Adolescent Attachment to Parents and Peers Andrea L. Barrocas The Emory Center for Myth and Ritual in American Life Working Working Paper o. !"
Abstract Attachment #onds e$ist in relationships relationships across the lifespan. Adolescence may #e a particularly crucial period for attachment relations. As relationships relationships %ith parents shift
&
and those %ith peers gain importance' importance' patterns of attachment may change as %ell. %ell. There is a huge gap in the attachment literature on the utili(ation of mother' father and peers attachment figures' specifically ho% attachment to parents relates to that %ith peers and' importantly' ho% adolescents are attached to their fathers as compared to their mothers. This study e$plores these patterns of attachment in adolescence. T%enty)four T%enty)four racially di*erse' mostly middle class adolescents+ ,grades eight and ten-' mothers+ and fathers+ attachment %as measured using the n*entory of Parent and Peer Attachment ,PPA,PPA/uestionnaire. 0*erall' adolescents rated attachment to mother and father similarly' similarly' suggesting generali(ation of attachment representations' #ut they rated attachment to peers' especially on the communication dimension' higher than to parents' suggesting a rise in the importance importance of peers. mplications of these findings findings for adolescent attachment attachment relationships are discussed.
Adolescent Attachment to Parents and Peers Adolescence is a ma1or transitional period in a person+s person+s life. With With the onset of pu#erty come not only physical changes' #ut also many other important changes. 2or e$ample' the social %orld of an adolescent changes to #ecome more peer focused than #efore. 3o%e*er' parents do not disappear from the daily life life of an adolescent. Therefore' a ma1or task of adolescence is to figure out %hich people can help satisfy %hat need. Attachment theory may #e helpful in understanding some of o f the patterns that e$ist across the transition to to adolescence. 4pecifically' 4pecifically' this pro1ect %ill %ill use attachment theory to understand the relationships that adolescents ha*e %ith % ith their mother' father and peers. Attachment Theory Attachment theory originated as an e$planation of the #ond that e$ists #et%een an infant and the primary caregi*er' typically typically the mother ,Bo%l#y' &566-. This #ond not only is important for general %ell #eing ,e.g. 7reppner 8 9llrich' &556- #ut also functions as a template for all relationships across the lifespan ,e.g. Waters' Waters' Merrick' Tre#ou$' Cro%ell' 8 Al#ershein' Al#ershein' :"""-. As early as infancy' infancy' children can mentally represent their attachment figures and construct ideas and e$pectations for relationships %ith #oth these original figures figures and others. Bo%l#y called this the internal internal %orking model of attachment. Attachment in infancy is conceptuali(ed as distinct #ut integrated #eha*iors that are e$hi#ited #y the infant in response to the caregi*er+s #eha*iors. There are three attachment related #eha*iorsi that define ho% attachment is seen across the lifespan; pro$imity seeking ,pro$imity maintenance-' separation protest' and secure #ase ,3a(an 8 4ha*er' &56<= 4egrin 8 2lora' :""!-. Pro$imity seeking descri#es the the physical closeness of infants to their caregi*er. 4eparation protest refers to the un%illingness un%illingness to separate from the caregi*er' %hich is translated through #eha*iors such as crying. When an
:
attachment figure is a child+s secure #ase' he or she is utili(ed as a foundation from %hich to e$plore the en*ironment and seek out non)attachment non)attachment related pursuits. Related to the secure #ase phenomenon is the idea of an attachment figure #eing a safe ha*en #ecause of representations of the attachment #ond. Representing an attachment figure as a safe ha*en' conse/uently' means that the infant goes to the caregi*er %hen distressed or in need of comfort or support' thus using the attachment figure as a secure #ase. Individual Differences. Early attachment theorist Mary Ains%orth e$plained attachment #eha*iors in infancy as an organi(ational construct' one that relies on the /uality of the primary caregi*er+s ,typically the mother- response to the infant+s #eha*iors that then influence the infant+s responses and interactions ,see 4roufe 8 Waters' Waters' &55<-. >ifferences in reactions and interactions lead to indi*idual indi*idual differences in the infant+s security. security. 2or e$ample' mothers %ho respond in a sensiti*e manner to their infants ha*e infants %ho think they %ill #e taken care of ,for more on this topic see 7aren' &55?-. More specifically' specifically' in mother)infant interactions interactions of this type' the infant learns that the mother is is a steady and secure person to to go to for comfort. comfort. n other %ords' the mother is percei*ed as a secure #ase for the infant. infant. Therefore' it %as thought that there %ould #e differences in ho% infants respond to and process caregi*ers+ #eha*iors' and that infants %ould sho% differences in their o%n #eha*iors as %ell. Ains%orth de*eloped a %ay to measure indi*idual differences in a testing situation called the 4trange 4ituation. 4ituation. n the 4trage 4ituation' the the infant and mother interact' generally in some form of play' and the infant is allo%ed to e$plore the surroundings ,for more detail see 4roufe 8 Waters' Waters' &55<-. The mother then lea*es the room and the infant is alone. At this point' in most most cases' a non)familiar adult enters. 2ollo%ing this' there is a reunion of mother and infant. infant. 0#ser*ing each point of change in the 4trange 4ituation can help e$plain the security of the infant in*ol*ed #y looking at the le*els of stress and comfort that the infant and mother e$hi#it. Children are classified in threeii categories according to their #eha*iors during reunion episodes; secure' an$ious)a*oidant and an$ious)am#i*alent. an$ious)am#i*alent. A secure infant seeks comfort from the caregi*er #ecause of representations that the caregi*er has #een and can #e used as a secure #ase' and e$plores the en*ironment %ith ease ,3a(an 8 4ha*er' &55?-. An an$ious)a*oidant infant does not sho% distress during the separation and upon reunion %ith the caregi*er a*oids contact due to conflicting representations of the caregi*er ,3a(an 8 4ha*er' 4ha *er' &55?= Waters' Waters' 3amilton' 3amilton' 8 Weinfield' Weinfield' :"""-. An an$ious)am#i*alent infant seeks the comfort of the caregi*er' yet is not soothed u pon reunion ,3a(an 8 4ha*er' &55?-. Stability of Attachment Over Time. Attachment researchers ha*e e$amined attachment #eyond infancy. Rather than looking at attachment #eha*iors' #eha*iors' ho%e*er' researchers ha*e looked at attachment representations. representations. t is thought is that one creates mental representations of ho% to interact %ith others' termed the internal %orking model of attachment' #ased on pre*ious attachment related relationships and interactions. Bo%l#y ,&566- #elie*ed that starting in infancy a child internali(es patterns of relating to people' generally the parents' and forms an idea of ho% to relate to others #ased on these representations. Through early interactions interactions %ith caregi*ers' caregi*ers' children internali(e internali(e and organi(e their understanding understanding of relationships relationships ,Laursen 8 Collins' :""?-. Each attachment relationship shapes the child+s mental schema and leads to the de*elopment of e$pectations for future relationships and interactions.
@
A #asic tenet of attachment theory is that it is sta#le o*er time' ho%e*er research yields mi$ed findings. Longitudinal studies ,e.g. 3amilton' :"""= Walters Walters'' 3amilton' 8 Weinfield' Weinfield' :"""- measured #oth infant attachment security status using the 4trange 4ituation and later attachment in adulthood. adulthood. Researchers found strong retention retention rates in classification of attachment. attachment. n other %ords' for most indi*iduals there %as continuity of attachment= ho%e*er' insta#ility and change in attachment classification for those %ho did change %ere e$plained #y significant life life e*ents. There %as *ery little little change in classification occurring for those %ho did not ha*e significant life e*ents ,e.g. Easter#rooks' &565= 3amilton' 3amilton' :"""= Walter Walterss et al.' :"""-. Thompson ,:"""- argues argues that security of attachment %ill remain sta#le only if other aspects of life that are related to attachment remain sta#le across transitions. 3e states' specifically' specifically' that if /uality of parental care is sta#le and de*elopment of solid self)concept and self)esteem occur' one is likely to remain securely attached to others. 0n the other hand' yet still supporting Thompson+s argument' some researchers found that attachment is not sta#le. 2or e$ample' Le%is' 2eirin' and Rosenthal ,:"""found no relationship #et%een attachment security status in infancy and adolescence. n the study' classification classification changed for a#out half ha lf of the participants. n addition' di*orce %as a huge mediating factor for change in attachment status' sho%ing that the internal %orking model of attachment can #e changed due to attachment related e$periences. Thompson+s ,:"""- argument supports these findings #ecause of o f his #elief that security can remain sta#le if there is sta#ility in relationships and /uality of care' #ut inter*ening occurrences that change these factors can cause security security to shift. These findings suggest that ne% e$perience #uilds upon pre*ious e$periences to create fle$i#le representations of ho% relationships are e$pected to #e. Stability of Attachment Across Caregivers. Caregivers. A related issue is the concordance of attachment' defined as the sta#ility of attachment #et%een # et%een attachment figures ,i.e. mother' father' peer' peer' si#ling-. Research on the concordance of children+s+ attachment to mother and father focuses primarily primarily on relationships during infancy. infancy. 4uch studies sho% mi$ed mi$ed findings. 4e*eral point to%ard strong concordance #et%een attachment attachment to mother and father ,e.g. Easter#rooks' &565- and others suggest that the mother)infant and father) infant attachment relationships are are independent ,Main 8 Weston' Weston' &56&-. Easter#rooks ,&565- found a rate of <" concordance #et%een the attachment relationships that :" month)old infants had %ith their mothers and fathers' %hich strongly supports the notion of concordance. 0ne e$planation for the strong strong rate of concordance is that parents parents %ho are more similar in childrearing approaches' such as sensiti*ity and a*aila#ility' %ill ha*e children %ho are attached similarly to #oth parents. More importantly' in an analysis of && studies of attachment that measure classification %ith Ains%orth+s Ains%orth+s 4trange 4ituation' 2o$' 7immerly' and 4chafer ,&55&found o*erall support for concordance of attachment to mother and father= those infants %ho %ere securely attached to their mother %ere more likely to #e securely attached to their father ,the same same patterns %ere found for insecurity-. insecurity-. These findings suggest that that children do internali(e representations of relationships and attachments and form e$pectations for other close relationships. relationships. 0n the other hand' some of the studies e$amined suggested a lack of concordance of attachment #et%een caregi*ers ,2o$' 7immerly' 7immerly' 8 4hafer' &55&-. 2or e$ample' Main and Weston Weston ,&56&- found that mother) mother) infant and father)infant attachment %ere not dependent on one another. another. They argue that
?
all relationships are different. different. Although mother)infant mother)infant and father)infant relationships relationships do interact %ith one another' mothers and fathers each ha*e specific %ays of raising and relating to their children. children. To date' concordance has only #een e$amined in infancy and early childhood' lea*ing concordance of attachment in adolescence as yet une$plored. Because of the de*elopmental changes that occur in adolescence' this period seems to #e one in %hich it is important to look at ho% attachment to each figure is related. Adolescent Development Adolescence is a period of significant cogniti*e' social and #eha*ioral transitions. Cogniti*ely' there are huge gains in reasoning and perspecti*e taking skills' as %ell as ac/uisition of #etter emotional emotional understanding. 4ocially' 4ocially' peer relationships #ecome much much more important than #efore. Physically' Physically' pu#erty #egins' sparking hormonal hormonal and physical changes. The de*elopmental changes that that occur in early early adolescence are related related to one another. mportantly' mportantly' they impact the desire for a more independent and autonomous life that comes %ith adolescence. t is during this this stage that an indi*idual indi*idual de*elops a more mature sense of identity ,Erikson' &56- #ecause of o f such ad*ances in #eha*ioral' social and cogniti*e realms ,3a#ermas 8 Bluck' :"""-. 2or e$ample' a#stract thinking allo%s for ne% thoughts related to identity ,e.g. Who am D What do likeD Who do stri*e to #eD-' %hich is one of the main transitions to adolescence= ho%e*er' it is /uestiona#le ho% much this affects the parent)child relationship in adolescence ad olescence ,Collins 8 Repinsky' &55?-. These social de*elopments ha*e the most most significant implications implications for adolescent attachment' although the #iological and cogniti*e changes are important as %ell. Parent-Adolescent Interactions. Interactions. The greatest markers of de*elopmental changes in the self in adolescence are seen through independence' autonomy and detachment from caregi*ers ,Erikson' &56= Ryan Ryan 8 Lynch' &565-. Ryan Ryan and Lynch ,&565- found that adolescents stri*e for more autonomy and indi*iduation from parents than #efore period of de*elopment and there is a higher higher le*el of detachment from parents. parents. Collins and Repinsky ,&55?- note that the amount of physical time that parents and their children spend together decreases during adolescence as %ell. %ell. Although there is is o#*ious physical distancing from parents' adolescents still sho% a desire for high le*els of support from them ,2urman 8 Buhrmester' &55:= oller' oller' &55?-. Representations of interactions %ith parents may' then' then' pro*ide a support #ase for adolescents. 3ill' 2onagy' 4afier' and 4argent ,:""@- state that communication in the family must #e' in the %ords of 4egrin 4e grin and 2lora ,:""!-' synchroni(ed and reciprocal for optimal attachments #et%een family mem#ers. mem#ers. 2amily 4ystems Theory states that there are smaller dyadic interactions that are separate from one another' #ut that also function %ithin the larger family unit ,Caffery 8 Erdman' :"""= 3ill' 2onagy' 4afier' 4afier' 8 4argent' :""@= 7reppner' :"":= 4egrin 4egrin 8 2lora' :""!-. Accordingly' Accordingly' as adolescents de*elop and #egin to search for autonomy and independence' there must #e a response to this change #y the family as a %hole. oller ,&55?- notes that families %ith %ith adolescents are constantly and increasingly renegotiating family roles that are #uffered #y ope n and fle$i#le communication. n addition' Allen Allen and Land ,&555- speak of the ongoing renegotiation that occurs in terms of family goals. They argue that that an adolescent %ith %ith secure attachment %ould #e part of a system %here goals are constantly re)set and the family mem#ers+ needs %ould #e in sync one another. Adolescent elationships !ith "others #ers #ersus us $athers. >oherty and Beaton ,:""?- e$plore the importance of looking at mother)child and father)child relationships
!
#oth separate from one another and as dyads that interact in a system' since each relationship has different /ualities /ualities that may impact later outcomes. As sho%n %ith the research on concordance' children can sometimes ha*e differing attachments to mothers and fathers. n addition' children start to e$perience their relationships relationships %ith their mothers mothers and fathers in differing differing %ays. 0#ser*ing parent)infant interactions interactions %hile playing' Leaper ,:"""- found that play %ith mother and father differed' according to #oth gender of parent and of child. This suggests that each parent contri#utes differently differently to children+s de*elopment and' importantly' points to%ard the differing influence that parents ha*e on social outcomes' sparking interest in e$ploring mothers and fathers impact on children+s outcomes. Adolescents ha*e #een #rought up spending more time %ith' and engage in more open sharing of emotion' %ith mothers than %ith fathers ,Laursen 8 Collins' :""?-. Paterson' 2ield' and Pryor ,&55?- supported earlier research that points to%ard the importance of mothers for attachment attachment related outcomes in adolescence. 4pecifically' 4pecifically' adolescents reported higher le*els of the /uality /ua lity of affect to%ard their mothers than fathers ,e.g. Founiss Founiss 8 4mollar' &56!-. Lei#erman' >oyle' and Markie%ic( ,&555- also found that percei*ed maternal a*aila#ility is important across the transition from childhood to adolescence' and an d the /uality of the mother)adolescent relationship is strongly linked to attachment security security as %ell ,Allen ,Allen et al.' :""@-. Allen and colleagues colleagues found that maternal #eha*iors' such as support and attunement' predicted security in the mother)adolescent relationship. relationship. Additionally' Additionally' they found that 5th and &"th graders %ho %ere more securely attached %ere #etter a#le to intellectually and emotionally use their strong relationship %ith their mother as a #ase for e$ploration and autonomy. More recent research has sho%n that older children differ in their utili(ation of mother and father as %ell as the /uality of affect to%ard mother mother and father. 2or e$ample' 3unter and Founiss Founiss ,&56:- and Paterson' 2ield' and Pryor ,&55?- found that adolescents rely on mothers for support support more than their fathers. fathers. This might #e #ecause of e$pectations a#out the roles of their mother and father that ha*e #ecome ingrained in the adolescent. Conse/uently' a child+s child+s mother mother and father are t%o different people %ho interact %ith and influence their children in distincti*e %ays. As sho%n' it is o#*ious that a strong #ond e$ists #et%een mothers and adolescents' ho%e*er the importance that the father plays in the child+s life should not #e o*erlooked. Recent research suggests that that fathers also are significant significant attachment figures for adolescents ,e.g. Williams Williams 8 7elly' 7elly' :""!-. t has #een sho%n that the father) adolescent relationship is related to se*eral attachment constructs= co nstructs= specifically' specifically' %armth' closeness and a*aila#ility ,Ca#rrera' Tamis)LeMonda' Tamis)LeMonda' Bradley' 3offerth' 8 Lam#' :"""-' and attachment to father significantly predicts friendship conflict for adolescents ,Lie#erman' >oyle' 8 Markie%ic(' &555-. Adolescent)father attachment impacts adolescent ad1ustment in a different %ay than does mother)adolescent attachment. Williams Williams and 7elly ,:""!- compared #oth mother)adolescent and father)adolescent relationships' finding that although there %ere more secure mother)adolescent attachment relationships' the father)adolescent attachment relationship %as related to adolescents+ #eha*ioral pro#lems. 4pecifically' 4pecifically' more paternal in*ol*ement in the parenting process %as related to more security of attachment. This suggests that fathers do play an important role for attachment related outcomes for
adolescents. t is un/uestiona#ly important important to study adolescent attachments to the fathers as %ell as to their mothers. Considering the importance of the reciprocal nature of the parent)adolescent attachment #ond' to date' there is not much research on ho% mothers and fathers are attached to their adolescents. There is information a#out ho% adolescents are attached to their parents' as %ell as ho% mem#ers of the family impact one another' ho%e*er the %ay parents are attached to their adolescents is an almost o*erlooked topic. >eko*ic and Buist ,:""!- ha*e 1ust recently found a relationship #et%een parental ratings of their attachment relationships %ith their children. t is important to e$plore this further' and especially' to e$tend this idea #y looking at the patterns among the consistency of attachment #et%een mem#ers of the same family. Adolescent-Peer elationships Throughout de*elopment' children create emotional #onds %ith not only their parents' #ut %ith other indi*iduals as %ell. Parent)child relationships impact social de*elopment' such as the creation of peer relationships. 3a(an and 4ha*er ,&55?- state that children must create #onds %ith other a*aila#le figures and' as de*elopment progresses' peers #ecome e$tremely important attachment figures. These ne% peer relationships' ho%e*er' look different than those %ith parents ,e.g. 2reeman 8 Bro%n' :""&= ickerson 8 agle' :""!-. 2or e$ample' adolescents #egin to spend spend less time %ith parents and much more time %ith their peers ,Collins 8 Repinsky' Repinsky' &55?-. 7erns ,&55?argues that the forming of closer peer #onds allo%s adolescents to e$plore independence from parents. Additionally' Additionally' ickerson and agle ,:""!- found that adolescents go to their peers in times of need ,pro$imity seeking #eha*iors#eh a*iors- more than #efore entering this de*elopmental period. mportantly' mportantly' peer relationships e$ist in different settings outside the family. family. Thus' the relationships that children ha*e %ith friends allo% for furthering of social de*elopment ,4a*in)Williams ,4a*in)Williams 8 Berndt' &55"-. E$amining the /uality of friendships friendships in adolescence' 4a*in)Williams and Berndt ,&55"- stated that #oth positi*e /ualities of peer relationships' such as trust and support' and negati*e /ualities' such as 1ealousy and resentment' help %ith de*elopment in social social and personal realms. Many researchers ha*e studied correlates of positi*e friendships and ha*e found relations to self)esteem ,e.g. Green#erg' 4iegel' 8 Leitch' &56@- and lo%er le*els of loneliness ,4a*in)Williams ,4a*in)Williams 8 Berndt' &55"-. Additionally' Additionally' Weimer' Weimer' 7erns' and 0lden#erg ,:""?- found a relationship #et%een positi*e friendship /ualities in a #est)friend dyad dyad and security of each partner in the dyad' suggesting that those dyads d yads %ith more security are made up of o f friends %ho feel #etter a#out not only the friendship #ut themsel*es as %ell. Better communication #et%een dyad partners %as also related to more security in the dyad' supporting the importance of communication for attachment as mentioned #efore' ,Weimer' ,Weimer' 7erns' and 0lden#urg' :""?-. As adolescents seek autonomy and independence from their parents' they turn to peers more than #efore ,ickerson 8 agle' :""!-. 2urman and Buhrmester ,&55:looked at important changes in peer relationships across the transition from childhood to adolescence. Most importantly' importantly' they found that support)seeking support)seeking needs are fulfilled fulfilled less #y parents and more #y peers as childhood ends and adolescence #egins. Founiss Founiss and 4mollar ,&56!- state that this change in utili(ation of p eers might occur #ecause ha*ing
<
high)/uality friendships ser*es to fulfill the social needs that emerge in adolescence. This does not undermine the importance of the parent)adolescent relationship' #ut points to%ard a gain in importance and influence of peer relationships for positi*e de*elopment. Related to this' parental understanding and fle$i#ility are related to adolescents+ friendship satisfaction and general %ell)#eing ,4illars' 7oerner' 8 2it(patrick' :""!-' %hich supports the argument that fle$i#ility is among the most important aspects of the relationships' as it is e$tremely important for communication ,Laursen 8 Collins' :""?-. elationship %et!een Parent and Peer Attachment Attachment Peers are central to adolescent de*elopment and social life. Because adolescents e$perience close #onds %ith peers' it is imperati*e to look at adolescent)peer attachments in con1unction to those %ith parents. parents. Easter#rooks and Lam# ,&5<5- found a relationship relationship #et%een mother)infant attachment ,using Ains%orth+s Ains%orth+s 4trange 4ituation- and peer competence at the same point in infancy infancy #y o#ser*ing dyads. n another early study' study' Waters' Waters' Wippman' Wippman' and 4roufe ,&5<5- found that attachment in infancy %as related to peer interaction at age three and a half. half. 4pecifically' 4pecifically' competence in the peer group %as predicted #y attachment status' suggesting that security of attachment to parents impacts impacts child)peer relationships. 2urthermore' 2urman' 4imon' 4imon' 4haffer' and Bouchey ,:"":found that' #ased on Bo%l#y+s pre*ious %ork' %orking models of friendships in late adolescence %ere related to those %ith %ith parents and romantic partners. Considering these findings and the de*elopmental trends that occur across the transition to adolescence' it seems as if' at this point' there %ould #e a relationship #et%een adolescent attachment to parents and to peers. nterestingly' nterestingly' 2urman' 4imon' 4haffer' and Bouchy also found significant differences in the attachment security status of adolescents to their parents and peers. 2or e$ample' some adolescents %ho %ere classified as dismissing dismissing %ith their parents %ere classified as secure %ith their peers. 0ne e$planation for this %as that at this point in de*elopment' some adolescents may not feel that their parents are responsi*e in times of need' and therefore seek this comfort from friends instead ,2urman' 4imon' 4haffer' 8 Bouchey' :"":-. While attachment theory says that a person has a style of interacting %ith others' it is important to note that mother)child and father)child relationships are discrete. Children e$perience each of o f their different relationships uni/uely and are influenced differently #y their mothers and fathers. As pre*iously e$isting e$isting relationships continue to #e important for de*elopment of ne% relationships in adolescence' it #ecomes apparent that adolescents start to differentiate among relationships %ith parents and %ith peers ,Collins 8 Repinsky' Repinsky' &55?-. Montemayor and Gregg ,&55?- speak of identity de*elopment in adolescence and an d its connection to interpersonal relationships= they note that as identity de*elops' adolescents sees that people they relate to see them in different %ays. Thus' in early adolescence one #egins to fully understand relationships relationships are distinct. As noted #efore' Williams Williams and 7elly ,:""!- also found differences differences in adolescent attachment to mother and to father. father. Therefore' it is important to e$amine attachment attachment relationships as different from one another. nterestingly' nterestingly' 2reeman and Bro%n ,:""&- conducted con ducted a study looking at the relationship #et%een attachment style and choice of attachment figure in adolescence. They found that' in general' g eneral' parents and peers %ere nominated /uantitati*ely e/ually' ho%e*er there %ere nomination nomination differences #ased on attachment status. Those
6
adolescents %ho %ere more secure nominated their mother more' and those %ho %ere more insecure %ere more likely to nominate their peers. n a study using an early self)report attachment measure' ickerson and agle ,:""!- found that attachment to parents and peers differed not only from one another #ut also across the adolescent transition. transition. t can #e strongly suggested suggested that attachment to parents changes in some manner across this transition #ecause communication and trust %ith parents decreased during these shifts ,from fourth to si$th to eighth grade- in adolescence ,ickerson 8 agle' :""!-. Additionally' Additionally' ickerson and agle found a change in the amount of reported pro$imity seeking and safe ha*en fulfillment %ith peers' such that' as mentioned se*eral times #efore' adolescents tend to seek out friends %hen needed instead of their parents. Although some /ualities /ualities of the parent)adolescent attachment relationship decrease %hile other /ualities of the adolescent)peer attachment relationship increase' parents still still are utili(ed as important attachment)figures. 0n the other hand' #oth ickerson and agle and 7erns' 7lepac' and Cole ,&55- found that use of parents to fulfill secure #ase needs did not change across this transition' %hich suggests that as adolescents e$plore ne% relationship realms' parents still remain an important #ase for security. ickerson and agle ,:""!- state that there are t%o *ie%s in e$plaining attachment to parents and peers. 0ne *ie% is that secure attachment in the parent)child relationship might allo% for for felt security in in other relationships. 0n the other hand' insecure attachment in the parent)child relationships might foster the desire to find security else%here. Currently' Currently' the ma1ority of research on peer attachment in adolescence e$amines the relationship relationship #et%een attachment and %ell)#eing. 2e% studies e$amine e$amine ho% adolescents are attached to their mother' father and peers and the relationship that e$ists #et%een these different attachments. Dimensions of Attachment The concept of a secure #ase seems to e$ist across all stages in de*elopment ,Armsden 8 Green#erg' &56<= Caffery 8 Erdman' Erdman' :"""-. Especially in adolescence' %hen e$ploration and autonomy from parents in #oth a physical and a psychological manner marks the adolescent transition' the presence and a*aila#ility of attachment figures is crucial ,e.g. ,e.g. Allen Allen 8 Land' &555-. Therefore' the #eha*iors of attachment attachment used to descri#e infant attachments should #e related to attachment in adolescence. 3a(an and 4ha*er ,&55?- state that in adult romantic relationships' the most important aspect of the relationship that relates to attachment is for each person to act as a relia#le ha*en of safety. safety. n adolescence' this idea translates translates to the support' %armth %armth and comfort that attachment figures can pro*ide. 4eparation protest is another #eha*ioral #eha*ioral aspect of *ery early attachments that translates to adolescence. 2or e$ample' #ecause of cogniti*e gains' a more comple$ understanding of separation %ill ha*e implications for attachment relationships %hen faced %ith more permanent separations' such as death. According to Armsden and Green#erg ,&56<- there are three underlying constructs of attachment that e$ist; communication' trust and alienation. The ma1ority of research pertaining to these three constructs focuses on communication and the relationship #et%een communication communication and attachment. Little research e$ists that e$amines the relationship #et%een #oth trust and attachment and alienation and communication. communication among parents and children has Communication. Bidirectional communication #een the focus of the ma1ority of research in this area. More specifically' specifically' 4egrin and
5
2lora ,:""!- argue that reciprocity' defined as mutual communication e$changes that are kno%ingly a*aila#le' and synchrony' defined as communication that occurs in a harmonious fashion' are aspects of communication that help create strong emotional #onds #et%een parents and children in infancy. Moreo*er' these strong parent)child parent)child relationship e$changes are important throughout life. n infancy' children seek pro$imity and comfort c omfort %hen they sense danger. Adolescents seek pro$imity and comfort in the form of o f ad*ice %hen they feel it is needed ,3a(an 8 4ha*er' &55?= 4chneider 8 Founger' Founger' &55-. Therefore' communication may #e e$tremely important in adolescence. >uring adolescence' the parent)child parent)child relationship depends on closeness' %hich is esta#lished and sustained from earlier stages' and conflict' %hich helps the adolescent distance' in a psychological sense' from the parents ,Laursen 8 Collins' :""?-. Additionally' Additionally' openness #et%een parents and adolescents is related to ha*ing a positi*e emotional climate climate ,Arnold' Pratt' 8 3icks' 3icks' :""?-. More open communication allo%s for understanding during a time of o f such important transition and changes' specifically in terms of acceptance of the ne% needs and desires adolescents face ,4illars' 7oerner' 8 2it(patrick' :""!-. The a#ility of parents and children to communicate %ith one another is related to security of attachment. 2or e$ample' attachment security is related to #etter connectedness of communication in the mother)child relationship ,2reitag' Belsky' Grossmann' Grossmann' 8 4cheuerer)English' 4cheuerer)English' &55-. 2reitag' Belsky' Belsky' Grossmann' Grossmann' Grossmann' and 4cheuerer)English ,&55- in*estigated the connection #et%een attachment and communication across across infancy and middle middle childhood. They conducted a cross cultural study that paralleled pre*ious ones conducted in the 9nited 4tates' finding an organi(ation in the parent)child relationship that is related to # oth attachment and communication. Additionally' Additionally' the authors refer to Bo%l#y+s Bo%l#y+s %ork on the partnership relationship that e$ists #et%een parent and child. The present study %ill e$amine the role that both the father)child and the mother)child relationships play in adolescent attachment #ecause these relationships co)e$ist in the family. family. The study conducted #y 2reitag and colleagues ,&55- looked at infancy through middle childhood' lea*ing adolescence une$plored' as %ith much of the research. Communication #et%een parents and children changes as the child passes through different de*elopmental stages ,Arnold' Pratt 8 3icks' :""?= Laursen 8 Collins' :""?-. Laursen and Collins ,:""?- speak of communication trends in families' mentioning that specifically in adolescence' the family unit functions #ased on prior interactions and communication patterns' #ut its mem#ers are a#le to re)e*aluate and adapt their communication to allo% for the adolescent+s changes. 4ince 2reitag and colleagues sho% the connection #et%een attachment security and communication across infancy and middle childhood' and it is kno%n that families that communicate #etter are #etter a#le to deal %ith de*elopmental transitions' it seems as if these communication patterns should persist across the transition transition to adolescence as %ell. This suggests that the #etter the communication #et%een parents and children in adolescence' the more a sense of felt security should e$ist across de*elopmental transitions. Beginning in infancy' infancy' parent)child communication creates a foundation for communication %ith others across the lifespan ,Bo%l#y' &566= 4egrin 8 2lora' :""!-. 7reppner and 9llrich ,&556- studied communication across the transition from childhood to adolescence. They found that after grouping grouping adolescents into different different
&"
attachment style categories ,secure' ha#itual' and am#i*alent-' differences in communication e$isted %ithin the family family unit. 4pecifically' 4pecifically' secure adolescents sho%ed less silence in their communication %ith parents than the other groups' and the secure group also sho%ed higher le*els of e$pressed closeness %ith mothers. 7reppner and 9llrich also found differences in adolescent communication %ith mother and father. nterestingly' those adolescents in the secure group e$hi#ited more distant #eha*ior from fathers' and higher le*els of e$pressed closeness %ith fathers %as seen %ith the ha#itual and am#i*alent groups. The secure group' ho%e*er' displayed #oth *ery high and *ery lo% e$pression of closeness %ith their fathers. Therefore' more *ariation *ariation in communication may lead an adolescent to #e a#le to ad1ust le*els of communication as needed= specifically' since one of the ma1or de*elopmental shifts in adolescence is that of autonomy and indi*iduation' #eing a#le to shift the amount of closeness and distance is optimal. feelings and Trust. The second dimension of trust can #e defined as the secure feelings #eliefs that another person %ill fulfill fulfill certain needs ,Armsden 8 Green#erg' &56<-. Trust Trust is a product of strong relationships' specifically those in %hich relationship partners feel that they can depend upon one another ,Collins 8 Repinsky' &55?-. 2urthermore' trust is one component of a strong relationship #et%een children and their attachment figures' as it is seen that children #uild trust in relationships #y learning that others are consistently there for them. The secure #ase phenomenon emphasi(es the kno%ledge of a*aila#ility of the attachment figure in times of need. n other %ords' a representation representation of the a#ility to trust the attachment figure e$ists #ecause of positi*e past situations related to trust. oller ,&55?- also states that trust%orthiness' trust%orthiness' as %ell as closeness' is an e$tremely important /uality of relationships. According to 4a*in)Williams and Berndt ,&55"- one of the ma1or /ualities of a strong peer relationship is trust. Additionally' Additionally' 3a(an and 4ha*er ,&55?- argue that' during adolescence' peer attachments #egin %ith such such pro$imity seeking #eha*iors. o matter %hom the attachment figure is' adolescents %ant to feel that they are close to' and can trust' those %ith %hom they ha*e relationships ,oller' &55?-. Alienation. The third dimension of attachment' alienation' is closely related to a*oidance and re1ection' t%o constructs that are *ery important to security of attachment. When one senses that the attachment figure is not a*aila#le' attachment #ecomes less secure' possi#ly #ased on feelings of alienation. 4urprisingly' 4urprisingly' gi*en the importance of the alienation dimension' no research on the relationship #et%een alienation and attachment e$ists. t is therefore therefore important to e$plore the relationship that feelings of alienation ha*e to attachment. Summary Attachment theory emphasi(es the importance of the emotional #onds in relationships. t has #een sho%n that parents parents play an important role as attachment figures for their children. Taking Taking into account the de*elopmental d e*elopmental changes that occur in adolescence' it also has #een found that parent and adolescent roles change as adolescents stri*e for more autonomy autonomy and indi*iduation ,4egrin 8 2lora' :""!-. Most importantly' importantly' secure adolescents ha*e #een found to #e more a#le to e$plore their en*ironment and achie*e more independence ,Lauren 8 Collins' :""?-. t also has #een sho%n sho%n that as children transition through adolescence' peers #ecome important figures as %ell as
&&
parents. 4ome of the research re*ie%ed ,e.g. ickerson 8 agle' :""!- suggests a smooth transition into the larger social %orld of adolescence %hen there is security %ithin the parent)child relationship' %hich then is related to more security in relationships that follo% the adolescent outside of the home. There are' ho%e*er' inconsistencies in the literature as to the patterns in attachment across different attachment figures ,e.g. Le%is' 2erin 8 Rosenthal' :"""= Main 8 Weston' &56&-' gi*ing merit to the present study. Attachment theorists e$amine attachment #eha*iors and representations across the lifespan. Most of the attachment attachment literature' ho%e*er' focuses on infancy ,e.g. 2o$' 7immerly' 7immerly' 8 4chafer' &55&= Easter#rooks' &565-' childhood ,e.g. 7erns' 7lepac' 8 Cole' &55- and adulthood ,e.g. 3a(an 8 Heifman' &555-' %ith much less research e$amining attachment in adolescence. onetheless' adolescence marks a significant period period of transition. Therefore it is important to e$amine attachment during such an important period as %ell. Research that e$amines adolescence typically concerns continuity of attachment from infancy ,e.g. 4froufe 8 Waters' Waters' &5<<- or attachment to the mother only ,e.g. Allen Allen et al.' :""@-. There is e*en less research on adolescent attachment to fathers and peers. The first o#1ecti*e of the present study %ill e$plore the p atterns of adolescent attachment to their mothers' mothers' fathers and peers. 4ince almost no studies of attachment in adolescence look at ho% relationships %ith all three figures relate to one another' this study is among the first first to do so. Considering the lack of research research and understanding of this interaction' specific specific hypotheses are not pro1ected. The second o#1ecti*e is to e$amine similarities %ithin mother and adolescent ratings of one another and father and adolescent ratings of one another. another. Again' #ecause this is is a ne% topic among attachment researchers' this study %ill e$plore relations %ithout pro1ecting specific hypotheses.
Method This study is part of a larger pro1ect' %hich e$amines family narrati*es and adolescent identity. identity. 0nly those methods related to the present pro1ect %ill #e discussed. Sample T%enty)four T%enty)four families %ith either an 6th grade or a &"th grade adolescent participated in the study. 2amilies %ere recruited through *arious sources such as schools and religious groups ,e.g. church-' as %ell as %ith fliers that %ere dispersed to #oth participating families and around a uni*ersity. uni*ersity. To #e included in the study' families families had to ha*e t%o parents li*ing in the home' either #iological parents' step parents' or adopti*e ad opti*e parents. Most families had other children li*ing in the household as %ell. 2amilies typically had #et%een one and four total children li*ing in the household. Thirteen families identified themsel*es as Caucasian' &" as African American' American' and one as 3ispanic. ineteen families %ere descri#ed descri#ed as traditional ,#oth ,#oth #iological parents li*ing in the home- and three as #lended ,one #iological parent and one non)#iological parent li*ing in the home-. 0ut of the :? families' families' t%o adolescents %ere adopted ,#oth ,#oth in infancy-. Ten adolescents %ere female ,mean age &?.- and &? %ere male ,mean age &?.!-. At the time of the study' study' ten adolescents %ere in 6th grade ,mean age &@.?- and &? %ere in &"th grade ,mean age &!.?-. All parents ga*e signed consent and all adolescents ga*e signed assent as appro*ed #y the Emory Emory 9ni*ersity nstitutional nstitutional Re*ie% Board. 2or
&:
participating in the study' families %ere compensated I!"' and adolescents %ere gi*en t%o mo*ie tickets and a I:! gift certificate. Procedure A female female researcher *isited the family+s family+s home on t%o separate occasions. >uring the *isit' se*eral narrati*es %ere collected from the mother and adolescent as part of the larger study. study. n addition' /uestionnaire packets %ere left left for #oth the mother mother and adolescent to complete separately. Another /uestionnaire packet %as left for the father father to complete as %ell. The researcher #riefly e$plained the /uestionnaires. Participants %ere asked to complete all /uestionnaires' although they the y %ere told that /uestions may #e skipped if there is any reason to do so' although this %as not encouraged. Ty Typically pically %ithin t%o %eeks' the same researcher returned to the home and collected the /uestionnaires. "easure Inventory of Parent and Peer Attachment-evised Attachment-evised &IPPA &IPPA' Armsden ( )reenberg* )reenberg* PPA' a self)report /uestionnaire that +,/. Each adolescent %as asked to complete the PPA' includes :! items that %ere designed to measure adolescent attachment to parent ,mother-' adolescent attachment to parent ,father- and adolescent ado lescent attachment to friends ,peers-. The measure assesses #oth #oth positi*e and negati*e affecti*e affecti*e and cogniti*e dimensions related to attachment. attachment. As discussed discussed in the introduction' the three three dimensions used to measure attachment are communication' communication' trust and alienation. The communication dimension is measured %ith %ith &" items. 2or e$ample; f my motherJfatherJfriends motherJfatherJfriends kno%s something is #othering me' heJshe asks me. The trust dimension dimension is measured %ith nine items. 2or e$ample; My motherJfatherJfriends motherJfatherJfriends respects respects my feelings. The alienation alienation dimension is measured %ith %ith si$ items. 2or e$ample; don+t get much attention from my motherJfatherJfriends. The adolescent is asked to complete each set of /uestions in relation to their mother' mother' their father father and their peers. Kuestions are ans%ered on a !)point !)point Likert scale' ranging from Almost e*er or e*er e* er True to Almost Al%ays Al%ays or Al%ays True. 9sing t%o samples of adolescents' ages ranging from & to :"' Armsden and Green#erg ,&56<- found good internal relia#ility for the PPA' PPA' %ith Chron#ach+s alphas . 6< for mother attachment' .65 for father attachment and .5: for peer attachment. A three) %eek test)retest relia#ility of .5@ for parent attachment and .6 for peer attachment %as also found ,Armsden 8 Green#erg' &56<-. The PPA also has good construct *alidity. *alidity. t has #een found to #e related to other measures' such as the 2amily 4elf)Concept ,r ".<6 %ith parent attachment= r ".:6 %ith peer attachment- and 4ocial 4elf)Concept ,r ".? %ith parent attachment= r ".!< %ith peer attachment- su#scales of the Tennessee 4elf)Concept 4cale and to se*eral su#scales of the 2amily En*ironmental 4cale ,2E4- ,Armsden and Green#erg' &56<-. 4pecifically' parent attachment %as positi*ely related to Cohesion ,r ".!-' E$pressi*eness ,r ".!:- and 0rgani(ation ,r ".@6-' and %as negati*ely related to Conflict ,r )".@- and Control ,r )".:"- ,Armsden 8 Green#erg' &56<-. n addition to asking adolescents to complete co mplete the PPA for their relationships %ith their mother' father and peers' each parent %as asked to complete the related *ersion of the same /uestionnaire to allo% for e$ploration of parent+s attachment attachment to their adolescents. ohnson' 7etring and A#shire A#shire ,:""@- re*ised the original PPA PPA to #e used for this purpose. This /uestionnaire consists consists of &" /uestions for for the communication
&@
dimension' se*en for the trust trust dimension and ! for the alienation dimension. 4e*eral items %ere re)%ritten in order to allo% parents to ans%er the / uestion #ased on their relationship %ith their children. children. 2or e$ample' the original /uestion' tell my motherJfatherJfriends a#out my pro#lems and trou#les' %as changed to' talk to my child a#out my difficulties. Additionally' Additionally' se*eral items %ere deleted from the original *ersion' such as' My motherJfatherJfriends doesn+t understand %hat +m going through theses days' and se*eral additions %ere made' such as' am constantly yelling and fighting %ith my child. Relia#ility for the different dimensions %ere #et%een alpha . <: and alpha .5! ,ohnson' 7etring 8 A#shire' :""@-. The re*ised *ersion of the PPA PPA %as also correlated %ith other measures looking at *aria#les related to attachment ,ohnson' 7etring 8 A#shire' :""@-. Scoring Each dimension of the PPA ,trust' communication and alienation- yields a separate score' and is scored separately. separately. 4e*eral /uestions on the trust and communication dimensions and the %hole alienation dimension are re*erse scored. Ans%ers to /uestions for for each dimension are then added together. together. f a /uestion is skipped' the mean of the ans%ered /uestions is then added as the score for the skipped /uestion. All three dimensions are also totaled totaled to create a total attachment attachment score dimension. The mean for each dimension is computed and used for analyses. Because se*eral fathers %ere not a*aila#le to complete co mplete the PPA' PPA' any analyses that include father ratings of attachment to adolescent %ere only conducted %ith scores from &5 families. iii
Results Results are discussed in t%o sections. The first section section e$amines ho% male and female adolescents rate their attachment to parents and peers as %ell as ho% mothers and fathers rate their attachment attachment to their adolescents. The second part e$amines the relations relations among attachment ratings that address the specific o#1ecti*es of the study. Differences among attachment ratings Ta#le & displays the mean ratings ,and standard de*iations- adolescents ga*e on each dimension to mother' father and peers. To look at the differences #et%een adolescent ratings of mother' father and peers on all dimensions' a : ,grade- N : ,genderN @ ,person- N ? ,dimension- Mi$ed)Model Analysis of Oariance ,A0OA- %as conducted %ith grade and gender as #et%een su#1ect factors and person and dimension as %ithin su#1ect factors. There %as a main effect effect of person ,2,:'?",2,:'?"- <.!' p ."&- and a main effect of dimension ,2,@'"- 5.&@' p .""&-' as %ell as a person #y dimension interaction ,2,'&:"- 6.:6:' p .""&-. o other main effects effects or interactions reached significance. Therefore' : ,grade- N : ,gender- N @ ,person- A0O A0OAs As %ere conducted for each dimension separately. separately. There %ere no main effects or interactions interactions on the trust dimension. There %as' ho%e*er' a main effect of person on the communication dimension ,2,:' ?"- &<.&"' p .""&-' the alienation dimension ,2,:' ?"- &6.!5' p . ""&-' and the total attachment score ,2,:' ?"- ?
&?
dimension' there %as a significant difference #et%een adolescent ratings of fathers and peers ,t ,:@,:@- ).?"' p .""&-' mothers and peers ,t ,:@,:@- )@.
&!
significance #et%een adolescent and father ratings of one another for the communication dimension. All other dimensions for #oth pairs pairs %ere not significantly significantly related. As an e$ploratory analysis' the concordance #et%een mother and father ratings of adolescents %as e$amined and is displayed in Ta#le Ta#le !. There %as a significant positi*e correlation #et%een mother and father ratings of adolescents ado lescents for the trust and alienation dimensions. The communication dimension dimension and the total attachment score %ere %ere not significantly related.
Discussion The present study e$amined patterns of attachment during adolescence. Because adolescence is such a crucial point in de*elopment' especially as it is related to social transitions' it is imperati*e that %e understand ho% attachment looks in adolescence. 4pecifically' patterns patterns in adolescent attachment to mothers' fathers and peers %ere e$plored' and the similarity #et%een parent and adolescent ratings of attachment to one another %as e$amined. 2indings and interpretations for each of these /uestions %ill %ill #e discussed' as %ell as limitations of the study and implications for future research Adolescents sho%ed different le*els of attachment to different figures. 4pecifically' peers %ere rated higher than mothers' and #oth peers and mothers %ere rated higher than fathers on #oth the communication dimension and the total attachment score. Mother and peers' rated e/ually on the alienation dimension' %ere #oth rated higher than fathers. 4chneider and Founger Founger ,&55,&55- and 3a(an and 4ha*er ,&55?- find that adolescents seek comfort from people %ho are most accessi#le. 4upporting the argument #y Collins and Repinsky ,:""?-' at this point in de*elopment' mothers may #e those in the family %ho are most easily accessi#le to satisfy these these needs. As pre*iously found #y Paterson' 2ield' and Pryor ,:""?-' adolescents ado lescents tend to rely on their mothers more than their fathers %hen in need of comfort and support. Therefore' present findings suggest that mothers may #e more accessi#le to adolescents than fathers inside the home' an e$planation for fathers #eing rated rated lo%est. t may #e that #ecause adolescents feel less alienated from their mothers than their fathers' they rate their communication higher and feel more secure attachment to their mothers as %ell. Additionally' Additionally' in general mothers and adolescents rated one another similarly' %hereas fathers and adolescents did not. n other %ords' mothers mothers and adolescents #oth reported similar le*els of ratings of attachment to one ano ther' %hereas fathers reported higher le*els of ratings of attachment to their adolescents ad olescents than their adolescents did to them. This suggests that that fathers percei*e their relationships relationships as stronger than ho% adolescents percei*e the relationship' #ut that adolescents ad olescents and mothers percei*e the strength of their relationships similarly. similarly. nterestingly' nterestingly' adolescents are more open %ith their mothers than their fathers on an emotional le*el ,Larusen 8 Collins' :""?-. t may #e that' #ecause of the le*el of a*aila#ility in the mother)adolescent relationship' relationship' adolescents feel more comfort and support from their mother. mother. 2athers' on the other hand' may #e less a*aila#le to satisfy their adolescents+ needs' yet recei*e a certain comfort and support from their adolescents. mportantly' mportantly' peers %ere rated highest for se*eral dimensions' supporting pre*ious research on the topic ,2urman 8 Buhrmester' &55:= ickerson 8 agle' :""!-. Amount of time spent %ith peers increases across adolescence ,Collins ,Collins 8 Repinsky' &55?-. As
&
%ith mothers *erses fathers' peers may #e more readily a*aila#le for satisfying needs of comfort and support. The present findings' therefore' therefore' support 3a(an and 4ha*er+s ,&55?argument that children create stronger #onds %ith those %ho are a*aila#le attachment figures' %hich also may e$plain %hy %hy peers tended to #e rated highest. highest. These findings fit %ith e$isting kno%ledge of adolescent de*elopment. As adolescents seek seek autonomy from the family their peer relationships relationships gain importance. 0n the other hand' adolescents continue to desire support from from their parents ,2reeman ,2reeman 8 Bro%n' :""&-. Therefore' an e$planation of the present findings' integrating that adolescents rate #oth peers and mothers high' is that adolescents de*elop more social needs' %hich lead them to seek out peers for comfort' #ut still need support from a strong strong home #ase. Regarding adolescent ratings of attachment #et%een mothers and fathers' findings suggest that there is concordance on ho% adolescents feel a#out their relationship %ith their mother and their father. father. 0*erall' adolescent ratings of mothers and fathers %ere %ere related on all dimensions' such that those adolescents ad olescents %ho are rating attachment to their mothers high are also rating attachment to their fathers high on the trust' communication and alienation dimensions as %ell as on the total attachment attachment score. The strong relation #et%een adolescent ratings of mothers and fathers suggests that adolescents generali(e across parental relationships and may *ie% parents as a unit. This idea supports the #asic tenet of attachment theory. The internal %orking model of attachment originally presented #y Bo%l#y ,&55- suggests that relationships %ith attachment figures figures are represented mentally' #ased on past e$periences' and that these ideas and e$pectations may then mold into a general style of attachment. Additionally' Additionally' se*eral researchers ha*e found high concordance rates #et%een infant attachment to their mother and father ,Easter#rooks' &565= 2o$' 7immerly' 7immerly' 8 4chafer' &55&-. Although this study is not longitudinal' %hen taking into account the research on concordance in infancy' findings of concordance suggest that children mentally represent relationships %ith their parents #ased on earlier points in the relationships. n other %ords' this study suggests that Bo%l#y %as correct= that there is' in fact' an internal %orking model of attachment. At a minimum' ho%e*er' this study supports the notion that' at this point in de*elopment' adolescents are a#le to generali(e attachment across their relationships %ith #oth parents. 4urprisingly' 4urprisingly' ho%e*er' this study did not no t find relations among adolescents attachment to parents and peers. As re*ie%ed' pre*ious pre*ious studies ,e.g. Easter#rooks Easter#rooks 8 Lam#' &5<5= Waters' Waters' Wippman' 8 4roufe' &5<5- ha*e found relationships #et%een #e t%een attachment to mother and peers in infancy and childhood. Additionally' Additionally' Lai#le' Carlo' and Roesch ,:""?- found a relationship #et%een attachment to parents and peers using the PPA. PPA. The study #y Lai#le and colleagues suggests that adolescent+s representations of their relationships are generali(ed across those %ith parents and peers' an important de*elopmental milestone socially. socially. The study #y Lai#le and colleagues' ho%e*er' ho%e*er' used an older sample ,mean age &6. years-' %hich might suggest that the younger sample in the present study has already generali(ed across parental relationships #ut not yet across all relationships' such as those %ith peers. nterestingly' nterestingly' another study #y Lai#le and colleagues ,:"""-' conducted %ith a sample of & year olds' sho%ed similar findings to the present study in terms of correlations among attachment to parents and peers. Again' it may may #e that at this point in de*elopment' adolescents are 1ust #eginning to generali(e across attachment figures. Research on friendships sho%s that across adolescence' peers are sought out in times of
&<
need more than #efore. As found in this study' study' peers do gain importance in adolescence' sho%n #y findings that' in general' adolescents are rating attachment to their peers highest. t may #e that early adolescents are using their peers to satisfy certain needs that are different than those that they seek out their parents for. for. 4ince adolescence is a time for identity de*elopment ,Erikson' &56-' adolescents typically stri*e for more autonomy from their parents and more more inclusion and acceptance into social realms. realms. Therefore' during this transition' adolescents may not yet #e a#le to generali(e across all relationships' #ut do rely on se*eral different relationships for comfort and support. 2uture research should attempt to find %hat *aria#les *a ria#les allo% for the generali(ing of attachment across figures and %hen' e$actly' this occurs. An additional /uestion addressed in the study %as the similarity #et%een adolescent and parental ratings of attachment attachment of one another. When comparing adolescents+ ratings ratings of their mothers %ith mothers+ ratings of their adolescents' mothers and adolescents rated one another similarly on the trust dimension as %ell as the total attachment score. n comparing similarities similarities in ho% adolescents rated rated their fathers and ho% fathers rated their adolescents' fathers and adolescents rated one another similarly on the communication dimension. Laursen and Collins ,:""?- argue argue that more time time is spent %ith mothers than %ith fathers and' more importantly' that there is more sharing of emotions %ith mothers as %ell. 0ther researchers ha*e also found relations #et%een /uality of the mother)adolescent relationship and attachment security ,Allen et al.' : ""@-' and percei*ed maternal a*aila#ility a*aila#ility ,Lei#erman' >oyle' >oyle' 8 Markie%ic(' &555-. n these stronger' higher)/uality higher)/uality relationships' there may also #e higher le*els of trust #et%een mothers and their adolescents. Additionally' Additionally' Benoit and Parker ,&55?- ha*e found found that secure mothers tend to ha*e secure children. Therefore' it may #e that mothers %ho are a*aila#le' specifically emotionally' are those %ho ha*e children %ho percei*e them to #e this %ay' as %ell as feel as if their children are a#le to satisfy some of their emotional needs. 2athers and adolescents percei*ed one another similarly on the communication dimension. Possi#le reasons for this finding are not immediately immediately apparent' particularly gi*en the findings that fathers and adolescents' generally' are not rating their relationships the same. >ra%ing upon findings #y 7reppner and 9llrich 9llrich ,&556- that *aria#ility *aria#ility in communication among fathers and adolescents is related to security of attachment' it may #e that communication %ithin the father)adolescent pair helps foster foster a #etter relationship. Williams Williams and 7elly ,:""!- also suggest that paternal p aternal in*ol*ement is related to adolescent attachment. Additionally' Additionally' mothers and fathers play %ith and influence their children differently ,Leaper' ,Leaper' :"""-. Therefore' relationships %ith %ith #oth mothers and fathers fathers are important during the adolescent transition= ho%e*er' as the pre*ious research and the present findings suggest' each relationship has its uni/ue /ualities and different different impact on adolescent de*elopment. Whereas' findings indicate that relations %ithin the mother)adolescent and father) adolescent relationships e$ists on different dimensions' dimensions' these dimensions are all related to one another. Many researchers find that communication in the parent)infant and parent) parent) child relationship is related to infant and child attachment security ,e.g. 2reitag' Belsky' Grossmann' Grossmann' 8 4cheuerer)English' &55-' such that secure relationships tend to #e more communicati*e' especially in open and emotional %ays. Because of the dyadic nature of communication and trust' and the findings that parental in*ol*ement is related
&6
to adolescent attachment ,Williams ,Williams 8 7elly' :""!-' it is the communication and trust in the parent)adolescent relationship that allo%s for adolescents to feel close to their parents and' 1ust as important' parents parents to feel close to their adolescents. 2inding some similarity similarity in ratings of attachment to one another in both the mother)adolescent and father) adolescent relationships suggests these relationships may #egin to #ecoming more reciprocal' as suggested #y other researchers ,Laursen 8 Collins' :""?-. Lastly' there %as concordance #et%een mother and father ratings of attachment to their adolescent. This finding makes sense in terms of 2amily 4ystems 4ystems Theory' Theory' %hich argues that the family functions as a %hole unit as %ell as separate dyadic units' %hich all interact ,7reppner' :"":-. The concordance #et%een all three dyadic relationships relationships ,mother)adolescent' father)adolescent and mother)father- indicates that these attachment relationships are not independent of one another. another. t may #e that through these these interactions' as a %hole family as %ell as separate dyads' d yads' comes the creation of a household climate. 0ther researchers ha*e also argued argued that families interact interact as a %hole' creating a family style of interacting ,Arnold' Pratt' 8 3icks' :""?= 7reppner' :"":-. These findings contri#ute to the importance of furthering research on the %hole family in addition to its separate mem#ers and dyads %hen trying to e$plain ho% families interact and impact one another. 0imitations There are se*eral limitations limitations to the present study study that %arrant attention. attention. 2irst' although there is a good amount of *aria#ility in the demographics of the sample' sample si(e is fairly small. small. Any general conclusions dra%n a#out attachment in adolescence must #e taken %ith the understanding u nderstanding that the sample si(e limits the po%er of findings. 2amilies also had some kno%ledge that the o*erall study %as a#out family relationships= therefore there may #e specific /ualities possessed #y the families that %ould lead them to participate in such a study. Additionally' Additionally' the PPA is a self)report /uestionnaire and may yield ans%ers that are #iased. There is the possi#ility possi#ility that reported ans%ers ans%ers may reflect a certain certain le*el of social' or e*en family' family' desira#ility. desira#ility. There is some contro*ersy in the literature a#out using self)report measures measures for attachment. 4ince attachment is thought to #e a mental representation of one+s emotional #onds and past e$periences in relationships' it is thought that the #est %ay to measure attachment is through narrati*es that tap into the implicit representations in the mind ,for a re*ie% different measures of attachment' see Cro%ell' 2raley' 8 4ha*er' &555= Lyddon' Lyddon' Bradford' 8 elson' &55@= 4ha*er 4 ha*er 8 Mikulincer' :""?-. 9n/uestiona#ly' though' the PPA PPA does measure *aria#les that are related to important aspects of the /uality of relationships and' importantly' importantly' that are related to attachment. Lastly' Lastly' this study does not yield data that speaks to%ard causality= therefore only relational conclusions can #e dra%n. onetheless' this study has sho%n that there are specific specific patterns of attachment in adolescence' ho%e*er more research is needed to further understand these these patterns. By adolescence' there is the capa#ility to generali(e across relationships' ho%e*er it is still unclear to %hat e$tent this occurs during early adolescence. Adolescents also ha*e different relationships %ith each parent' yet these relationships seem to #e highly related to one another in regard to attachment attachment dimensions. t is e$tremely important important to e$plore ho% these relationships %ork together in a family unit and' 1ust as important' to e$plore ho% relations inside the family impact the adolescent+s adolescent+s e$tra)familial relationships. relationships. The
&5
present study demonstrates the e$isting comple$ities of the adolescent period and the transitions that occur at this time in de*elopment' %ith hop es to spark more curiosity and e$ploration a#out relationships %ithin and outside of the family during adolescence. Conclusi*ely' parents are o#*iously still important figures in their adolescents+ li*es' %hile at the same time time peers are important as %ell. %ell. t seems as if in early adolescence' there is still a strong need for parental support' especially in the form of a good relationship' %hich may help adolescents form close #onds %ith friends as they enter into a larger social %orld.
References Allen' . P.' P.' 8 Land' >. ,&555-. Attachment in adolescence. n . Cassidy 8 P. 4ha*er ,Eds.-' 1andboo2 of attachment' Theory* research research and clinical clinical implications ,pp. @&5)@@!-. e% Fork; Guilford Press. Allen' . P.' P.' McElhaney' 7. B.' Land' >. .' 7uperminc' G. P.' P.' Moore' C. W.' W.' 0+Beirne) 7elly' 7elly' 3.' et al. ,:""@-. A secure #ase in adolescence; Markers of attachment security in the mother)adolescent relationship. Child Development* 3' :5:)@"<. Armsden' G. C.' 8 Green#erg' M. T. ,&56<-. The in*entory of parent and peer attachment; ndi*idual differences and their relationships to psychological %ell) #eing in adolescence. 4ournal of 5outh 5outh and Adolescence* +6 ' ?:<)?!?. Arnold' M. L.' Pratt' M. W.' 8 3icks' C. ,:""?-. Adolescents+ representations of parents+ *oices in family stories; Oalue Oalue lessons' personal p ersonal ad1ustment' and identity de*elopment. n M. Pratt 8 B. 2iese ,Eds. /* /* $amily stories and the life course' Across time time and generations ,pp. &@)&6-. Mah%ah' ; La%rence Erl#aum Associates' Pu#lishers. Benoit' >.' 8 Parker' 7. C. ,&55?-. 4ta#ility and transmission of attachment across three generations. Child Development* 67' &???)&?!. Bo%l#y' . ,&55-. Attachment and loss' #ol. #ol. +. Attachment. e% Fork; Basic Books.
:"
Bo%l#y' . ,&566-. A secure base' Parent-child Parent-child attachment and healthy human development . e% Fork; Basic Books. Ca#rrera' . .' Tamis)LeMonda' C. 4.' Bradley' R. 3.' 3offerth' 4.' 8 Lam#' M. E. ,:"""-. 2atherhood in the t%enty)first century. Child Development* +' &:<)&@. Caffery' Caffery' T.' 8 Erdman' P. ,:"""-. Conceptuali(ing parent)adolescent conflict; Applications from systems and attachment theories. The $amily 4ournal' Counseling and Therapy for Couples and an d $amilies* ' &?):&. Collins' W. A.' 8 Repinski' >. . ,&55?-. Relationships during adolescence; Continuity and change in interpersonal prespecti*e. n R. Montemayor' G. Adams' 8 T. Gullotta ,Eds.-' Personal relationships relationships during adolescence ,pp. &)@-. Thousand 0aks' CA; 4age. Cro%ley' . A.' 2raley' 2raley' R. C.' 8 4ha*er' P. R. ,&555-. Measurement of indi*idual differences in adolescent and adult attachment. n . Cassidy 8 P. 4ha*er ,Eds.-' 1andboo2 of attachment' Theory* research research and clinical clinical implications ,pp. @@) @!?-. e% Fork; Guilford Press. >eko*ic' M.' 8 Buist' 7. L. ,:""!-. Multiple perspecti*es %ithin the family; 2amily relationship patterns. 4ournal of $amily Issues* 86 ' ?<)?5". >oherty' W. W. .' 8 Beaton' . M. ,:""?-. Mothers and fathers parenting together. n A. L. Oangelistsi ,Ed.-' 1andboo2 of family communication ,pp. :5):6-. Mah%ah' ; La%rence Erl#aum Associates. Associates. Easter#rooks' M. A. ,&565-. Kuality of attachment to mother and to father; Effects of perinatal risk status. Child Development* 69* 6:!)6@". Easter#rooks' M. A.' 8 Lam#' M. E. ,&5<5-. The relationship #et%een /uality of infant) mother attachment and infant competence in initial encounters %ith peers. Child Development* 79' @6")@6<. Erikson' E. ,&56-. Identity' 5outh e % Fork; Fork; W. W. W. orton 8 Company. Comp any. 5outh and crisis. e% 2o$' . A.' 7immerly' 7immerly' . L' 8 4chafer' W. W. >. ,&55&-. Attachment to motherJattachment to father; A meta)analysis. meta)analysis. Child Development* 68' :&")::!. 2reeman' 3.' 8 Bro%n' B. B. ,:""&-. Primary attachment to parents and peers during adolescence; >ifferences #y attachment style. 4ournal of 5outh 5outh and Adolescence* :9' !@). ,&55:-. Age and se$ differences in perceptions of net%orks and personal relationships. Child Development* 6:' &"@)&&!. 2urman' W.' 4imon' O. A.' 4haffer' L.' 8 Bouchey' 3. A. ,:"":-. Adolescents+ %orking models and styles for relationships %ith parents' friends' and romantic partners. Child Development* :' :?&):!!. Green#erg' M. T.' T.' 4iegel' . M.' 8 Leitch C. . ,&56:-. The nature and importance of attachment relationships to parents and peers during adolescence. 4ournal of 5outh 5outh and Adolescence* +8' @<@)@6. 3a#ermas' T.' T.' 8 Bluck' 4. ,:"""-. Getting a life; The emergence of the life story in adolescence. Psychological %ulletin* +86 '
:&
3a(an' C.' 8 4ha*er' P. R. ,&55?-. Attachment as an organi(ational frame%ork for research on close relationships. Psychological In;uiry* 7' &)::. 3a(an' C.' 8 4ha*er' 4ha*e r' P. P. ,&56<-. Romantic lo*e conceptuali(ed as an attachment process. 4ournal of Personality and Social Psychology* 78' !&&)!:?. 3a(an' C.' 8 Heifman' >. ,&555-. Pair #onds as attachments; E*aluating the e*idence. n . Cassidy 8 P. 4ha*er ,Eds.-' 1andboo2 of attachment' Theory* researc research h and Fork; Guilford Press. clinical implications ,pp. @@)@!?-. e% Fork; 3ill' .' 2onagy' P.' P.' 4afier' E.' 8 4argent' . ,:""@-. The ecology of attachment in the family. $amily Process* 38' :"!)::&. 3unter' 2. T.' T.' 8 Founiss Founiss . ,&56:-. Changes in functions of three relations during adolescence. Developmental Psychology* +' 6")6&&. ohnson' L. .' 7etring' 4. A.' 8 A#shire' C. ,:""@-. The re*ised in*entory of parent and peer attachment; Measuring attachment in families. Contemporary $amily Therapy' An International 4ournal* 87' @@@)@?5. 7aren' R. ,&55?-. %ecoming attached' $irst relationships relationships and ho! they shape our capacity to love. e% Fork; 0$ford 9ni*ersity Press. 7erns' 7. A. ,&55?-. A de*elopmental model of the relations #et%een mother)child attachment and friendship. n R. Er#er 8 R. Gilmour ,Eds.-' Theoretical frame!or2s for personal relationships relationships ,pp. &:5)&!-. 3illsdale' England; La%rence Erl#aum Associates. 7erns' 7. A.' 7lepac' L.' 8 Cole' A. 7. ,&55-. Peer relationships and preadolescents+ perceptions of security in the child)mother relationship. relationship. Developmental Psychology* :8' ?!<)?. 7reppner' 7. ,:"":-. Retrospect and prospect in the psychological study of families as systems. n . Mc3ale 8 W. Grolnick ,Eds.-' etrospect and prospect prospect in the psychological study of families ,pp. ::!):!<-. Mah%ah' ; La%rence Erl#aum Associates. 7reppner' 7.' 8 9llrich' M. ,&556-. Talk Talk to mom and dad and listen to %hat is in #et%een; A differential approach to family communication communication and its impact on adolescent de*elopment. n M. 3offer' . Founiss Founiss'' 8 P. oack ,Eds.-' #erbal Westport' interaction and development in families !ith adolescents ,pp. 6@)&"6-. Westport' CT; A#le$ Pu#lishing. Lai#le' >. .' Carlo' G.' 8 Rafaelli' M. ,:"""-. The differential relations of parent and peer attachment to adolescent ad1ustment. 4ournal of 5outh 5outh and Adolescence* 8,* ?!)!5. Lai#le' >. .' Carlo' G.' 8 Roesch' 4. C. ,:""?-. Path%ays to self esteem in late adolescence; The role of parent and peer attachment' empathy' and social #eha*iours. 4ournal of Adolescence* 8* <"@)<&. Laursen' B.' 8 Collins' W. W. A. A. ,:""?-. Parent)child communication during adolescence. n A. L. Oangelistsi ,Ed.-' 1andboo2 of family communication ,pp. @@@)@?6-. Mah%ah' ; La%rence Erl#aum Associates. Associates. Leaper' C. ,:"""-. Gender' affiliation' assertion' and the interacti*e conte$t of parent) child play. Developmental Psychology* :6 ' @6&)@5@. Lie#erman' M.' >oyle' A.' 8 Markie%ic(' >. ,&555-. >e*elopmental patterns in security of attachment to mother and father in late childhood and early adolescence; Associations %ith peer relations. Child Development* 9' :":):&@.
::
Le%is' M.' 2eiring' C.' 8 Rosenthal' 4. ,:"""-. Attachment o*er time. Child Development* +' <"<)<:". Lyddon' Ly ddon' W. .' Bradford' E.' 8 elson' . P. ,&55@-. Assessing Assessing adolescent and adult attachment; A re*ie% re*ie% of current self)report measures. 4ournal of Counseling and Development* +' @5")@5!. Main' M.' 8 Weston' Weston' >. R. ,&56&-. The /uality of the toddler+s relationship to mother and father; Related to conflict #eha*ior and the readiness to esta#lish ne% relationships. Child Development* 78' 5@:)5?". Montemayor' R.' 8 Gregg' O. R. ,&55?-. Current theory and research on personal relationships during adolescence. n R. Montemayor' G. Adams' 8 T. Gullotta Gullotta ,Eds.-' Personal relationships relationships in adolescence ,pp. :@):?!-. Thousand 0aks' CA; 4age Pu#lications. ickerson' A. B.' 8 agle' R. . ,:""!-. Parent and peer attachment in late childhood and early adolescence. 4ournal of
.' Cro%ell' .' 8 Al#ershein' L. ,:"""-. Attachment security in infancy and early adulthood; A t%enty)year t%enty)year longitudinal study. Child Development* +' 6?)65.
:@
Waters' Waters' E.' Wippman' Wippman' .' 8 4roufe' . A. ,&5<5-. Attachment' positi*e affect' and competence in the peer group; t%o studies in construct *alidation. Child Development* 79' 6:&)6:5. Weimer' Weimer' B. L.' 7erns' 7. A.' 8 0lden#urg' C. M. ,:""@-. Adolescents+ interactions %ith a #est friend; Associations Associations %ith attachment style. style. 4ournal of <=perimental Child Psychology* ' &":)&:". Williams' Williams' 4. 7.' 8 7elly' 2. 2. >. ,:""!-. Relationships among in*ol*ement' in*ol*e ment' attachment' and #eha*ioral pro#lems in adolescence; E$amining father+s influence. 4ournal of relationships relationships !ith mothers* fathers and friends. Chicago' llinois; 9ni*ersity of Chicago Press.
Ta#le &. "eans &standard deviations/ deviations/ for adolescent ratings of attachment by person and dimension of attachment.
Dimensions (range 1-5) Trust Communication Alienation Total
Mother
Ratings by person 2ather
?.@ ,.<<@.< ,.<[email protected] ,.<:?." ,.!-
?.& ,.6"@.@ ,.5&@.? ,.5<@. ,.
Peers ?.: ,.&?.? ,.!
Ta#le :. "eans &standard deviations/ deviations/ for attachment ratings bet!een mothers and adolescents and fathers and adolescents by dimension of attachment.
:?
Dimensions (range 1-5) Trust Communication Alienation Total
Adolescent to Mother
Ratings by person Mother to to Adolescent Adolescent to 2ather
2ather to Adolescent
?.@ ,.<<@.< ,.<[email protected] ,.<:?." ,.!-
?.: ,.?6@.< ,[email protected] ,.!@.5 ,.?"-
?.& ,.!:@. ,[email protected] ,.6"@.5 ,.!:-
?.& ,.6"@.@ ,.5&@.? ,.5<@. ,.
Ta#le @. Correlations of adolescents ratings of attachment to each person by dimension of attachment.
Dimensions Trust Communication Alienation Total
Ratings by person Mother and 2ather Mother and Peers
2ather and Peers
.? @ Q .? : 5Q .! " ! Q .? 5 :Q
.&? 6 .?& "Q ) ." " 5 .:< "
." 6? .& @ ! .: & .& @ 5
Q p ."!' QQ p ."&
:!
Ta#le ?. Correlations Correlations of attachment ratings bet!een mother and an d adolescent and father and adolescent by dimension of attachment.
Ratings by person Moth other and Adol dolescent 2ather her and Adolescent Dimensions Trust Communication Alienation Total
.! " Q .: ? 6 .&6? .?66 Q
.&&" .@? 5 .:& .& < 5
p .&"' Q p ."!' QQ p ."&
Ta#le !. Correlations Correlations of attachment ratings to adolescent bet!een mother and father by dimension of attachment.
Ratings by person Mother and 2ather Dimensions Trust Communication Alienation Total
.! : : Q ." ? 6 .! @ Q .@6&
Q p ."!' QQ p ."&
:
2igure &. "eans for ratings of attachment bet!een father and adolescent* by grade and gender of adolescent* for the trust dimension. Male 2emale Gender and Grade of Adolescent
! ?.! ? @.! s g n i @ t a R f :.! o e u : l a O &.!
Adolescent to 2ather 2ather to Adolescent
& ".! " 6t h
&"t h
6t h
&"t h
:<
2igure :. "eans for ratings of attachment bet!een father and adolescent* by grade and gender of adolescent* for total attachment score. Male 2emale Gender and Grade of Adolescent
! ?.! ? @.! s g n i @ t a R f :.! o e u : l a O &.!
Adolescent to 2ather 2ather to Adolescent
& ".! " 6t h
&"t h
6t h
&"t h
:6
i
Although %ill focus on these three attachment #eha*iors' it is necessary to note that some researchers #elie*e there are more than three. ii Ains%orth and others consistently found a group of children %ho %ere unclassifia#le. As a result' a fourth category' the disorgani(edJdisoriented type' %as introduced. Children of this classification sho% #eha*iors #eha*iors of #oth the am#i*alent and a*oidant types. %ill not #e discussing this category. category. iii ran all analyses using 1ust the &5 families' ho%e*er results %ere not different= therefore' to o#tain higher po%er' all analyses that did not include these father ratings of attachment to adolescent utili(ed all :? original families.
