The expression and understanding of jealousy in children

The expression and understanding ofjealousy in children with autism

NIRIT BAUMINGERBar-Ilan University

AbstractWe investigated the expression and understanding of jealousy in 16 high-functioning children with autism and 17typically developing children matched for IQ, chronological age, gender, and maternal education. We examined theexpression of jealousy via children’s behaviors, verbalizations, and affects demonstrated during twojealousy-provoking triadic scenarios~drawing and playing! enacted among the child in the experimental group~autism or typical!, that child’s main caregiver~mostly mothers!, and a familiar peer or sibling. The two scenarioscorresponded with the two types of jealousy described in past studies: social-comparison jealousy~drawingscenario! and social-relational jealousy~playing scenario!. To tap children’s understanding, we asked them toidentify jealousy from a picture, to provide examples of times they felt jealous, and to offer suggestions for copingwith jealousy. The main results revealed that children with autism expressed jealousy in situations similar to theirtypical age mates but manifested it in different behaviors. Moreover, children with autism revealed a less coherentunderstanding of the feeling. We discuss the meaning of the gap between demonstrating and understanding jealousyin light of the two central theoretical views conceptualizing the core emotional deficit in children with autism.

Jealousy is a complex, unpleasant feeling thatis highly dependent on social context and theability to make spontaneous comparisons~Izard, 1991; Miller, Volling, & McElwain,2000!. By definition, an individual experi-ences jealousy when a potential threat existsthat a valued relationship will be lost to arival ~Izard, 1991; Parrott, 1991; Salovey &Rothman, 1991!. Thus, jealousy primarily ap-pears in triadic contexts involving the jealousindividual, the rival, and the valued relation-

ship with a significant other~Hansen, 1991;Parrott, 1991!.

Although researchers have often examinedjealousy~mainly romantic jealousy! in adults,much less work has focused on childhood jeal-ousy. One issue that complicates the study ofjealousy~like other complex emotions such asembarrassment or pride! concerns the vague-ness of the facial indices related to jealousy,which precludes identification of jealousybased solely on facial expression. Researcherstend to agree that the affective expression ofjealousy comprises a composite of several emo-tions ~e.g., fear, frustration, sadness, anger!;however, agreement is lacking about the spe-cific emotional components that constitute jeal-ousy~Arnold, 1960; Hupka, 1984; Izrad, 1991;Parrott, 1991!. Because of this unreliability offacial indices associated with jealousy, re-searchers believe its manifestation is orga-nized in a script-based pattern rather than as asimple blend of basic level emotions. There-fore, studies have focused on the identifica-

The Internal Grants Program of Bar-Ilan University sup-ported this research. The author extends special thanks tothe children and families who were willing to take part inthis study and expresses her appreciation to Dee B. Ankon-ina for her editorial contribution and to Dov Har-Even forhis statistical assistance. The author would also like tothank Galit Halevy–Tendler for her inspiring remarks, CoryShulman for her help in data collection, and the four anon-ymous reviewers for their thought-provoking comments.

Address correspondence and reprint requests to: NiritBauminger, School of Education, Bar-Ilan University,Ramat-Gan 52900, Israel; E-mail: [email protected].

| |DPP16~1! 445 1021 02013004 11:40 am REVISED PROOF

Development and Psychopathology16 ~2004!, 157–177Copyright © 2004 Cambridge University PressPrinted in the United States of AmericaDOI: 10.10170S0954579404044451

157

tion of situations that provoke jealousy inchildren and on the detection of behaviors andaction components that indicate jealousy inthese situations~Masciuch & Kienapple, 1993!.

When considering situations that provokejealousy, another issue to be taken into ac-count constitutes the obscurity of the distinc-tion between jealousy and envy~Parrott, 1991!.Jealousyalwaysinvolves a triadic situation inwhich the child’s own loss equals the rival’sgain, and it involves complex projections aboutthe self in regard to others. In contrast, envymay involve only two-person situations, andthis feeling comprises the wish to have an-other person’s possession or success and0orthe wish that the other person did not possessthis desired characteristic or object~Parrott,1991!. According to these definitions, the ma-jor differentiation between the two feelingsconsists of jealousy’s necessary loss of a rela-tionship within a triadic situation, whereas envydoes not require this loss. Furthermore, Par-rott and Smith~1993! have suggested that inenvy one’s own appraisal leads to dissatisfac-tion with oneself whereas in jealousy the re-flected appraisal or attention of another leadsto a lack of security and confidence. However,complicating the differentiation is that in jeal-ousy the perceived threat may not necessarilyinvolve the loss of love, and the child mayexperience jealousy related to the significantother’s appreciation of the rival’s higher suc-cess~Parrott, 1991!. Moreover, both jealousyand envy are concerned with losses of self-esteem stemming from social comparison,demonstrate similar behavioral manifesta-tions, and may co-occur in the same situations~Bers & Rodin, 1984; Parrott, 1991; Salovey& Rodin, 1984; Silver & Sabini, 1978!.

In an attempt to better differentiate jeal-ousy from envy in children, investigators de-vised two different situations within a socialtriad to distinctly elicit jealousy. One focuseson the child’s loss of love and0or attention,which is social-relations jealousy. The otherfocuses on the child’s loss of admiration be-cause of another child’s higher success, whichis social-comparison jealousy. Situations ofsocial-relations jealousy challenge one’s ex-clusivity in a relationship, whereas situationsof social-comparison jealousy challenge the

child’s superiority or equality~Bers & Rodin,1984; Masciuch & Kienapple, 1993; Milleret al., 2000!.

Distinct behaviors and actions have beenidentified as indices of jealousy among chil-dren of different ages. In infancy throughpreschool, children in both types of jealousy-provoking situations~social comparison andsocial relations! evidence behaviors such asgazing directly at their main caregiver~moth-ers and0or fathers! and0or the other child,discontinuing work and focusing attention onthe triad, frowning, making attempts to inter-fere with or enter into the rival interactionusing attention-provoking behaviors, takingthe other child’s objects, hugging or climbingon the main caregiver, answering questionsthat were addressed to the other child, attempt-ing to correct the other child, or trying tochange the situation by complaining~Masci-uch & Kienapple, 1993; Miller et al., 2000!.As children reach school age, other behaviorsalso emerge, such as comparing oneself tothe other child, expressing a desire for theobject or relations, attempting to protect one-self from being demeaned, behaving nega-tively or making negative comments towardanother person or about oneself, and attempt-ing to do at least as well as~equalization! orbetter than~competition! the rival ~Bers &Rodin, 1984!. It should be noted that preado-lescents and adolescents, in particular, are so-cialized to show their anger, distress, fear, oranxiety more indirectly through more subtlebehaviors rather than direct actions or ex-plicit facial expressions~Blumberg & Izard,1991; Harris, 1989!. Thus, children’s capac-ity for spontaneous comparisons~which al-lows for jealousy! increases with age, whereastheir overall explicit negative affect associ-ated with jealousy diminishes with age~Bers& Rodin, 1984!.

Wide consensus exists that emotional dif-ficulties comprise one of the chief character-istics of the autism syndrome, manifestingthemselves in both the expression and under-standing of emotion. However, the affectiveversus cognitive nature of these difficulties re-mains ambiguous~Happé, 1994; Hobson,1993a; Travis & Sigman, 1998!. Although anextensive body of research has investigated


158 N. Bauminger

simple emotions such as happiness, sadness,fear, and anger~see review in Dissanayake &Sigman, 2001!, few studies have focused onthe understanding of complex, self-consciousemotions such as pride, embarrassment, andguilt ~e.g., Capps, Yirmiya, & Sigman, 1992;Kasari, Chamberlain, & Bauminger, 2001;Yirmiya, Sigman, Kasari, & Mundy, 1992! oron the behavioral manifestations and expres-sions of such emotions~e.g., Dawson & Mc-Kissick, 1984; Kasari, Sigman, Baumgartner,& Stipek, 1993; Spiker & Ricks, 1984!. Inself-conscious emotions, the child expressesawareness or concern for others’ evaluations.These emotions hold particular importance forstudying the nature of the emotional deficit inautism, in that they touch upon one of thesechildren’s major difficulties: the metarepresen-tation of one’s own and others’ mental states~Baron–Cohen & Swettenham, 1997; Hob-son, 1993a!. Complex emotions involve pro-jection of one’s own mental state vis-à-visothers. For example, a child may be hurt, sad,or even angry after slipping and falling butwill feel embarrassed if an audience observedthe situation. In Kasari, Sigman, Baumgart-ner, and Stipek’s~1993! study on pride, youngchildren with autism could express pleasurefrom their success in completing a puzzle asoften as typically developing children, but theyfailed to share their success or to look for praisefrom another and even looked away when theywere given praise~which is different than typ-ical controls!. Thus, Kasari and colleagues con-cluded that autism may involve a specificaffective deficit that is related to emotions andsituations that contain some social interactivecomponent.

Like pride, jealousy comprises a self-reflective, socially mediated, complex emo-tion that is highly dependent on the individualinterpretation of social reactions. Jealousy oc-curs when one believes that a significant otherprefers a rival; thus, like in pride, one’s ownmental state is reflected vis-à-vis the reflec-tion of another person’s mental state. How-ever, jealousy seems to require multipleinferences on the part of the child. It appearsthat the child needs to infer someone else’smental perspective~significant other! towardtwo individuals~rival and oneself!. Further-

more, conceptualization of a rival relation-ship seems to call for an understanding orbeliefs regarding the quality of the interper-sonal relations between the jealous individ-ual and a significant other~i.e., mother!, aswell as the understanding of the interpersonalrelationship between the significant other anda rival ~i.e., peer or sibling!. Thus, jealousyappears to require inferences regarding a net-work of interpersonal relationships involvingthe self and others~Volling, McElwain, &Miller, 2002!. Moreover, the experience ofjealousy involves the loss of “formative atten-tion,” which influences the particular aspectsof the self that are intrinsically interpersonal~Neu, 1980; Tov–Ruach, 1980!. Thus, one mayconsider the child’s “interpersonal self” to beimportant for jealousy to occur~i.e., the par-ticular aspect of the self that concerns theself as distinct from others, as socially effec-tive, and as an object of others’ regard; Neisser,1988!. Hobson~1990! and Lee and Hobson~1998! suggested that the development of theself in relation to the physical world~i.e., the“ecological self” according to Neisser, 1988!is intact in children with autism, whereas theirinterpersonal self fails to develop normally,resulting in difficulties in self-conscious emo-tions such as jealousy.

The Present Study

The social complexity and cognitive and af-fective prerequisites of jealousy enable itsuse as a means to shed light upon severaldebated issues in the affective deficit of chil-dren with autism. In order to attain a thor-ough understanding of both the expressionsand understanding of jealousy, the presentstudy combined observations of children’s be-havioral reactions to two in vivo triadic situ-ations with an actual rival, which were knownto engender jealousy in typically developingchildren, with the examination of children’ssocial cognitive processing of jealousy suchas its recognition and the experience of andcoping with jealousy.

The first issue of debate relates to the natureof the affective deficit in autism. Inasmuch asjealousy involves both social–affective capa-


Jealousy in autism 159

bilities ~e.g., the ability to form an interper-sonal relationship with a significant other, theability to perceive the evolving rival relation-ship as interpersonal! and social–cognitive ca-pabilities~e.g., self-evaluation vis-à-vis others,sense of competition, possessiveness!, it is verylikely that both theories concerning the under-standing of autism~affective and cognitive!would predict difficulties in the expression andunderstanding of jealousy. The affective view~Hobson, 1993a, 1993b; Rogers & Penning-ton, 1991! highlights the child’s disturbance inintersubjective personal engagement with oth-ers, which is the lack of intersubjective sharingin autism, which seriously disrupts the child’sability to experience or understand interper-sonal relationships as such. Thus, this approachwould emphasize the child’s inability to graspthe relations between the mother and a rival asinterpersonal and as relations that could poten-tially jeopardize the child’s interpersonal rela-tions with the mother. The cognitive view, thatis, the theory of mind explanation to the affec-tive deficit in autism, would emphasize thechild’s inability to take another person’s viewsinto account, leading to difficulties in attribut-ing mental states to others and to oneself in re-gard to others~e.g., Tager–Flusberg, 2001!.Thus, such difficulties would necessarily im-pede the manifestation or understanding of self-reflective emotions such as jealousy.

In line with research in typical develop-ment, the present study implemented two dif-ferent jealousy-provoking social situations totest these two theories in children with au-tism. The social-relations scenario comprisedan affectively laden situation designed to evokejealousy as the child’s reaction toward an in-terpersonal interaction~shared play! betweenanother child~a rival! and the mother, whoignored her own child. If the affective ap-proach best characterizes the affective deficitin autism, then the child would be expected toact as a “behaviorist”~standing outside rela-tionships and only watching behaviors; Hob-son, 1993b!; therefore, jealousy would beexperienced less in this situation. In the “social-comparison scenario,” the mother praised an-other child’s drawing~but did not actuallysocially interact with that rival! while ignor-ing her own child’s drawing. This scenario

aimed to challenge the child’s sense of self-evaluation vis-à-vis the reflection of anotherperson’s mental state~e.g., “I believe shethinks0feels my drawing is not as good.”!.Thus, if theory of mind best characterizes thenature of the affective deficit, children wouldbe less likely to experience jealousy in thissituation. The different foci of the two situa-tions ~interpersonal vs. cognitive processes!aimed to help elucidate the nature of the emo-tional deficit in autism.

We also posed the second question of devi-ant versus delayed expression and understand-ing of jealousy. It is already well documentedthat high-functioning children with autism pos-sess higher social–emotional capabilities com-pared with their low-functioning peers~e.g.,higher prosocial abilities and emotional respon-siveness; Bacon, Fein, Morris, Waterhouse, &Allen, 1998!. Furthermore, in the understand-ing of social emotions such as empathy and em-barrassment, children with autism compensatefor their emotional deficit by utilizing theirhigher cognitive capabilities, which is the cog-nitive compensation hypothesis~Capps et al.,1992; Kasari et al., 2001; Yirmiya et al., 1992!.Finally, certain capabilities develop later inthese children, for example, theory of mindcapabilities~Happé, 1995!. Considering allthese previous studies together, we hypoth-esized that the ability to express a self-reflectiveemotion such as jealousy may already bepresent among older~preadolescent and ado-lescent! high-functioning children with au-tism, even if a similar self-reflective emotionsuch as pride was not shown to exist yet inyounger, low-functioning children with autism.

A third issue of particular interest was theexamination of differences between childrenwith autism and children with typical devel-opment in regard to their expressions and0orunderstanding of jealousy. According to Lewis~1993!, children may sustain an emotional statebut not necessarily experience consciousawareness of that state. For example, chil-dren as young as 2–3 years can express pridein the presence of others, but it is not until 7or 8 years that they recognize the role thatothers play in the evaluation of their ownaccomplishments. In addition, the majority ofschool-aged children’s descriptions of pride-


160 N. Bauminger

evoking events do refer to an audience~Kasari,Sigman, Yirmiya, & Mundy, 1993; Seidner,Stipek, & Fesbach, 1988!. Thus, the investi-gation of children’s expressions and under-standing of jealousy can help clarify whethera gap exists between a more automatic behav-ioral process of affective expression and ahigher level, conscious-awareness process thatencompasses children’s ability to understandthe factors eliciting this emotion, includingthe role of others.

The present study sought to examine theexpression and understanding of jealousy inhigh-functioning children with autism. Morespecifically, the study posed three objectives:to describe the manifestations of jealousy inchildren with autism compared with typicallydeveloping age-matched children, to explorethe differences in understanding jealousy be-tween the two samples, and to examine howunderstanding and exhibiting jealousy is re-lated to children’s mental age.

Method

Participants

A sample of 33 preadolescents and adoles-cents participated in the study, including 16~2female! high-functioning individuals with au-tism and 17~2 female! typically developingindividuals. All participants in the autism sam-ple met the criteria for autism on the AutismDiagnostic Interview—Revised~ADI-R; Lord,Rutter, & Le Couteur, 1994!. In addition, allchildren but one had the diagnosis of autismbased on theDiagnostic and Statistical Man-ual of Mental Disorders~DSM-IV; AmericanPsychiatric Association, 1994! prior to theirparticipation in the study as determined by li-censed psychologists unassociated with the cur-rent study. One child was diagnosed withAsperger syndrome prior to his participationin the study, but he met the ADI-R criteria forautism. TheDSM-IV criteria included~a! on-set prior to 36 months of age;~b! qualitativeimpairment in social interaction;~c! qualita-tive impairment in communication~e.g., def-icits or abnormalities in language developmentor deficit in play, particularly symbolic play!;and ~d! restricted and repetitive stereotyped

behaviors, which may include bizarre re-sponses to various aspects of the environ-ment, such as resistance to change. The authoradministered the ADI-R to the parents of thechildren to verify diagnosis and to provide ad-ditional information about the children’s de-velopmental histories. The ADI-R focuses onmeeting criteria for autism in three main areas:reciprocal social interaction; communicationand language; and repetitive, restrictive, andstereotyped behaviors. The child also needs toshow evidence of developmental delay or de-viance prior to the age of 36 months. All 16children met the criteria for autism on all fourADI-R criteria.

The mean age was 11.14 years~SD5 3.01!for the children with autism and 11.51 years~SD 5 2.62! for the typically developingchildren. Mean full-scale IQ scores, as mea-sured on the Wechsler Intelligence Scale forChildren—Revised~Wechsler, 1974!, were92.81 ~SD 5 14.15! for the children withautism and 98.35~SD 5 7.19! for the typi-cally developing children. We matched thegroup of typically developing children to thechildren with autism on chronological age,mental age, all IQ scales~full, verbal, andperformance!, gender, and maternal educa-tion. As can be seen in Table 1, studentt testsrevealed no significant differences betweengroups regarding any demographic variables.We recruited the children with autism throughthe Special Education Department in the Is-raeli Ministry of Education. We recruited typ-ical children from local public schools.

Measures

To examine children’s expressions of jealousyin the two different jealousy-provoking situa-tions, we manipulated and videorecorded twoexperimental triadic scenarios. Assessment ofchildren’s jealousy-provoked behaviors, ver-balizations, and affects utilized three codingscales: explicitness of jealousy manifesta-tions, quantity of different jealousy behaviors,and response time. To investigate the under-standing of jealousy, we asked children to rec-ognize jealousy from a picture, to generateexamples of times they experienced jealous



feelings, and to suggest ways for coping withsuch feelings.

Jealousy-provoking experimental scenarios.Based on Masciuch and Kienapple~1993!, wemanipulated two experimental scenarios in thecurrent study to provoke jealousy in the chil-dren. The drawing scenario corresponded withsocial-comparison jealousy, and the playingscenario corresponded with social-relationsjealousy. Each scenario included a triad con-sisting of the child in the experimental group~autism or typical!, his or her main caregiver~all mothers except for two families having achild with autism, where the father, who wasthe main caregiver, participated!, and anotherfamiliar child ~the rival child! who was eitherthe child’s friend or sibling. The interchange-able participation of siblings and peers in thescenarios, in accordance with parents’ prefer-ences and accessibility considerations, was jus-tifiable because of the procedure’s similarability to provoke jealousy in children withboth a sibling and a peer~Masciuch & Kienap-ple, 1993; Miller et al., 2000!. Ten parents ineach sample preferred to implement the sce-narios with a sibling.

In the drawing scenario~i.e., social-comparison jealousy!, based on Masciuch andKienapple~1993!, we gave each child a boxwith colored markers and a blank sheet ofpaper and instructed the child to complete adrawing of his or her choice. Parents re-

ceived prior written instructions~a! to sit inclose proximity to the two children,~b! tocomplete a demographic questionnaire whileignoring the two children, and~c! to praisethe rival child’s picture while ignoring his orher own child upon the children’s completionof their drawings. Videotaping began whenthe parent started to praise the rival child andlasted 5 min.

In the playing scenario~i.e., social-relationsjealousy!, adapted from Masciuch and Kienap-ple’s ~1993! reading scenario, we told chil-dren that they were going to play separatelywith some games. We instructed children tochoose a game from a box that included sevengames: a falling tower construction game, anoptical viewing device with a choice of differ-ent 3-dimensional slides of animated moviescenes, two different assembly toys, domi-noes, one magnetic construction game, and onemagnetic mosaic game where the child neededto copy a shape from a picture. Prior to thebeginning of the scenario, parents receivedwritten instructions~a! to sit in close proxim-ity to the children,~b! to ignore both childrenwhile completing a questionnaire for 2 min,and ~c! after 2 min to join the rival child inlively affectionate play for 5 min while ignor-ing his or her own child. Videotaping beganwhen the parent joined the rival child and lasted5 min.

We adapted the playing scenario from Ma-sciuch and Kienapple’s~1993! reading sce-

Table 1. Sample characteristics

Autism~n 5 16!

Typical~n 5 17!

M SD M SD Significance

Chronological age~months! 133.75 36.17 138.18 31.53 nsMental age~months! 121.93 28.37 135.33 29.54 nsFull-scale IQ 92.81 14.15 98.35 7.19 nsVerbal IQ 91.75 15.36 95.06 5.38 nsPerformance IQ 95.69 14.74 102.53 11.50 nsMale0female ratio 1402 1502 nsMother’s education 4.50 1.36 3.70 1.16 ns

Note: IQ and mental age scores are based on the WISC-R. Mother’s education was calculated on a 1–6 scale~15 lessthan 8 years of study, 25 high school without matriculation, 35 high school matriculation, 45 special professionaltraining after high school, 5 5 bachelor’s degree, 6 5 second degree and above!.


162 N. Bauminger

nario in order to suit the ages of the childrenin the current sample. The original study onpreschoolers instructed the mother to place therival child on her lap while reading him or hera story, which would have been inappropriatefor preadolescents and adolescents. Prior tothe beginning of the present study, we imple-mented a pilot study of the current playingscenario with children unrelated to the presentsample~three typically developing children andthree high-functioning children with autism!.The pilot study confirmed that the playing sce-nario provoked behaviors and verbalizationssimilar to the ones reported in the reading sce-nario by Masciuch and Kienapple.

Coding of videotaped jealousy expressions.Asmentioned above, we assessed children’sjealousy-provoked behaviors, verbalizations,and affects using three coding scales: explic-itness, quantity of different jealousy behav-iors, and response time.

Explicitness: Hierarchical jealousy scale.We developed a hierarchical jealousy scale forthe purpose of this study to measure the ex-plicitness of the jealousy manifestations evi-denced by the child during each of the twoscenarios. The scale derived from the behav-iors, verbalizations, and affects identified asjealousy indices by previous research~e.g.,Bers & Rodin, 1984; Hupka, 1984; Masciuch& Kienapple, 1993; Miller et al., 2000!. Thisscale included six scores describing the behav-iors, verbalizations, and affective expressionsof jealousy in hierarchical order from an ab-sence of explicit jealousy indices to the mostexplicit indices of jealousy. Scoring was asfollows: a score of 1 indicated that the childdid not seem to pay attention to any of theongoing scenario; a score of 2 indicated onebrief eye gaze at the parent, rival child, ordyadic interaction; a score of 3 indicated onelong gaze or a number of several short eyegazes directed at the parent, rival child, ordyadic interaction, with or without stoppinghis0her own activity; a score of 4 indicatedbehaviors or verbal comments that indirectlyintervened into the interaction between the par-ent and the rival child, such as asking the par-ent different questions or grabbing objects from

the other child’s hand; a score of 5 indicateddirect behaviors or verbalizations that focusedparent’s attention on the drawing or playing ofthe child in the experimental group, such asputting one’s picture in front of the parent’seyes after the parent finished praising the otherchild’s picture, with or without asking the par-ent to look at the picture or game; and a scoreof 6 indicated a direct declaration of compar-ison and lack of equality, with or without frus-tration, such as “Mom, why don’t you alsoplay with me?” and0or when a child expressednegative affects such as frustration, anger,crying, sadness, or a depressed facial expres-sion as a reaction to the mother’s behavior. Itshould be noted that behaviors and verbaliza-tions were rated as 5 for only one child, butboth coders gave him a final score of 6 basedon his negative affective reaction~intenselyfrustrated tone of voice and facial expres-sion!. In all other cases, children’s negativeaffective reactions to the mother coincided witha full, explicit behavioral and0or verbal ex-pression of jealousy.

On this scale, separately for each of thetwo scenarios~drawing and playing!, the childwas assigned the highest score evidenced overthe 5-min scenario. A score of 4 and aboveindicated explicit behaviors, verbalizations, andaffects that reflected jealousy whereas a scorebelow 4 indicated only eye gaze in differentdegrees. Appendix A presents a more system-atic description of children’s jealousy re-sponses on the hierarchical scale.

Two different trained coders rated all of thechildren’s responses separately for the draw-ing and the playing scenarios. Pearson corre-lations between the two coders were .82 forthe drawing scenario and .81 for the playingscenario.

Quantity of different jealousy manifesta-tions: Behavioral coding category scale.Weutilized the behavioral coding category scale~Masciuch & Kienapple, 1993! to assess thequantity of jealousy behaviors and vocaliza-tions of different categories that we observedin each of the two 5-min jealousy scenarios~drawing and playing!. The scale included 10indices of jealousy comprising three main cat-egories: the child’s gaze direction, verbaliza-



tions, and actions. We separately counted andsummed up the total number of jealousy indi-ces observed for each of the three categoriesduring the 5 min of observation for the draw-ing and playing scenarios; thus, a higher scorein a particular category indicated a higher quan-tity of that category of jealousy manifesta-tions. The child’s gaze direction categoryincluded two main gaze behaviors: the eyes ofthe child in the experimental group directed atparent and0or directed at the peer0sibling. Thechild’s verbalizations category included fivecomponents:~a! attention seeking: the childmakes verbal attempts to draw the parent’s at-tention to self or to the child’s own drawing0playing, such as “My drawing is prettier,”“Mom, we used to play this together”;~b! self-deprecatory: the child makes comments thatrefer negatively to self or the child’s owndrawing0playing, such as “My picture is ugly”;~c! prosocial comments: the child makes pos-itive comments about the rival child’s drawing0playing, such as “That’s such a nice game”;~d! interactive comments: the child makes com-ments that enter into the ongoing actions andconversations between the parent and the rivalchild, such as “Yes, I remember that familytrip to Eilat, do you remember diving with thedolphins?”; and~e! negative comments: thechild makes comments that signal contemptor personal disregard for the rival child or theparent, such as “Your assembly game is veryeasy compared to mine.” The category forchild’s actions included three components:~a!desisting0modifying activity: the child stopsplaying when the parent joins the interactionwith the rival child, or the child tries to im-prove his0her own drawing after the parentpraises the other child’s~i.e., resumes draw-ing after having stopped earlier!; ~b! attentionseeking: the child takes actions to draw theparent’s attention to self or to his0her owndrawing0playing, such as placing his0her pic-ture in front of the parent’s eyes or caressingthe parent’s hair; and~c! involvement behav-iors: the child attempts to physically inter-vene in the interaction between the parent andthe rival child, such as initiating play with thatdyad.

Two observers underwent training in cod-ing the three categories~children’s eye gaze

behaviors, verbalizations, and actions! usingthe six pilot study videotapes~three autismand three typical development!, until an inter-observer agreement level of 85% or higherwas obtained on each of the scale’s three cat-egories for both of the scenarios~drawing andplaying!. Then, these two observers indepen-dently rated a randomly selected 50% of chil-dren’s responses across participants andscenarios. The intraclass correlation coeffi-cients for the drawing scenario were .99 foreye gaze and .98 for verbalizations and foractions. The intraclass correlation coeffi-cients for the playing scenario were 1.00 foreye gaze, .98 for verbalizations, and .90 foractions.

Response time.We measured the responsetime to examine whether children with au-tism, because of their well-documented emo-tional difficulties in linking emotions withsocial situations~Dennis, Lockyer, & La-zenby, 2000!, would need a longer duration torespond in both scenarios compared to theirtypical age mates. We coded children’s initialjealousy response on the jealousy behavioralcoding scale in seconds and separately for eachof the two scenarios~drawing and playing!.

Assessing the understanding of jealousy.Weassessed the understanding of jealousy throughthree tasks measuring the child’s ability to rec-ognize jealousy in a picture, elicit examplesof different situations that provoke jealousy,and provide strategies for coping with jeal-ousy. In the first task, in order to assess therecognition of jealousy, we showed childrena color drawing depicting a typical social-relations triadic scenario~similar to Milleret al., 2000!, in which a mother is hugging hernew baby while an older sibling is watching.Fifteen typically developing preadolescents andadolescents~unrelated to the present study!who viewed the picture in a prior pilot studycould easily recognize that the older siblingwas jealous and wanted to obtain the mother’sattention as well. After looking at the picture,we asked the children to identify the child’s~older sibling’s! feelings in the picture. A scoreof 0 indicated erroneous identification of thechild’s feeling~e.g., very happy, joyful!, a score


164 N. Bauminger

of 1 indicated abasicfeeling sharing the samehedonic tone~e.g., sad, angry!, a score of 2indicated acomplexfeeling from the same he-donic tone~e.g., lonely, neglected!, and a scoreof 3 indicated the recognition ofjealousyinthe picture.

In the second task, the understanding of thedifferent situations that elicit jealousy, we askedchildren to provide examples of situations inwhich they or other people would feel jealous.We coded children’s generated examples ofjealousy first in line with the two theorizedtypes of jealousy~social relations and socialcognitive! and second in line with the per-sonal immediacy of the example given~a per-sonal example vs. an example about otherpersons!. For the first analysis, we coded af-fective jealousy~corresponding to social-relations jealousy! when participants’examplesindicated that jealousy involved negative feel-ings associated with a child’s responses to asocial triangle in which the parent, another fa-miliar adult ~e.g., teacher, grandmother, orgrandfather!, or a peer paid exclusive atten-tion to another peer or a sibling~Masciuch &Kienapple, 1993!. Affective jealousy reflectssituations in which the child’s exclusivity in arelationship is threatened.

We coded social–cognitive jealousy~corre-sponding with social-comparison jealousy!when participants’ examples indicated thatjealousy arose when one child enjoyed moresuccess or possessions compared with an-other child, challenging the first child’s supe-riority or equality ~Bers & Rodin, 1984!.Children obtained a score of 2 if they pro-vided both types of examples~affective andsocial–cognitive jealousy!, a score of 1 if onlyone type was specified, and a score of 0if neither type was specified. Appendix Bpresents examples of children’s suggestions fortheir experiences of jealousy according to thejealousy type.

In addition, to determine whether the par-ticipants who included both types of jealousyin their examples did so because they weremore verbally productive, we computed a pro-ductivity score for each participant to mea-sure verbal output regarding their descriptionsof jealousy experiences. Similar productivityrates in the two groups would eliminate the

possibility that verbal output rendered con-founding effects on the dependent variable oftypes of jealousy. To obtain this score, we cal-culated the total number of words stated bythe child in the description of the experienceof jealousy. Thet-test analysis revealed thatno significant differences emerged between theautism group~sum5 470, range5 3–85,M 529.37, SD 5 25.04! and the typical group~sum 5 508, range5 6–70, M 5 29.88,SD 5 17.73! with regard to productivity,t ~31! 5 .07,p . .05.

The intraclass correlation coefficients cal-culated between two raters who indepen-dently coded all of the children’s responseswere 1.00 for social–cognitive jealousy and.96 for affective jealousy. Raters discussed alldisagreements on the affective scale until theyattained agreement.

We conducted another analysis for the sec-ond task with regard to the personal immedi-acy level demonstrated in children’s examplesof the experience of jealousy. This examina-tion compared children’s generation of per-sonal examples of jealousy, which was scored1 ~e.g., “I’m jealous of my friend because she’sprettier than me.”! versus children’s genera-tion of examples regarding other persons,which was scored 0~e.g., “When everyone getsthe credit, but he does not”!.

To further the assessment of jealousy un-derstanding, the third task sought to assess thequality and quantity of children’s repertoire ofstrategies for coping with jealousy. We codedcoping strategies to deal with jealousy alongtwo dimensions: the number of suggestionsoffered by the child and the various contentareas suggested by the child. Because of thepaucity of research specifically investigatingchildren’s coping strategies for jealousy, thepresent content analysis utilized Harris’~1989!reported coping strategies implemented by chil-dren when dealing with similar unpleasant,painful feelings~e.g., sadness, disappoint-ment, distress!. Harris identified two main strat-egies for coping with an unpleasant feeling.First, children attempt to change a situationby moving to or creating a situation or activitythat is more enjoyable. Usually this will in-volve partial or total restoration of the loss.“When confronting sad feelings, children may



try to modify this feeling by deliberately en-gaging in an activity that is normally associ-ated with feeling happy”~Harris, 1989, p. 156!.Second, children may attempt to reduce an un-pleasant feeling by implementing cognitive–mentalistic techniques such as trying not tothink about the unpleasant event, forgetting it,or occupying the mind with other things. Byoffering the first solution, children acknowl-edge that a later positive event can mitigatean earlier negative emotion; by suggesting thesecond solution, children demonstrate aware-ness of the connection between cognition andemotion. According to Harris, the types of so-lutions offered by children change over devel-opment: younger children~at around age 6years! are more likely to offer the first solu-tion whereby unpleasant feelings are modi-fied by carrying out some activity; and by theage of 10 years or over, children are more likelyto suggest cognitive solutions as well.

Two coders examined the adaptability ofHarris’ ~1989! categories to the present sam-ple’s solutions for coping with jealousy. All ofthe children’s suggestions but four could becoded in line with one of the two criteria, andthere was 92% agreement between coders. Thefour exceptions all comprised aggressive so-lutions such as hitting the other child~sug-gested by a child with autism! or destroyingthe other child’s beautiful car~suggested by atypically developing child!. Thus, we con-sidered aggressive solutions as a separatecontent category; however, because of theirlow frequency, they were not included in theanalysis. Appendix C presents examples ofchildren’s suggestions for each of the twocategories.

Procedure

We contacted the parents of the children inboth the special and regular education set-tings through their school principal and0orthrough the child’s teacher. After obtainingwritten parental consent for participation, wearranged home visits by telephone with par-ents. We advised parents in advance aboutthe nature of the research and the need toarrange for another child to be present in thehome during the experimenter’s visit. We con-

ducted the research at the home of the childin the experimental group in a quiet roomduring one home visit. In half of the casesthe children completed the three-part jeal-ousy understanding task prior to enactmentof the two experimental jealousy scenarios,and the sequence was reversed for the otherhalf. No significant order effect or interactionof Group 3 Order emerged for any of theexpression or understanding variables. Tocounteract possible negative effects of thejealousy-provoking situations, we instructedparents to compliment the drawing producedby the child with autism or to join in thechild’s play immediately after we turned offthe videocamera.

Results

Expressions of jealousy

The first set of analyses examined the differ-ences in the explicitness, quantity, and re-sponse time of jealousy expressions~using thehierarchical scale, behavioral coding scale, andresponse time measure, respectively! betweenchildren with autism and typically developingchildren in the drawing and playing scenarios.

As can be seen in Table 2, which de-scribes the distribution of the two groups onthe jealousy hierarchical scale, the majorityof children in both groups expressed explicitbehaviors, verbalizations, and affects that in-dicated jealousy~i.e., a score of 4–6!. Onthis scale, a score of 4 and above indicatedexplicit behaviors, verbalizations, and affectsthat reflected jealousy whereas a score below4 indicated only eye gaze in different de-grees. A jealousy index of 4 or above for thedrawing and the playing scenarios emergedfor 88 and 75% of the children with autismand 67 and 73% of the typically developingchildren, respectively. Indeed, a 23 2 Group-ing ~autism0typical! 3 Scenario~drawing0playing! analysis of variance~ANOVA ! withrepeated measures on scenario, which we ex-ecuted to examine group differences for theexplicit expression of jealousy on the hierar-chical scale, was not significant for group ef-fect, scenario effect, or the interaction of groupand scenario. Children in both groups~au-


166 N. Bauminger

Table 2. Distribution of the two groups according to explicitness of jealousy scores on the hierarchical jealousy scalefor the drawing and playing scenarios

Drawing Playing

Autism Typical Autism Typical

Level of Explicitness of Jealousy n % n % n % n %

Level 1: no particular indication of jealousy 1 6.25 0 0.00 1 6.25 1 6.66Level 2: one brief eye gaze 1 6.25 4 26.66 0 0.00 0 0.00Level 3: long gaze or number of several short eye gazes 0 0.00 1 6.66 3 18.75 3 20.00Level 4: behaviors or verbalizations that indirectly intervene into the parent—rival

child interaction5 31.25 4 26.66 4 25.00 5 33.33

Level 5: direct behaviors0verbalizations aimed at focusing parent’s attention to theexperimental child’s drawing0playing

5 31.25 4 26.66 4 25.00 4 26.66

Level 6: direct declaration of comparison and lack of equality, or negative affect 4 25.00 2 13.33 4 25.00 2 13.33

||

DP

P1

6~1!4

45

1102

10

201300

41

1:4

0a

mR

EV

ISE

DP

RO

OF

16

7

tism and typical! displayed jealousy at a sim-ilar level of explicitness on the two scenarios~M 5 4.46,SD5 1.34 for drawing andM 54.31, SD 5 1.36 for playing in the autismsample, andM 5 3.80,SD5 1.37 for draw-ing andM 5 4.16,SD5 1.33 for playing inthe control group!.

Next, we conducted a 23 2 ~Group3 Sce-nario! multivariate ANOVA~MANOVA ! withrepeated measures on scenario to investigategroup differences on the quantity of the jeal-ousy behaviors observed for each of the threecategory types~gaze, verbalization, and ac-tion!. The results of the MANOVA revealed asignificant main effect of group,F ~Wilk’s cri-terion! ~3, 27! 5 7.20,p , .001. Neither themain effect of the scenario nor the interactioneffect of Group3 Scenario was significant.Table 3 presents the means and standard devi-ations for the three jealousy categories’ scoreson the behavioral coding category scale evi-denced by the children with autism and chil-dren with typical development. As can be seenin the table, univariate ANOVAs revealed sig-nificant group differences on gaze and action.Compared to typically developing children,children with autism were significantly lesslikely to look at the parent and0or the rivalchild but were significantly more likely to act

toward them, regardless of the scenario~draw-ing or playing!.

Further, to examine the differences amongthe three behaviors of the jealousy coding scale~eye gaze, verbalizations, and actions!, weperformed a 23 2 3 3 ANOVA ~Group 3Scenario3 Type of Jealousy Behaviors! withrepeated measures on the type of scenarios andbehaviors. The results of the ANOVA yieldeda significant behavior effect,F ~2, 58! 532.08, p , .001, and interaction effect~Group3 Behavior!, F ~2, 58! 5 5.87, p ,.01. The interactions for Type3 Scenarioand for Group3 Type 3 Scenario were notsignificant. To determine the source of thesignificant interaction, simple effect testsexamined the differences between the threejealousy behaviors~eye gaze, verbalizations,and actions! within each group. A signifi-cant F value emerged for each group, butthe difference between the behaviors in thetypically developing group,F ~2, 28! 522.18,p . .001, was higher compared withthe difference within the group of childrenwith autism, F ~2, 30! 5 9.11, p , .001.Indeed, a further set of paired comparisontests according to Newman–Keuls~ p , .05!within each group revealed significant differ-ences between all three jealousy behaviors for

Table 3. Means (M), standard deviations (SD), and F values for differences betweenchildren with autism and children with typical development on the behavioral codingcategory scale

Autism Typical

Draw Play Draw PlayF ~1, 29!Group

F ~1, 29!Scenario

F ~1, 29!Interaction

GazeM 3.87 4.75 5.53 7.60 4.10* 2.81 .46SD 2.57 2.93 3.60 5.92

VerbalizationM 2.43 2.81 1.87 3.80 .07 2.45 1.12SD 2.36 2.80 2.23 4.29

ActionM 1.81 2.10 .40 .73 14.87** 1.22 .02SD 1.47 1.65 .63 .79

Note:Group, autism0typical; scenario, drawing0playing. In the verbalization and action categories, severalSDs werehigher than their means; therefore, an additional nonparametric Mann–Whitney test for independent samples wasperformed for these cases, which mirrored the ANOVA results.*p , .05. **p , .001.


168 N. Bauminger

the typical control group and only betweeneye gaze and the other two jealousy behaviors~e.g., verbalizations and actions! for the au-tism group.

To examine the group differences on re-sponse time, we conducted a 23 2 ~Group3Scenario! ANOVA with repeated measures.We calculated the response time using loga-rithm values in order to decrease the large stan-dard deviations. The results of the ANOVAyielded a significant main scenario type ef-fect, F ~1, 29! 5 10.84,p , .01, and nonsig-nificant effects for group and for the interactionof scenario and group. Children in both groups~autism and typical! responded faster in thedrawing scenario compared with the playingscenario.

Last, nonparametric Mann–Whitney testsexamined within-group differences betweenthose scenarios implemented with siblings~n 5 10 in each group! and those imple-mented with peers regarding all three scalesfor the expression of jealousy~hierarchicalscale, behavioral coding scale, and responsetime!. No significant differences emerged onany of the scales for either the drawing or theplaying scenarios.

Correlations between mental age andexpressions of jealousy: Within-groupexamination

We computed the correlations between the chil-dren’s mental age and the three measures ofjealousy manifestation~hierarchical scale, jeal-ousy behavioral coding scale, and responsetime! in each group~autism and typical! forthe drawing and the playing scenarios. Fewsignificant correlations emerged for eithergroup. For children with autism, the child’smental age correlated negatively with the ac-tion scale for the playing situation~r 5 2.56,p , .05!. In the autism group, children with ahigher mental age were less likely to displayactions that expressed jealousy during the playscenario. In the typically developing group,during the play scenario, children with a highermental age were less likely to display jealousyexplicitly ~r 5 2.46, p , .05!. A Fisher Zanalysis revealed significant group differ-ences only in regard to the correlation be-

tween the child’s mental age and action on theplay scenario~FisherZ 5 2.00,p , .05; au-tism, r 5 2.56; typical,r 5 .06!.

Understanding jealousy

The next set of analyses focused on children’sunderstanding of jealousy along the three maindimensions tapped by the tasks: recognitionof jealousy in a picture, providing examplesof the experience of jealousy, and describingways to cope with jealousy. Regarding the firsttask, which assessed children’s ability to rec-ognize jealousy, the majority of children withtypical development~n 5 13; 76.5%! accu-rately recognized jealousy in the picture ver-sus only 4 children~25.0%! in the autismgroup. However, among the remaining 12 chil-dren with autism who could not recognize jeal-ousy in the picture, 7 children~comprising43.8% of the autism sample! were neverthe-less able to identify basic and complex emo-tions with an accurate hedonic tone~e.g., sad!,as were 3 out of the 4 remaining typically de-veloping children who had failed in the recog-nition scale~17.6% of the typical sample!. AnANOVA with group as the independent vari-able and the child’s score on the recognitionof jealousy as the dependent variable yieldeda significant group difference. Children withautism were less likely to recognize jealousyin the picture compared with typically devel-oping children,F ~1, 31! 5 12.39,p , .001~M 5 1.31,SD5 1.19, for autism, andM 52.59,SD5 0.87, for the typical group!.

Regarding children’s ability to provide ex-amples of the experience of jealousy, a chi-square analysis revealed that the two groupsdiffered in the proportion of children whoincluded both the affective and the social–cognitive jealousy types in the examples ofjealousy experiences that they provided,x2

~2, 33! 5 8.11, p , .05. Only 2 childrenwith autism ~12.5%! provided both types,versus 10 typically developing children~58.8%!. An ANOVA with group as the inde-pendent variable and the total jealousy exam-ple score as the dependent variable revealedthat, compared to typical children, childrenwith autism were significantly less likely tooffer both types of examples~affective and



social–cognitive jealousy!, F ~1, 31! 5 10.10,p , .01 ~M 5 1.06, SD 5 0.44, for autism,and M 5 1.58, SD 5 0.50, for the typicalgroup!. The next analyses separately exam-ined the proportions of each group that pro-vided examples of each type of jealousy. Thepercentage of affective jealousy examples dif-fered between the two groups, provided byonly 31.2% of the children with autism com-pared to 70.5% in the control group,x2 ~1,33! 5 5.10, p , .05. In contrast, no signifi-cant differences emerged in the percentagesof children who provided examples related tosocial–cognitive jealousy~75.0% in autismvs. 88.2% in the typical group, Fisher exacttest,ns!.

A significant difference appeared in the per-centages of children who provided personalexamples of jealousy versus children whoprovided examples of jealousy relating toother people. Only 56% percent of the chil-dren with autism~n 5 9! described per-sonal examples versus 100% of the childrenwith typical development~Fisher exact test,p , .006!.

Regarding the third task, children’s strat-egies for coping with jealousy, we conductedanalyses to examine the number of solutionssuggested by the child and the type of contentareas suggested. Children with autism sug-gested a lower number of solutions for copingwith jealousy compared to their typical agemates,F ~1, 31! 5 11.51,p , .01 ~M 5 1.06,SD50.57;M52.11,SD51.11, respectively!.We computed an ANOVA to test for group dif-ferences regarding the number of solutions sug-gested by the children in each of the two maincoping categories~situation activity andcognitive–mentalistic!. Children with autismsuggested fewer situation-activity solutionscompared with typically developing children,F ~1, 31!54.47,p, .05~M50.68,SD50.47;M 51.35,SD51.69, respectively!, and a sim-ilar number of cognitive–mentalistic solutions~M 5 0.31,SD5 0.47;M 5 0.59,SD5 0.87,respectively!. Because of the large standard de-viations found in comparison to the means forthe cognitive–mentalistic domain, we also per-formed a nonparametric Mann–Whitney test fortwo independent samples, which mirrored theANOVA results.

Correlations between mental age andunderstanding jealousy: Within-groupexamination

We computed the correlations in each group~autism and typical! between the children’smental age and the three tasks measuring theunderstanding of jealousy: picture recogni-tion, the child’s examples of jealousy, and thechild’s suggestions for coping with jealousy.Few significant correlations emerged for ei-ther group. For children with autism, the abil-ity to provide examples of social–cognitivejealousy and the ability to provide both typesof jealousy examples~both affective andsocial–cognitive! each correlated positivelywith the children’s mental age~r 5 .52, p ,.05; r 5 .60, p , .01, respectively!. Simi-larly, the ability for children with autism tosuggest cognitive–mentalistic solutions forcoping with jealousy correlated positively withchildren’s mental age~r 5 .56, p , .05!. Fortypically developing children, child’s mentalage correlated negatively with providing asituation-activity solution to cope with jeal-ousy and correlated positively with the abil-ity to provide cognitive–mentalistic solutions~r 5 2.42,p , .05; r 5 .45,p , .05, respec-tively!. The FisherZ test revealed significantgroup differences for the following correla-tions: mental age and children’s ability to pro-vide both types of examples of jealousy~Z 51.87,p , .05; r 5 .60 for autism;r 5 2.03,for typical! and mental age and the ability toprovide only the social-comparison examplesof jealousy~Z 5 1.69, p , .05; r 5 .52 forautism;r 5 2.07 for typical!.

Discussion

The present study explored the expressionand understanding of jealousy among high-functioning children with autism. The mainfindings revealed that children with autismexpressed jealousy in situations similar tothose that provoked jealousy in their typi-cally developing age mates~the experimentaldrawing and playing scenarios!, yet, the man-ifestations of jealousy differed in autism ver-sus typical development. Typically developingchildren demonstrated more eye gaze behav-


170 N. Bauminger

iors toward the parent and0or toward the ri-val child in each of the jealousy-provokingsituations~drawing and playing!, whereas chil-dren with autism displayed more actions toexpress their jealous feelings. Furthermore,children with autism exhibited a less coher-ent understanding of the feeling and of situa-tions that provoke jealousy, compared to thetypical control group. This discussion consid-ers the meaning of these differences betweenautism and typical development and dis-cusses the implications of these gaps for un-derstanding the emotional deficit in autism.

The finding that children with autism ex-pressed jealousy in both situations, which in-volved the risk of losing formative attention~play! and evaluation~draw!, may suggest thatthese older, high-functioning children’s aware-ness of themselves as objects for others’ eval-uations and0or concerns~“interpersonal self”!is developed, even if expressed differently thanamong typical controls. What was absent fromthe capacity of 2-year-old children with au-tism and mental retardation, namely indica-tions of pride, embarrassment, possessiveness,or competition~Hobson, 1990; Kasari et al.,1993!, may already have evolved within olderhigh-functioning children with autism. Fur-thermore, the current sample’s showing behav-iors~e.g., “Mom, look!”! and direct and indirectspontaneous attempts to share attention withthe caregiver regarding a third object~e.g., thechild’s painting! call for further examinationof the possibility of later development of sec-ondary intersubjectivity and, by extrapola-tion, joint attention~the shared attention oftwo persons regarding a third event or object!in these children. Indeed, theDSM-IV-TR~American Psychiatric Association, 2000! con-siders the lack of a spontaneous search forshared experience to be a cardinal symptomof autism, but less is known about this behav-ior in older high-functioning children with au-tism ~Rogers & Bennetto, 2001!.

The different expressions of jealousy dem-onstrated by the children with autism in thepresent study as compared to their typicalage mates may stem from a deficit in under-standing socially accepted rules for emo-tional display, which children gain throughthe socialization process during typical devel-

opment. The considerably more explicit, ac-tive expressions of jealousy demonstratedby children with autism compared to theirtypical age mates in our study resembled theless mature, more explicit jealousy-provokedbehaviors reported as characterizing younger,typically developing children~e.g., Bers &Rodin, 1984; Masciuch & Kienapple, 1993!.For example, one child with autism took hisdrawing and pushed it in front of his moth-er’s face after she had praised the rival child.He repeated this behavior seven times, in-cluding one instance when he jumped overthe other child’s head in an attempt to reachhis mother’s face and show her his picture.The extent of active behaviors exhibited bythe children with autism was unparalleled inthe typical sample, whose more implicit ex-pressions of jealousy were generally limitedto gazing at the parent or rival child. In ad-dition, only the mental age of children withautism negatively correlated with the actionjealousy scale during the play situations, in-dicating that children with a higher mentalage exhibited fewer actions during this sce-nario. Although the current outcome regard-ing manifestations of jealousy in autismare possibly linked to deficits in emotionalunderstanding rather than to a qualitativedeviance explanation, this study calls for fu-ture research that directly compares childrenwith autism to young typical children. Fur-ther studies would also do well to investigatethe expression of other self-reflective emo-tions ~pride, embarrassment! in older high-functioning children with autism.

What implications arise from such findingsin terms of the affective versus cognitive in-terpretations of the emotional deficit in au-tism? In contrast with expectations, each ofthe two scenarios~drawing and playing! ren-dered the same jealousy-provoking effecton children with autism. Albeit there were dif-ferent roots for the two jealousy experiencesmanipulated in the present study~playing: af-fective and interpersonal; drawing: social–cognitive and comparison processes!, bothcomprised classic jealousy situations in whichthe child lost something~either attention orevaluation! from a valued person~caregiver!to a rival; hence, the loss experienced by the



child in the experimental group equaled therival’s gain. Thus, the children with autism inthe present study did seem to experience a self-reflective, socially mediated emotion such asjealousy. Although not enough is known aboutthe demands of the current jealousy tasks todraw clear-cut conclusions, expressing jeal-ousy in the play scenario could possibly implychildren’s sense of the intersubjectivity of thenetwork of relations within the triadic sce-nario~caregiver–rival–child in the experimen-tal group!. Expressing jealousy in the drawingscenario may possibly imply children’s capa-bilities to attribute their caregiver’s implicitmental states toward their own picture. How-ever, the lack of superiority for one theoreticalexplanation~affective vs. cognitive! over theother calls for an integrative developmentaltheoretical model that allows for the evolve-ment of higher social–emotional capabilitiesin high-functioning children with autism andalso accepts the existence of the expression ofself-reflective emotions and sharing capabili-ties. In addition, this model needs to take intoaccount the gap between these children’s abil-ity to experience an emotion that is sociallymediated and rooted in social interaction incomparison to their seriously lagging abilityto understand this emotion.

Indeed, the present study demonstrated amore severe deficit in the ability to con-sciously describe the experience and to under-stand jealousy than in the expression of theemotion. According to Saarni~1999!, the abil-ity to describe emotional experience requiresthe development of a network of concepts,which are scripts for representing children’sown emotional responses within a multidimen-sional matrix of causes, goals, values, socialrelations, and beliefs about emotion manage-ment. Typically developing children of aboutage 6–8 have well-defined scripts that revealsuch a multidimensional matrix. In the caseof autism, these scripts seem to fail to de-velop in the normative way. Not all types ofexperiences are difficult for these children toreflect upon. It appears that when the emo-tional experience requires a projection aboutthe self vis-à-vis the representation of socialrelations, like in social-relations jealousy, thechild with autism is less likely to succeed.

The children with autism in the present studywere less accurate in identifying jealousy froma social-relations picture compared with typ-ically developing children, and they providedexamples of social-relations jealousy less of-ten ~fewer than one-third of the autism groupsucceeded vs. more than two-thirds in the typ-ical group!. Such intergroup differences didnot emerge for social-comparison jealousy.Thus, although children with autism ex-pressed affective jealousy~in the playing sce-nario! as often as typical controls, they wereless capable of recognizing the emotion insuch a situation and less competent at identi-fying the conditions that reflect this type ofjealousy.

Similarly to these findings, the children withautism in Bauminger and Kasari’s~2000! studyreported greater loneliness compared with typ-ical controls, but they failed to include the moreaffective dimension of loneliness~being leftout of intimate close relationships! in their def-initions of the emotion. Is it possible that whendealing with the more affective root of an emo-tion that is linked to social relatedness~e.g.,emotional loneliness, affective-relational jeal-ousy!, high-functioning children with autismare able to experience the emotion but cannotdescribe this experience because of their af-fective deficit? Interestingly, Lee and Hobson~1998! reported that high-functioning individ-uals with autism were deficient in their abilityto describe issues related to their interper-sonal self; none of the participants in their studyprovided social self-statements that referredto friends or to being a member of a socialgroup. However, that study did not examinechildren’s actual social-interpersonal rela-tions, and it is possible that these children wereinvolved in interpersonal relationships with-out the ability to reflect on them. Other recentstudies have reported friendships in these high-functioning older children with autism~Baum-inger & Kasari, 2000; Bauminger & Shulman,2003!.

The social-comparison examples were rel-atively intact in children with autism. Theseexamples are based on the child’s sense of com-petition and0or feeling of lack of equality, andthey are rooted in a social–cognitive ratherthan affective process of comparison—the


172 N. Bauminger

child is projecting about the self through theaccomplishments or possessions of others. Fur-thermore, in contrast with the experimental sce-nario that elicited jealousy, in the examples ofsocial-comparison jealousy it is hard to dis-count the possibility that children provided ex-amples of envy~“I want for myself what theother child has.”!. Inasmuch as jealousy in-volves complex projections about the self vis-à-vis others whereas envy does not necessarily,the children with autism in the present studycould more easily provide less mature exam-ples of jealousy~or envy!, because of theirdifficulties in performing the more complexprojections about the self required for the re-flection of jealousy. They could more easilyfurnish examples stemming from their ownself-needs or their own appraisal that led todissatisfaction, rather than examples rooted ininterpersonal relationships and dealing withthe fear of losing these relationships.

Overall, what seems to remain distortedeven at older ages is the ability to reflect aboutemotional experience with others, a reflectionthat requires the consideration of interper-sonal relationships. One common explanationfor the superior social–emotional functioningof high versus low-functioning children withautism comprises the cognitive compensationor logicoaffective hypothesis~Capps et al.,1992; Hermelin & O’Connor, 1985; Kasariet al., 2001; Yirmiya et al., 1992!. This hypoth-esis suggests that children with autism learnstrategies to recognize emotions that “comenaturally” to individuals with typical develop-ment. The major dilemma of this hypothesisconcerns the boundaries of this strategy in fullycompensating for the affective deficit in au-tism ~Kasari et al., 2001!. In the present study,only for the autism sample, children’s mentalages correlated positively with the ability toprovide social-comparison examples of jeal-ousy but not with the ability to provide social-relations examples of jealousy. Thus, cognitivecapabilities seemed less helpful in producingexamples that are rooted in interpersonal rela-tionships~e.g., social-relations jealousy!. More-over, future studies utilizing social–cognitiveanalogue tasks such as recognition of affec-tive jealousy in a picture should take into ac-count the possibility that such tasks contain

affective elements that may be difficult for chil-dren with autism to reflect upon.

In terms of coping with jealousy, the chil-dren’s ability to provide cognitive–mentalisticsuggestions was positively linked with mentalage for both samples; however, overall, chil-dren with autism provided a lower number ofsolutions for coping with jealousy in compar-ison to their typical age mates. This findingemphasizes the difficulties in emotional un-derstanding in the autism sample; they haveless knowledge of how to deal with an unpleas-ant feeling such as jealousy.

The following study limitations should benoted. First, because of the fact that only high-functioning older children participated in thestudy, questions still remain regarding the chro-nology of the development of jealousy in au-tism and its universality to the disorder. Futurestudies may examine whether low-functioningchildren with autism experience and under-stand jealousy in a similar manner and mayattempt to identify the onset of jealousy amonglow- and high-functioning children with au-tism. Along these lines, future research maydo well to include participants with a nar-rower IQ range and to utilize a one on onematching procedure. Second, the peer0siblingparticipation factor should be mentioned. Be-cause of the fact that the manipulation of jeal-ousy scenarios largely depended on the parent’sability to perform the scenarios accurately, itwas especially important that parents felt com-fortable. Therefore, parents were given the op-tion of performing the scenario with a peer orwith a sibling, in light of past studies’ reportsthat both siblings and peers provoke jealousyin similar situations~Masciuch & Kienapple,1993; Miller et al., 2000!. Indeed, a similarpercentage of peers and siblings was selectedby parents in the two groups~autism and typ-ical!, and nonsignificant differences emergedbetween scenarios that were implemented withsiblings versus those with peers. However, apossibility remains that because the study wasnot limited to only siblings or to only familiarpeers for all of the scenarios, the results mayhave been influenced in some way.

Third, another limitation regards the exper-imental manipulation of jealousy. The presentstudy followed the same paradigm that suc-



cessfully provoked jealousy in typically de-veloping children, altering none of itscomponents. However, in order to claim un-equivocally that the presence of the other peeror the mother was necessary to provoke jeal-ousy, future research should compare theseresults with nontriadic scenarios comprisingonly the mother and child while the mothersolely praises her own drawing and0or withtwo triadic scenarios~one including the motherand the other including an unfamiliar adult!.Previous research on prosocial behaviors re-vealed that children with autism were rela-tively indifferent to mothers’ behaviors~e.g.,pretending to be sick; Sigman, Kasari, Kwon,& Yirmiya, 1990!, but researchers have nottested this issue for jealousy. In addition, tobetter tease out the differentiation of jealousyfrom envy, another experimental conditioncould include only two peers, for example,with one receiving what the other one de-sires. Fourth, because the present study em-ployed only a single picture to tap affectivejealousy, future studies would do well to in-clude control pictures~e.g., including otheremotions! in order to claim specificity for the

difficulties these children exhibit in under-standing affective jealousy.

In conclusion, high-functioning childrenwith autism manifested jealousy in similar sit-uations as did their typically developing coun-terparts; yet, their understanding of the feelingwas less coherent compared with their typi-cal age mates. These outcomes call for adevelopmental–integral model to explicate theaffective deficit of these children and takeinto consideration the better social–emotionalperformance of individuals with autism whohave higher cognitive capabilities. In partic-ular, the major contribution of the presentstudy revealed a gap between a more intactcapability to experience jealousy~a self-reflective, socially mediated emotion! and adeficit in the capacity to fully reflect on theexperience of such an emotion. This gap be-tween expression and understanding of themore affective–interpersonal roots of emo-tions should be studied further. Along theselines, the current study’s outcomes empha-size the need to promote emotional under-standing capabilities in high-functioningchildren with autism.

References

American Psychiatric Association.~1994!. Diagnostic andstatistical manual of mental disorders~4th ed., Re-vised!. Washington, DC: Author.

American Psychiatric Association.~2000!. Diagnostic andstatistical manual of mental disorders: Text revision~4th ed., Revised!. Washington, DC: Author.

Arnold, M. B. ~1960!. Emotion and personality. New York:Columbia University Press.

Bacon, A. L., Fein, D., Morris, R., Waterhouse, L., &Allen, D. ~1998!. The responses of autistic children tothe distress of others.Journal of Autism and Develop-mental Disorders, 28, 129–142.

Baron–Cohen, S., & Swettenham, J.~1997!. Theoryof mind in autism: Its relations to executive func-tion and central coherence. In D. J. Cohen & F. R.Volkmar ~Eds.!, Handbook of autism and devel-opmental disorders~pp. 880–893!. New York:Wiley.

Bauminger, N., & Kasari, C.~2000!. Loneliness and friend-ship in high-functioning children with autism.ChildDevelopment, 71, 447–456.

Bauminger, N., & Shulman, C.~2003!. The developmentand maintenance of friendship in high-functioning chil-dren with autism: Maternal perceptions.Autism, 7,81–97.

Bers, S. A., & Rodin, J.~1984!. Social comparison jeal-ousy: A developmental and motivational study.Jour-nal of Personality and Social Psychology, 47,766–779.

Blumberg, S., & Izard, C.~1991!. Patterns of emotionexperiences as predictors of facial expressions of emo-tion. Merrill–Palmer Quarterly, 37, 97–183.

Capps, L., Yirmiya, N., & Sigman, M.~1992!. Under-standing of simple and complex emotions in non-retarded children with autism.Journal of ChildPsychology and Psychiatry and Allied Disciplines,33, 1169–1182.

Dawson, G., & McKissick, F. C.~1984!. Self-recognitionin autistic children.Journal of Autism and Develop-mental Disorders, 14, 383–394.

Dennis, M., Lockyer, L., & Lazenby, A. L.~2000!. Howhigh-functioning children with autism understand realand deceptive emotion.Autism, 4, 370–381.

Dissanayake, C., & Sigman, M.~2001!. Attachment andemotional responsiveness in children with autism.In-ternational Review of Research in Mental Retarda-tion, 23, 239–266.

Hansen, G. L.~1991!. Jealousy: Its conceptualization, mea-surement, and integration with family stress theory. InP. Salovey~Ed.!, The psychology of jealousy and envy~pp. 211–230!. New York: Guilford Press.

Happé, F. G. E.~1994!. Annotation: Current psychologi-cal theories of autism: The “theory of mind” accountand rival theories.Journal of Child Psychology andPsychiatry, 35, 215–229.

Happé, F. G. E.~1995!. The role of age and verbal abilityin the theory of mind task performance of subjectswith autism.Child Development, 66, 843–855.


174 N. Bauminger

Harris, P. L.~1989!. Children and emotion: The develop-ment of psychological understanding. Oxford: BasilBlackwell.

Hermelin, B., & O’Connor, N.~1985!. The logico-affectivedisorder in autism. In E. Schopler & G. B. Mesibov~Eds.!, Communication problems in autism~pp. 283–310!. New York: Plenum Press.

Hobson, R. P.~1990!. On the origins of self and the caseof autism. Development and Psychopathology, 2,163–181.

Hobson, R. P.~1993a!. Autism and the development ofmind. Hove, UK: Erlbaum.

Hobson, R. P.~1993b!. The emotional origins of interper-sonal understanding.Philosophical Psychology, 6,227–249.

Hupka, R. B.~1984!. Jealousy: Compound emotion orlabel for a particular situation.Motivation and Emo-tion, 8, 141–155.

Izard, C. E.~1991!. The psychology of emotions. NewYork: Plenum Press.

Kasari, C., Chamberlain, B., & Bauminger, N.~2001!.Social emotions and social relationships in autism:Can children with autism compensate? In J. Burack,T. Charman, N. Yirmiya, & P. Zelazo~Eds.!, Devel-opment and autism: Perspectives from theory and re-search~pp. 309–323! Hillsdale, NJ: Erlbaum.

Kasari, C., Sigman, M., Baumgartner, P., & Stipek, D.~1993!. Pride and mastery in children with autism.Journal of Child Psychology and Psychiatry, 34,353–362.

Kasari, C., Sigman, M., Yirmiya, N., & Mundy, P.~1993!.Affective development and communication in chil-dren with autism. In A. P. Kaiser & D. B. Gray~Eds.!,Enhancing children’s communication: Research foun-dations for intervention~pp. 201–222!. New York:Brookes.

Lee, A., & Hobson, R. P.~1998!. On developing self-concepts: A controlled study of children and adoles-cents with autism.Journal of Child Psychology andPsychiatry, 39, 1131–1144.

Lewis, M. ~1993!. The emergence of human emotions. InM. Lewis & J. M. Haviland~Eds.!, Handbook of emo-tions ~pp. 223–235!. New York: Guilford Press.

Lord, C., Rutter, M., & LeCouteur, A.~1994!. AutismDiagnostic Interview—Revised: A revised version ofa diagnostic interview for caregivers of individualswith possible pervasive developmental disorders.Jour-nal of Autism and Developmental Disorders, 19,185–212.

Masciuch, S., & Kienapple, K.~1993!. The emergence ofjealousy in children 4 months to 7 years of age.Jour-nal of Social and Personal Relationships, 10, 421–435.

Miller, A. L., Volling, B. L., & McElwain, N. L. ~2000!.Sibling jealousy in a triadic context with mothers andfathers.Social Development, 9, 433–457.

Neisser, U.~1988!. Five kinds of self-knowledge.Philo-sophical Psychology, 1, 35–59.

Neu, J.~1980!. Jealous thoughts. In A. O. Rorty~Ed.!,Explaining emotions~pp. 425–463!. Berkeley, CA:University of California Press.

Parrott, W. G.~1991!. The emotional experience of envyand jealousy. In P. Salovey~Ed.!, The psychology ofjealousy and envy~pp. 3–31!. New York: GuilfordPress.

Parrott, W. G., & Smith, R. H.~1993!. Distinguishing theexperiences of envy and jealousy.Journal of Person-ality and Social Psychology, 64, 906–920.

Rogers, S. R., & Bennetto, L.~2001!. Intersubjectivity inautism: The role of imitation and executive function.In A. M. Wetherby & B. M. Prizant~Eds.!, Autismspectrum disorders: A transactional developmental per-spective~Vol. 9, pp. 79–107!. Baltimore, MD: Paul H.Brooks.

Rogers, S. R., & Pennington, B. F.~1991!. A theoreticalapproach to the deficits in infantile autism.Develop-ment and Psychopathology, 3, 137–162.

Saarni, C.~1999!. The development of emotional compe-tence. New York: Guilford Press.

Salovey, P., & Rodin, J.~1984!. Some antecedents andconsequences of social-comparison jealousy.Journalof Personality and Social Psychology, 4, 780–792.

Salovey, P., & Rothman, A.~1991!. Envy and jealousy:Self and society. In P. Salovey~Ed.!, The psychologyof jealousy and envy~pp. 271–286!. New York: Guil-ford Press.

Seidner, L. B., Stipek, D. J., & Fesbach, N. D.~1988!. Adevelopmental analysis of elementary school-aged chil-dren’s concept of pride and embarrassment.Child De-velopment, 59, 367–377.

Silver, M., & Sabini, J.~1978!. The perception of envy.Social Psychology, 41, 105–117.

Sigman, M., Kasari, C., Kwon, J., & Yirmiya, N.~1990!.Responses to negative emotion of others by autistic,mentally retarded, and normal children.Child Devel-opment, 63, 796–807.

Spiker, D., & Ricks, M.~1984!. Visual self-recognition inautistic children: Developmental relationships.ChildDevelopment, 55, 214–225.

Tager–Flusberg, H.~2001!. A reexamination of the theoryof mind hypothsis of autism. In J. Burack, T. Char-man, N. Yirmiya, & P. Zelazo~Eds.!, Developmentand autism: Perspectives from theory and research~pp. 173–193!. Hillsdale, NJ: Erlbaum.

Tov–Ruach, L.~1980!. Jealousy, attention, and loss. In A.O. Rorty ~Ed.!, Explaining emotions~pp. 465–488!.Berkeley, CA: University of California Press.

Travis, L. L., & Sigman, M.~1998!. Social deficit andinterpersonal relationship in autism.Mental Retarda-tion and Developmental Disabilities, 4, 65–72.

Volling, B. L., McElwain, N. L., & Miller, A. ~2002!.Emotion regulation in context: The jealousy complexbetween young siblings and its relations with childand family characteristics.Child Development, 73,581–600.

Wechsler, D.~1974!. WISC-R manual: Wechsler Intelli-gence Scale for Children—Revised. San Antonio, TX:Psychological Corporation.

Yirmiya, N., Sigman, M., Kasari, C., & Mundy, P.~1992!.Empathy and cognition in high-functioning childrenwith autism.Child Development, 63, 150–165.



Appendix A

Examples of children’s manifestations of jealousy on the explicitness scale for the drawing and playing scenarios

Autism Typical DevelopmentLevel of Explicitness

of Jealousy Draw Play Draw Play

Level 4Behaviors or verbalizations

that indirectly interveneinto the parent—rival childinteraction

Stands in close proximity tomother and caresses her hair

Makes loud comment “I am thechamp,” with eye gaze towardthe father

“Mom, I love you . . .”

Makes repeated loud commentsabout own game such as“Wow, what a beautiful slide. . .”

Starts again after stopping ownwork and says “Just a minute,I did not finish yet, Mom . . .@plus eye gaze#

Stops own work, looks at momand other child, grabs colorsand stamps from the otherchild’s color box, and usesthem in her picture

Stops own game, grabs objectsfrom the other child’sconstruction game, and startsto build own model

Level 5Direct behaviors0

verbalizations aimed atfocusing parent’s attentionto the drawing0playing ofthe child in theexperimental group

“Now look at my picture,” whilepushing picture towardmother’s face several times

“Look at this.”

“See mine.”

“Look what I did . . .”

“Mom, look,” while pushinggame toward the mother

Says “Mom, look” and verygently places own picturecloser to mom’s sight

“Yeah! This is something Icould never do before withmy own game . . . Wow!Mom, look . . .”

“Mom, look . . .”

Level 6Direct declaration of

comparison and lack ofequality, and0or anynegative affect

“I did not think of that idea, sonow I will make the samepicture as his, so you@mother# will say mine ispretty too”

“I put a lot of effort intodrawing this, so why aren’tyou saying anything about mypicture?”~whining!

“But I drew more . . .”

“Mom, why don’t you play withme too?”

“Mom . . . there is somethingthat is really making meangry: Why do you only helpher? . . . Iwon’t be yourfriend if you keep playingonly with her” ~whining!

“Mom, see what I am doing.Why don’t you ask me what Iam doing? We used to playthis together”

“Mine is prettier. Here, look.”

“Is mine pretty too? Do youknow what I am painting? Iam painting a doubledeckerbus.”

“Mom @in sad tone of voice# ,why can’t I also play withboth of you?”

“. . . And what about me? Am Ialone?” with facial expressionof frustration

||

DP

P1

6~1!4

45

2002

10

201300

41

1:4

0a

mR

EV

ISE

DP

RO

OF

17

6

Appendix B

Children’s examples for the experience of jealousy

Type of Example Autism Typical Development

Social–relations jealousy • When someone walks with his girlfriend, his other friendsfeel jealous. They also want to be friends with the kid whois my friend.

• When my best friend gets more attention, then I don’t.• When a kid from class is going to play with another kid

from class, and not with me, I feel sad.

• I was jealous when my little sister was born and everyonepaid attention only to her.

• When my father hugged my brother, mom was not home,and I also wanted my dad.

• When my friend meets another friend of hers, I feel jealous,because I feel left alone.

• When you are insulted by a friend and your other friendsare on the side of your rival

Social–cognitive jealousy • When somebody gets something and the other one does notget anything, then it is possible to be very jealous.

• When kids in school can buy whatever they want wheneverthey want

• I feel jealous of my brother. He has more fun than I do, hegoes to parties, and I help with the housework, like I do thelaundry.

• I’m jealous of my friend because she’s prettier than me.• When someone’s good at sports and I’m not• In school when someone gets good grades and I don’t

Appendix C

Examples of children’s suggested solutions for coping with jealousy

Type of Solution Autism Typical Development

Situation activity To ask for the desired game as a present for PassoverTo go and play with another childTo join the child who has what I want

To save money and buy the desired objectTo ask your friend if you can try his bicycleTo go see the movie at a different time

Cognitive mentalistic Get out for a walk and wait for the feeling to passIgnore the eventSay to yourself that you don’t want the other thingGet out of that place so you won’t see it or have to think

about it

Use self-talk to encourage yourself, such as “I’m also pretty”Think about other things and forget about itConvince yourself that it is not so important for youThink about the other person’s disadvantages and remember the

things that you like in yourself

||

DP

P1

6~1!4

45

2102

10

201300

41

1:4

0a

mR

EV

ISE

DP

RO

OF

17

7

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

Documents

The expression and understanding of jealousy in children