega tively with attachment. In effecti veparental monitoring is highl y associat ed with yout h involvem ent with deviant peers(46). Diffi culties with attachm ent or othe rwise a tt e nua ted pa rental re lationships(wh ether by abuse or neglect ) a re likely to play a rol e in th e choice of deviant peers.While our m easure of attachment is not sophisticat ed e no ug h to det ermine types ofattachment organizations, a dismissing attachment org a niza t ion has previously beenass ocia te d with conduc t and subs ta nce abuse disorders in hospitali zed adolescent s(47). While we need to be ca re ful not to infer beyond th e associa t ion, ou r findin g ofth e significan t relation ships between peer relations a nd a ttachment is consiste n twith a mod el of socia l development where ea rly family expe rie nce is intimat elyinvolved in th e development of ext ra -fa milia l relation sh ips. T he ea rly risk factor ofpoor family relationships appears to cascade down th e developmen tal pathway,compounding risks along th e way.

Althou gh preliminary, a nd in need of ve rificat ion with longitudi na l st udy, thi sst udy is cons iste nt with a mod el of pe rson ality developmen t where early familyexperiences a re internalized a nd t he conseque nt person al ity structures regul at e th edevelopment of di fferent types of ext ra-fa milia l relationships, i.e. , normal , ga ng-,delinquent- or drug-r elated . This research ex pands th e cum ula t ive knowledge re­ga rd ing th e social and in terpersonal corre la tes of ind ivid ual trai ts pr edi cting crim­inal recidi vism, a nd provides lead s for further study of the development of pee rrelation ships in crim ina l delinquen cy.

ACKNOWLEDGMENTS

This research is supporte d by The Californ ia Wellness Founda tion 's grant to Dr.St ein er. We would also like to thank Karen Thomas and Marcia Scipio for provid ingratings of ward narratives.

REFERENC ES

I. Ste ine r H. Practice pa rameter s for th e assessment and tr eatment of child re n andadolescents with conduct d isord er. J Ame r Acad C hild Adolesc Psychia try 1997; 36:IS-1 8S.

2. St einer H, Cauffman E.Juven ile justic e, del inq uency, and psych iat ry. Child & AdolescentPsychi at ry C linics of North Am erica 1998; 7: 653-672.

PERSONALITY TRAITS IN JUVENILE DELINQUEl\'TS 43

3. Fink elh or D, Berl iner L. Research on th e treat ment of sex ua lly abused ch ildre n: a reviewa nd recommendat ions . J Am Acad Child Adolesc Psychiat ry 1995; 34: 1408-1 423.

4. Mill er LS, Wasserma n GA, Neugebauer R, et al. Wi tnessed com m unity violen ce a nda nt isocial beh avior in high -r isk, ur ba n boys. J C lin C hild Psychol 1998; 28: 2-11.

5. Aicho rn A. Waywa rd Youth . New York: Viking Press, 1935.6. Bowlby J . Fourty-fou r juvenile th ieves: thei r characters an d home-life. In t J Psychoa na l

1944; 25: 19- 53.7. Tyson P, T yson RL. Psychoan alyti c T heori es of Development : An In tegra tion. New Haven ,

CT: Yale Unive rsity Pr ess, 1990.8. Bandura, A. The Soc ia l Founda tions of T hough t and Act ion. En glewood Cliffs, NJ:

Pr en tice Hall , 1986.9. HarrisJR. T he Nurt ur e Assumptio n: Why C hildren Turn Out th e Wa y that They Do. New

York : Free Pr ess, 1998.10. Shaw DS, Ow en s EB, VondraJI , ct a l. Ea rly risk factors and pathways in th c developm ent

of early di sruptive beh avior problem s. Dev Psych opat hol 1996; 8: 679-699.II . Loeber R, Sto u thame r-Loe be r M. Fa mi ly factors as cor relates and pr edi ctor s of j uven ile

conduc t pr obl ems and delinquen cy, in Cr ime a nd J us tice : An Annua l Review of Resear ch ,Vol. 7. Ed ite d by T onry M, Mor ris N. C hicago: Unive rsity of Chicago Pr ess, 1986: 29-149.

12. Patterson GR, Reid JB, Dishi on °1]. A develo pme ntal perspective on ant isocial beh avior.Am Psych ol 1989; 44: 329-335.

13. Ellio tt DS, Huizin ga D, Aget on SS. Explai ning Del inqu en cy and Drug Use. Newbu ry Park,CA: Sage, 1985.

14. DiLalla LF, Mitch ell CM, Art hur MW , et al. Aggression an d del inq uen cy: family a ndenviro nmental factors. ,] Youth Ado lesc 1988; 17: 233-246.

15. Ainsworth MD S. Attachments beyond in fancy. Am Psych ol 1989; 44: 709-716.16. Pa rk e RD , Ladd GW. Fam ily-Peer Relationships: Mod es of Lin kages. Hillsdale, J\!J:

Erl ba um, 1992.17. Thienneman M, Shaw Rj, Ste iner H. Defen se style a nd fami ly environment. C hild

Psychi at ry Hu m Dev 1998; 28: 189-1 98.18. Bra ck G, Mathen y KB, Gay MF. Relati onshi ps bet ween a ttachm ent and coping resources

a mo ng lat e ado lescen ts . J Co llege Student Dev 1993; 34: 2 12-2 15.19. Frau en glass S, Roth DK. Family suppo rt decreases influe nce of deviant peers on Hispanic

ado lesce nts ' substance use.,] Cl in C hild Psych o] 1997; 26: 15- 23.20. Fe ld ma n SS, Weinberger DA. Self-r estrai nt as a med ia tor of fam ily influen ces on boys'

delinquent beh avior : a lon gitudinal study. C hild Dev 1994; 65: 195- 21 1.2 1. Weinberger DA, Schwa rtz GE. Distress & rest rai nt as su perord inate dimensions of

sel f-re ported adj ust me nt : a typo logica l persp ecti ve. j ou rn al Pers 1990; 58: 38 1-417.22. Huckaby W,J, Kohl er M, Garner EH, et a l. A compar ison between the Weinb erger

Adjustmen t Invent ory and the Minnesota Mul tiphasic Person al ity Inventory with incar ­cera te d delinquent males. C hild Psychi atry Hum Dev 1998; 28: 273-285.

23. Tinklenberg JA, Steiner H , H uckab y Wj , et al. Crimina l recid ivism pr edi ct ed fro mnarra tives of violent j uveni le delinque nts . C hild Psych iat ry Hu m Dev 1996; 27: 69-79.

24. Steine r H, Will iam s SE, Benton -H ardy L. Violen t crime paths in incarcerated juven iles:psych ological , environmenta l, a nd biological fac tors, in Plenu m Publ ish ing Co ., NY, 1997.

25. St einer H, Cauffman E, Du xbury E. Person ality t raits in j uven ile del inqu en ts: relation tocr imina l beh avior a nd recidivism .J Am Acad C hild Ado lesc Psychia try 1999; 38: 256-262.

26. Achenbach TM, Edel brock C . Manual for th e C hild Beh avior C heck list and Revised C hildBehavior Profile. Burlingt on: Univers ity of Vermont, 1983.

44 J EFFERSON JOURNAL OF PSYCHIATRY

27 . Ste ine r H. Ind ividual psycho the ra pe utic a pproaches to violent youth. Lectu re a t InstituteII of th e 43 rd a nnua l meeting of th e Am eri can Acad em y of Child a nd AdolescentPsychiat ry, O ct ober 1996 .

28. St einer H, Garcia IG , Matthews Z. Posttraumatic st re ss di sorder in incarcera ted j uveniledelinquents. J Am Acad Child Adol esc Psychi atry 1997; 36: 357-365.

29. St ein er H , Feldman SS. Two approac hes to th e mea surem ent of adaptive style: compa r­ison of normal , psychosomat icall y ill, a nd del inquent ado lescent s. J Am Acad C hildAdol esc Psychi atry 1995; 34: 180 -1 90.

30. Weinberge r DA. Distress a nd self-res traint as measu res of adjustmen t across th e life­spa n: confirmato ry fact or a na lysis in clinical a nd non -cli nical sa mples. Psych ol Assess1997; 9: 132-1 35.

3 1. Feldman S, Wentzel K. Relat ion s bet ween family inte ract ion patt erns, classroo m self­restraint a nd acade m ic achi evem ent in pr e-adolescent boys . J Edu c Psychol 1990 ; 82:813- 819 .

32. Hinshaw SP, Zupan B, Simmel C , et a l. Peer sta tus in boys with and wit hout a tte ntiondeficit hyperactivity disorder: pr edi ctions from overt a nd covert a nt isocia l beh avior, socia lisol ation , and authoritative parenting beli efs. C hild Dev 1997; 68: 880-896.

33. Gu evevr em ont DC , Dumas rvIC. Peer rel ation ship problems a nd disru pt ive beh aviordisorders. J Em otional Beh avior al Disorders 1994; 2: 164 -1 72.

34. Landau S, Mili ch R, Dien er MB . Peer relat ion s of child re n wit h attent ion deficit hyper­ac tivity dis order. Reading and Writing Quart erly: Overcom ing Lea rn ing Disabilit ies 1988;14: 83- 105.

35 . Patt erson GR. Performan ce Models for Parenting: A Socia l In terven tiona l Perspective.ew York: Wil ey and Son s, 1997.

36 . Robins LN, Rutt er M: Strai ght a nd Deviou s Pathways fro m C hild hood to Adu lth ood .Cambridge: Cambridge Universit y Press , 1990.

37. Bri ereJN, Elliott DM. Im m ed ia te a nd long term impact s of child sex ual abus e. Fut ur e ofChildren 1994; 4: 54-69.

38. Bradl ey AR , WoodJM: How do children tell ? The di sclosure process in child sexua l abuse .Child Abuse Neglect 1996; 20: 88 1-89 I.

39. Arauj o K, St ein er H. Adolescen t defen se sty le a nd life stressor s. C hild Psychia try H umDev 1999; 30( 1): 19 - 28.

40. Rodgers B: Social a nd psych osocial wellbeing of children from d ivor ced families: Aus tra­lian research findings. Aust Psych ol 1996; 3 1: 174-1 82.

4 I. Swa im RC , O etting ER, Edw ards RW , et al. Links fro m emot iona l d istress to ad olescentdrug usc: a path mod el. J Consult C lin Psychol 1989; 57: 227-23 I.

42. Farrell AD, Danish SJ. Peer drug as socia t ions and emot iona l rest rain t: ca uses or conse ­qu ences of adolescents' drug us c? J Consult Clin Psychol 1993; 6 1: 327-334.

43. Dawes MA, T arter RE , Kirisci L. Behavioral se lf-regulat ion: correla te s a nd 2 yea rfollow-ups for boys a t ri sk for substance abuse. Drug Alcoh ol Dep end 1997; 45: 165-176.

44 . Dishi on TJ , Loeb er R. Adolescent marijuana a nd alcoho l abuse: the role of paren ts an dpeers revisit ed . Am J Drug Alcoh ol Abu se 1985; II : 11-25.

45. Ru chkin VV, Eise mann M, Hagglof B, e t a l. In terrela tions between tempera ment,character a nd parental rearing in male delinquent ado lescents in north ern Russia. Co mprPsychi at ry 1998; 39: 225-230.

46. Dishion TJ , Capa ldi D, Sprackl en KM, et al. Peer eco logy of mal e ado lescen t drug lise. DevPsychopathol 1995; 7: 803-824.

PERSONALITY TRAITS IN JUVENILE DELI 'QUEf\'TS 45

47. Rosenstein DS, Horowitz HA. Adol escent a ttachme nt and psychopathology.] Consu lt ClinPsychol 1996; 64: 244- 253.

48. Wein berger DA, Barth olcmew K. Social -emotional adjustment and patt erns of alcohol uscamo ng young adults. ] Pers 1996; 64: 495-527.

49. McCrae RR, Costa PT ]r. Val idat ion of th e five-factor model of persona lity acrossinstrumen ts and observ ers. ] Pers Soc Psychol 1987; 52: 81-90 .

50. Stein er H, WilsonlJ. Conduct disorder, in Disru ptive Behavior Disorders in Children andAdol escents. Washington, DC : Am eri can Psychia t ric Pr ess, 1999: 47-98.