Literature review paper

Physcialabuse_congitive_emotionalprocesses_behaviorproblems.pdf

Home >Psychology homework help >Literature review paper

Articles

Physical Abuse, Cognitive and Emotional Processes, and Aggressive/Disruptive Behavior Problems Michael Teisl, Mt. Hope Family Center, University of Rochester, Dante Cicchetti, Institute of Child Development, University of Minnesota

Abstract

Cognitive and emotional processes were examined in maltreated children with a history of physical abuse (n = 76), children with a history of maltreatment other than physical abuse (i.e., sexual abuse, physical neglect, and emotional maltreatment; n = 91), and a group of non-maltreated comparison children (N = 100). Physical abuse was associated with errors in cue interpretation, easy access to aggressive responses to conflict, and poor emotion regulation. In turn, maladaptive cognitive and emotional processes made unique contributions to explaining the relationship between physical abuse and peer nominations of aggression and disruptive behavior. Maltreated chil- dren who had not been physically abused evidenced a trend toward cognitive and emotional maladaptation, but only poor emotion regulation accounted for aggression and disruptive behavior in these children. The findings emphasize the importance of considering multiple domains of development in efforts to understand how varying forms of child maltreatment contribute to later maladjustment.

Keywords: child maltreatment; aggression; social information processing; emotion regulation

Introduction

The development of aggressive behavior places children at serious risk for future maladaptation in the form of continued aggression, conduct problems, and overall difficulties in the domain of social relationships. In fact, research has noted that early forms of aggression remain stable over time (Huesmann, Eron, Lefkowitz, & Walder, 1984; Olweus, 1979) and that childhood aggression is a consistent predictor of externalizing behavior problems during adolescence as well as antisocial behavior in adulthood (Coie, Lochman, Terry, & Hyman, 1992; Farrington, 1995). Particularly disconcerting is Moffitt’s (1993) account of the life-course-persistent antisocial type of youth. This perspective identifies individuals who evidence an aggressive and antiso- cial behavioral style during childhood as being at even greater risk for continued

Correspondence should be addressed to Michael Teisl, Mt. Hope Family Center, University of Rochester, 187 Edinburgh St., Rochester, NY 14608. Email: [email protected]

doi: 10.1111/j.1467-9507.2007.00412.x

mailto:[email protected]

aggression and criminality in adulthood than are individuals who first display antisocial behavior as adolescents (Moffitt & Caspi, 2001; Moffitt, Caspi, Harrington & Milne, 2002). Given that aggression and other forms of externalizing behavior problems place children at increased risk for a future marked by hostility and other antisocial behavior, it is essential that researchers achieve a greater understanding of the mechanisms guiding the early development of these behavioral problems.

Empirical evidence linking the experience of child maltreatment with aggression and disruptive behavior problems indicates the important impact that the study of maltreatment may have on understanding the development of these behavioral out- comes in children. One of the most consistent findings in maltreatment literature is that children who have been maltreated are more likely to exhibit atypical behavior in the form of higher levels of externalizing and/or internalizing symptomatology than are non-maltreated children from comparable socioeconomic backgrounds (Bolger & Patterson, 2001; Manly, Kim, Rogosch, & Cicchetti, 2001). Physical abuse, in particu- lar, has been found to place children at increased risk for exhibiting aggressive or delinquent behavior (Cicchetti & Lynch, 1995; Smith & Thornberry, 1995; Trickett & McBride-Chang, 1995). Considering the work of Moffitt and others, maltreated chil- dren may thus be at risk for progressing down a path toward increasing maladjustment, marked by limited social competence with the potential for continued aggression and antisocial behavior in the future as well as an ultimate risk for the development of psychopathology in adulthood.

Although there is strong empirical evidence that maltreated children, particularly those who were physically abused, are more likely to exhibit aggressive and other externalizing behavior, less is known about the underlying processes governing this relationship. Consistent with a developmental psychopathology perspective, identify- ing these processes may help to explain the development of aggressive and disruptive behavior in maltreated children, as well as to elucidate factors important in the development of problematic behavior in other normative or at-risk groups (Cicchetti, 1993; Sroufe, 1990). Furthermore, the study of child maltreatment may provide some insight into how deviations from the average expectable environment can result in subsequent maladaptation and failure to successfully negotiate certain stage-salient tasks throughout development, such as establishing successful peer relationships (Cicchetti & Lynch, 1995).

Social Information Processing

Considerable research on the development of aggression and other socially incompe- tent behavior has focused on the role of cognitive processing during social situations. This work has suggested that aggressive behavioral development is associated with distorted perceptions regarding the social environment. Among the most widely studied models in this area of inquiry is the social information-processing model (Crick & Dodge, 1994; Dodge, 1986), which suggests that, when presented with social cues, children engage in six stages of cognitive processing. These stages include the encoding of internal and external cues, the interpretation of cues, clarification of goals, response access or construction, response decision, and enactment. Distortions in these various stages of processing have been noted to have a negative impact on children’s social adjustment, placing them at heightened risk for aggressive interactions with peers (Dodge, Pettit, McClaskey, & Brown, 1986; Slaby & Guerra, 1988). The

2 Michael Teisl and Dante Cicchetti

interpretation stage of the social information-processing model, in particular, has attracted considerable attention from researchers interested in aggression and peer relations. Research in this area has found that aggressive children are more likely than others to interpret both ambiguous and benign cues as hostile, and that attributional biases increase the likelihood that a child will behave aggressively (Dodge, 1980; Nasby, Hayden, & DePaulo, 1980).

Additional research on cognitive functioning has shown that maltreated children display deviations in processing social information similar to the deviations evident in aggressive children. For example, Pollak, Cicchetti, Hornung, and Reed (2000), and Rieder and Cicchetti (1989) found that, compared to non-maltreated peers, mal- treated children appeared to be hypervigilant to threat-related cues in the environment. Pollak and colleagues also noted that maltreated children evidence a bias toward misperceiving anger when processing facial expressions, and they further discovered that physically abused children were at greatest risk for exhibiting this bias (Pollak & Sinha, 2002; Pollak & Tolley-Schell, 2003; Pollak et al., 2000).

Other work also has indicated that physically abused children evidence disruptions in the ability to effectively process social information. In fact, physically abused children have been noted to make errors encoding social cues, exhibit hostile attribu- tional biases, access more aggressive behavioral responses, and view aggression more favorably than children who have not been harmed (Dodge, Bates, & Pettit, 1990; Dodge, Pettit, Bates, & Valente, 1995). Moreover, the results of these investigations suggest that these patterns of processing social information, in part, account for the relationship between physical maltreatment and the development of aggression and other behavior problems.

The hypothesis that cognitive processes mediate the link between maltreatment and children’s aggressive behavior was further evaluated in the current study. Based on the evidence that abused children are at risk for developing biases toward perceiving threats from others, we focused our investigation on the interpretation stage of the social information-processing model (Crick & Dodge, 1994; Dodge, 1986). Also, given their likely exposure to models of aggressive behavior, we expected that physi- cally abused children, in particular, would have easy access to a mental cache of aggressive responses to situations of conflict. Thus, we included this pattern of pro- cessing in our study as well.

Prior investigations of maltreatment, social cognition, and aggression have assessed children’s histories of maltreatment or harsh discipline through parent self- report (Dodge et al., 1990; Dodge et al., 1995). The present study extends these prior investigations by focusing on cases of maltreatment that have been identified through the coding of social services records and interviews with social services workers. This method of identifying maltreatment has advantages over self-report or parent- report measures in that the effects of social desirability and denial are reduced, resulting in more accurate detection of maltreatment occurrences. Furthermore, an extensive review of case histories allows for a detailed examination of additional dimensions of maltreatment, such as the presence or absence of specific subtypes (see Barnett, Manly, & Cicchetti, 1993). This is not meant to understate the impor- tance of investigations of maltreatment using self-report assessments, as this approach has the advantage of identifying cases that have escaped the attention of social service agencies. However, it is important to note that the cases included in the current study represent a distinct subset of maltreating families, given their involvement with social service agencies.

Abuse and Aggressive/Disruptive Behavior 3

Another distinctive aspect of the present investigation is that we examined whether or not physical abuse has a unique link to aggression over and above that of other forms of maltreatment. Whereas previous studies focused solely on the impact of harsh parenting or physical discipline (Dodge et al., 1990; Dodge et al., 1995), this investigation included physical neglect, emotional maltreatment, and sexual abuse, in addition to physical abuse, in evaluating the associations among child maltreatment, social cognition, and aggression. An investigation including other forms of maltreat- ment will allow us to discern whether or not there are elements specific to the experience of physical abuse that lead to the development of aggressive behavioral problems or whether or not these risk factors are consistent across subtypes of maltreatment.

Emotion Regulation

Despite their importance, it is doubtful that cognitive processes are the only factors influencing the development of externalizing behavior problems. Instead, it is likely that emotion processes also play an integral role, as these behaviors are often as- sociated with feelings of anger (Cole & Zahn-Waxler, 1992). Although there is an association between experiences of negative affect, such as anger, and the manifesta- tion of aggression and externalizing behavior, individuals who are highly reactive to anger-provoking situations do not necessarily act on those angry feelings. Instead, the ability, or inability, to effectively regulate one’s own emotional experience seems most important in determining the manner in which those feelings are ultimately expressed (Eisenberg, Fabes, Guthrie, Murphy, Maszk, & Holmgren, 1996; Fabes & Eisenberg, 1992).

Emotion regulation is a developmental construct widely considered to have its roots early in childhood, where interactions between children and their caregivers provide opportunities to learn about emotion and its management (Cole, Martin, & Dennis, 2004). In light of the important role that the parent–child relationship plays in the development of emotion regulation, it stands to reason that hostile, neglectful, or inconsistent parenting, typically seen in maltreating families (Rogosch, Cicchetti, Shields, & Toth, 1995), could result in the development of aberrant patterns of emotion regulation. Empirical research on child maltreatment has, in fact, noted that maltreated children are more likely to evidence problems with emotion regulation (Maughan & Cicchetti, 2002; Shields & Cicchetti, 1998), and has indicated that this dysregulation may be the result of inadequate instruction or modeling by parents (Shipman & Zeman, 2001).

Emotion dysregulation, in turn, has been identified for its potential in explaining the maladjustment noted in maltreated children. The findings of Shields, Cicchetti, and Ryan (1994), for example, indicate that emotion dysregulation may at least partially account for the relationship between child maltreatment and attenuated social compe- tence. Other research has also noted that the representations maltreated children form regarding themselves and their caregivers place them at risk for dysregulated emotion, as well as concomitant difficulties in their relationships with peers (Shields, Ryan, & Cicchetti, 2001). Given that maltreated children are at risk for difficulties with emotion regulation, in addition to the evidence linking emotional regulation to children’s social functioning, we hypothesize that emotion regulation will also mediate the relationship between experiences of maltreatment and maladaptive behavior exhibited in the peer arena.

4 Michael Teisl and Dante Cicchetti

Integration of Cognitive and Emotional Processes

Although research on social cognition provides clear evidence of a relationship between errors in cognition and later social dysfunction, it also has been suggested that affective processes play a significant role through their influence on social interactions (Calkins, Gill, Johnson, & Smith, 1999; Thompson, 1994). Researchers in the area of social cognition have recognized the important influence that emotion has on social competence and the exhibition of behavioral problems (Crick & Dodge, 1994; Dodge, 1991; Lemerise & Arsenio, 2000). Despite this recognition, there remains a relative paucity of work examining the relative effects of cognitive and emotional processes in predicting the development of aggression and disruptive behavioral problems. This is particularly true with respect to investigations of the mechanisms underlying the long-term impact of child maltreatment.

Research that has attempted to account for the importance of both emotion and cognition in understanding the development of problematic behavior with peers has focused on the experience of negative affect (Dodge & Somberg, 1987; Graham, Hudley, & Williams, 1992; Orobio de Castro, Slot, Bosch, Koops, & Veerman, 2003). The present study builds upon this existing literature by examining the broader process of emotion regulation, as opposed to the experience of negative affect alone, in addition to social information processing. We believe that the joint investigation of these cognitive and affective processes in children with varying maltreatment experi- ences will advance the understanding of how aversive child rearing promotes the development of aggressive behavior. Our hypotheses in this study were as follows:

(1) Maltreated children would receive more nominations of aggressive and disruptive behavior from peers than would non-maltreated children. This was expected to be particularly true for physically abused children.

(2) Maltreated children would display more social information processing errors typical of aggressive children than would non-maltreated children. Aggressive processing errors were expected to be primarily evident in physically abused children.

(3) In accordance with prior research on maltreatment, maltreated children were expected to receive lower ratings of emotion regulation than their non-maltreated peers. Physically abused children were not necessarily expected to receive lower ratings of emotion regulation relative to other maltreated children.

(4) Aggressive patterns of processing social information and poor emotion regulation would independently mediate the relationship between maltreatment experiences and peers’ nominations of aggression and disruptiveness. Aggressive processing patterns were expected to be primarily operative in explaining aggressive and disruptive behavior for physically abused children.

Method

Participants

Participants for this study were recruited in two cohorts, each hailing from two different cities in the northeastern USA (Cohort 1, N = 121; Cohort 2, N = 146). Both cohorts were drawn from urban communities, and each consisted of a group of children with documented cases of child maltreatment, as well as a group of non-maltreated comparison children. Maltreated children were recruited through the Massachusetts

Abuse and Aggressive/Disruptive Behavior 5

Department of Social Services, for Cohort 1, and through the Monroe County, NY Department of Social Services, for Cohort 2. Non-maltreated comparison children in each cohort were recruited through fliers posted in low-income urban neighborhoods comparable to the neighborhoods where children in the maltreatment groups resided. The maltreatment groups were representative of families receiving services through both the Massachusetts and Monroe County Department of Social Services (DSS) (Cicchetti & Manly, 1990). All participating families signed release forms granting access to their DSS records. The presence or absence of maltreatment was subse- quently verified through interviews with social workers and extensive examinations of each family’s case history.

Maltreatment and non-maltreatment groups in each cohort were compared on a number of demographic variables (see Table 1). Given the demographic equivalence of these groups across cohorts, the samples were combined in order to increase the power of analyses conducted in this investigation, particularly for analysis of maltreatment subtype effects. Children included in the study ranged in age from six to 12 years (M = 8.04, SD = 1.63), 57.7 percent of whom were male. Children’s race/ethnicity was converted into a dichotomous variable representing minority (i.e., African-American, Hispanic, Other) and non-minority (i.e., White) group membership, with 70.8 percent of participants classified as non-minority. Given the evidence associating the experi- ence of physical abuse, in particular, with aggressive social information-processing styles, as well as heightened aggression and externalizing behavior, we expected to see the most severe disruptions in social, emotional, and cognitive functioning in this subset of maltreated children. Consequently, children were split into three groups consisting of physically abused children (N = 76), children only experiencing other forms of maltreatment (i.e., sexual abuse, emotional maltreatment, and physical neglect; N = 91), and non-maltreated comparison children (N = 100). Groups were found comparable with respect to age and gender, but significant differences arose when considering race/ethnicity, with the other maltreatment group evidencing more minority subjects than either of the other two groups (see Table 2). There were also significant differences in terms of the number of children in the family, although the groups were comparable on a number of other family demographic features. Because the number of children in the family did not correlate with any variables of interest in the study, however, it was not considered further.

Maltreatment Classification

Social workers assigned to families in the maltreatment group in Cohort 1 were interviewed by trained research associates using a modified version of the 87-item checklist of maltreatment events developed by Giovannoni and Becerra (1979). This checklist provides information regarding the specific subtypes of maltreatment that children experienced. The social workers had worked closely with the families and reported on the nature of maltreatment based upon intimate knowledge of each family’s history and case record. The maltreatment classification system developed by Barnett et al. (1993) was used to code the Monroe County DSS records of families in Cohort 2. Trained doctoral students and clinical psychologists carefully reviewed each family’s DSS case history in order to determine the parameters of maltreatment that had occurred.

Barnett et al. (1993) derived definitions of maltreatment subtypes based largely on the work conducted by Giovannoni and Becerra (1979). Thus, the criteria used by

6 Michael Teisl and Dante Cicchetti

T ab

le 1.

D em

og ra

p h

ic C

om p

ar is

on of

M al

tr ea

tm en

t an

d N

on -m

al tr

ea tm

en t

G ro

u p

s in

C oh

or t

1 an

d C

oh or

t 2

M al

tr ea

tm en

t gr

ou ps

N on

-m al

tr ea

tm en

t gr

ou ps

C oh

or t

1 M

(S D

) C

oh or

t 2

M (S

D )

t C

oh or

t 1

M (S

D )

C oh

or t

2 M

(S D

) t

C hi

ld ’s

ag e

8. 23

(1 .7

9) 8.

74 (1

.4 9)

-1 .9

8* 8.

14 (1

.7 7)

8. 29

(1 .5

4) -.

43 C

hi ld

re n

in th

e fa

m il

y 3.

08 (1

.1 1)

3. 04

(1 .5

4) .2

0 2.

54 (1

.0 6)

2. 59

(1 .0

7) -.

22 M

ot he

r’ s

ed uc

at io

n 10

.9 7

(1 .9

3) 10

.8 9

(1 .5

1) .2

9 10

.9 3

(1 .6

2) 11

.3 8

(1 .3

7) -1

.4 4

C oh

or t

1 (%

) C

oh or

t 2

(% )

c2 %

C oh

or t

1 %

C oh

or t

2 c2

C hi

ld ’s

ge nd

er (m

al e)

50 .0

57 .9

.9 8

57 .4

59 .2

1. 42

C hi

ld ’s

ra ce

(n on

-m in

or it

y) 88

.3 56

.1 18

.2 8*

** 91

.8 51

.3 21

.4 2*

** S

in gl

e pa

re nt

fa m

il y

48 .3

54 .2

.5 3

57 .4

46 .2

1. 20

H is

to ry

of pu

bl ic

as si

st an

ce 95

.0 96

.3 .1

5 98

.4 92

.3 2.

27 <$

10 00

0 to

ta l

in co

m e

71 .7

80 .4

1. 66

83 .6

71 .8

2. 00

* p

< .0

5, **

* p

< .0

01 .

Abuse and Aggressive/Disruptive Behavior 7

each system to operationalize subtypes of maltreatment were comparable. In each cohort, physical abuse was rated as present if a child sustained injury from a caregiver through non-accidental means. Physical neglect was indicated if caregivers failed to provide for medical, nutritional, cleanliness, educational needs, or adequate supervision. Sexual abuse was designated if there was evidence of inappropriate sexual contact. Emotional maltreatment was identified if a child’s basic emotional needs (i.e., security, positive regard, autonomy) had been thwarted. Prior research has indicated that the identification of pure maltreatment subtypes is far less common than cases of maltreatment with co-morbid subtypes (Cicchetti & Barnett, 1991; Cicchetti & Rizley, 1981; Manly, Kim, Rogosch, & Cicchetti, 2001). Consistent with the previous work on maltreatment, the majority of maltreated subjects (60.7 percent) in the current study were found to have experienced multiple forms of maltreatment, with 45 percent experiencing physical abuse, 81.5 percent experienc- ing physical neglect, 56 percent experiencing emotional maltreatment, and 7.7 percent experiencing sexual abuse.

Although prior investigations have considered the role of physical maltreatment in the development of aggressive behavior, the present investigation was most interested in the relative impact of physical abuse. Given the high degree of subtype overlap commonly found in maltreated samples, any effort to select out pure maltreatment subtypes would result in an unnatural representation of maltreated children, not to mention a considerably smaller sample. Instead, maltreated subjects were categorized into two groups. One group represented children with a history of any form of maltreatment other than physical abuse, including 86.8 percent who had experienced physical neglect, 42.9 percent who had experienced emotional maltreatment, and 8.8 percent who had experienced sexual abuse. The other group represented children who had been physically abused in addition to any other maltreatment experience. Of those in the physical abuse group, 75 percent also experienced physical neglect, 72.4 percent

Table 2. Demographic Comparison of Maltreatment Status Groups

NM M (SD)

PA M (SD)

OM M (SD) F

Child’s age 8.20 (1.68) 8.50 (1.65) 8.61 (1.60) 1.60 Children in the family 2.56 (1.06) 2.97 (1.22) 3.12 (1.53) 4.91** Mother’s education 11.11 (1.54) 10.88 (1.52) 10.95 (1.79) .48

NM(%) PA(%) OM(%) c2

Child’s gender (male) 62.0 53.9 56.0 1.29 Race/ethnicity (non-minority) 76.0 76.3 60.4 7.15* Single parent family 53.0 52.6 51.6 .04 History of public assistance 96.0 96.1 95.6 .03 <$10 000 in total income 79.0 76.3 78.0 .18

NM = non-maltreatment; PA = physical abuse, OM = other maltreatment. * p < .05, ** p < .01.

8 Michael Teisl and Dante Cicchetti

also experienced emotional maltreatment, and 6.6 percent also experienced sexual abuse. Physical abuse and emotional maltreatment are two subtypes with a particularly high likelihood of being identified together and are often considered interdependent (Hart & Brassard, 1991; McGee & Wolfe, 1991). Although the two maltreatment groups varied in the degree of emotional maltreatment experienced, we determined that any effort to parse out the effects of this particular form of maltreatment would result in an inaccurate representation of physically maltreated children. Thus, we proceeded by investigating the two identified maltreatment groups as originally defined, noting that the primary factor of distinction was the presence versus absence of physically abusive experiences.

Procedures

The recruitment staff first obtained permission from parents for their children to attend a recreational summer camp. Prior to camp attendance, parents also provided informed consent for their children to participate in research activities conducted at camp. Conducting research in a camp environment allows for convenient evaluation of a large number of children in a natural and ecologically valid setting. Furthermore, it affords the opportunity for intensive observation of children’s interactions with their peers in addition to providing an ideal setting for peers to evaluate one another (see Cicchetti & Manly, 1990 for a detailed description of research summer camp procedures).

The structure and organization of the summer camps attended by children in each cohort were comparable. All children attended camp daily from 0900 to 1600 h for one week and were assigned to groups of approximately eight same-sex children, each within one year of another’s age. Groups consisted of roughly equal ratios of mal- treated to non-maltreated children. At the camp, children participated in a variety of group-oriented and recreational activities. Additionally, children were asked to com- plete a series of research measures during assessments and interviews with trained research assistants. At the end of each week, children completed a set of nominations on their peers, while group counselors, who were unaware of children’s maltreatment status and study hypotheses, completed a set of ratings on each of the children in their respective groups.

Measures

Social Information Processing. In order to assess the children’s ability to interpret social cues accurately, we administered to them a modified version of a measure of intention-cue detection developed by Dodge, Murphy, and Buchsbaum (1984). Prior research has noted that deficits on this task are not attributable to intelligence or general cue detection ability (Dodge et al., 1984; Dodge, Price, Bachorowski, & Newman, 1990). The intention-cue detection task was administered in a group format, with each of four research assistants responsible for overseeing two children. Research assistants were available to ensure that the measure was completed properly and to prevent any communication or observation between children, simulating individual administration conditions. Children were presented with a series of 12 videotaped vignettes in which two children are involved in a provocative situation. One child is clearly identified as the provocateur by a colored and numbered T-shirt. Vignettes varied by the provocateur’s intent, with three vignettes of each of the following types: hostile intent, prosocial intent, accidental intent, or ambiguous intent. Following each

Abuse and Aggressive/Disruptive Behavior 9

vignette, the tape was stopped and children were asked two questions. First, children were asked to indicate whether or not the intent of the provocateur in each vignette was either hostile, prosocial, or accidental. Based on the responses, the proportion of hostile interpretations of ambiguous vignettes (three items,

! ! = .50 ) was calculated as

a measure of hostile attributional bias. The proportion of incorrect attributions of hostility, or hostile interpretations of both prosocial and accidental vignettes (six items,

! ! = .62 ), was also calculated as a measure of errors in cue interpretation.

The second question asked children to describe in their own words what they would do if the provocative situation had happened to them. Children’s responses were initially coded into one of eight categories. Consistent with other work on response accessing (Dodge et al., 1995), categories were later collapsed into three more inclu- sive categories representing aggressive, passive, and competent responses. Aggressive responses involved either verbal or physical retaliation, including requests for punish- ment from the teacher. Passive responses included crying in addition to a general failure to act or respond. Competent responses were identified as those geared toward resolving the situation such as inviting the child to play, asking questions about the incident, or other attempts at prosocial interaction. All responses were coded by independent raters with interrater agreement ranging from k = .78 to k = .98. Any disagreements were recoded by an independent third rater. The proportion of aggres- sive responses across all vignettes was then calculated to represent the tendency to access aggressive responses (12 items,

! ! = .81 ).

Peer Nominations. At the end of each camp week, children were asked to rate the other children in their group using a sociometric measure developed by Coie and Dodge (1983). Nominations were based on approximately 30–35 hours of interaction. Coie and Kupersmidt (1983) noted significant correlations between nominations made by familiar peers and unfamiliar peers based upon six playgroup sessions, attesting to the validity of the current use of this measure. Specifically, children were instructed to nominate the two peers they liked the most and the two peers they liked the least, as well as to nominate one peer who best fit each of a set of items describing various behavioral characteristics, including acts shy, cooperative, leader, disruptive, and starts fights. Coie and Dodge (1983) reported a moderate degree of stability over four years for the behavioral description variables, particularly for disruptive (r of grade 3 to 7 = .38, p < .01) and starts fights (r of grade 3 to 7 = .44, p < .01). Children com- pleted this measure individually and were assured that their responses would be kept confidential. The total number of nominations each child received from their peers was determined and converted to proportions of total possible nominations per category. Results were then standardized within each cohort examined. A significant correlation was noted for scores on the disruptive and the starts fights variables (r = .61, p < .01). Scores for these variables were subsequently averaged to generate one score represent- ing overall aggressive/disruptive behavior.

California Child Q-Set. The California Child Q-Set (CCQ; Block & Block, 1980) asks raters familiar with a child to sort 100 items describing children’s behavioral and personality characteristics into nine categories ranging from least descriptive to most descriptive, using a forced-choice format (i.e., a fixed number of items were assigned to one and only one category). At the conclusion of each week of camp, counselors created q-sort profiles for each child in their group based on approxi- mately 30–35 hours of observation. Ratings of emotion regulation were subsequently

10 Michael Teisl and Dante Cicchetti

generated using the emotion regulation q-scale, as developed by Shields and Cicchetti (1997). Intraclass correlations were calculated to determine reliability between counselors in each cohort, with correlations across the two cohorts ranging from .70 to .72.

The emotion regulation q-scale was originally derived from q-sort profiles of a prototypically well-regulated child, as created by 17 doctoral-level psychologists with expertise in emotional development. This q-scale is considered to assess reactivity, empathy, and socially appropriate expressions. Scores on the q-scale were determined from ratings on five positively weighted items (i.e., ‘Can recover from stress’, ‘Can admit negative feelings’, ‘warm/responsive’, ‘Is empathic’, ‘Has genuine/close rela- tionships’) and five negatively weighted items. (i.e., ‘Emotional reactions are inappro- priate’, ‘Is rigid/repetitive when stressed’, ‘Goes to pieces under stress’, ‘Is emotionally labile,’ ‘Is easily irritated’). Prior investigations have indicated that emotion regulation, as assessed using these items, is a distinct construct, which can be differentiated from other aspects of social functioning, such as behavioral regulation (Shields & Cicchetti, 2001), and have highlighted the notable correlation between scores generated from the q-scale and those of other widely used assessments of emotion regulation (see Shields & Cicchetti, 1997; Shields et al., 2001). Nevertheless, in order to avoid overlap between scores on the emotion regulation scale and socio- metric nominations reflective of peer relationships, the item ‘Has genuine/close rela- tionships’ was removed from emotion regulation scale scores used in the present investigation.

Results

Analyses addressed each of the stated hypotheses in order. Children’s age, gender, and race/ethnicity were significantly correlated with dependent variables of interest (see Table 3). Thus, each of these variables was controlled for in all analyses for purposes of consistency. Dependent variables of interest included in this study are peer nomi- nations of aggressive and disruptive behavior and emotion regulation q-scale scores. Additionally, three scores reflecting social information processing errors considered to

Table 3. Correlations Among Variables

1 2 3 4 5 6 7 8

(1) Age — .05 .08 .02 .02 -.31** -.37** .04 (2) Gender — .02 -.03 .13* .07 -.08 .09 (3) Race/ethnicity — .14* -.04 .03 .10 -.07 (4) A/D behavior — .04 .16** .08 -.48** (5) HI (amb) — .40** .16** .06 (6) HI (non) — .36** -.07 (7) Aggressive responses — -.17** (8) Emotion regulation —

A/D behavior = aggressive/disruptive behavior; HI (amb) = hostile interpretations of ambigu- ous vignettes; HI (non) = hostile interpretations of non-hostile vignettes * p < .05, ** p < .01.

Abuse and Aggressive/Disruptive Behavior 11

be associated with aggressive behavior in children were drawn from the intention-cue task and examined as dependent variables. These scores included incorrect attributions of hostility, hostile interpretations of ambiguous vignettes, and aggressive behavioral responses generated by children to all types of vignettes. The skewness of these proportion score variables was examined and found to be within acceptable limits (skewness <1).

A multiple analysis of covariance (MANCOVA) was initially conducted, with age, gender, and race/ethnicity entered as covariates, maltreatment status entered as the independent variable, and all information processing variables, emotion regulation, and peer nominations of aggressive and disruptive behavior entered as dependent variables. Results of the omnibus MANCOVA included a significant overall F for age, F(5, 257) = 14.68, p < .001, hp2 = .22, and for race/ethnicity, F(5, 257) = 2.52, p = .03, hp2 = .05. A significant overall F was also noted for maltreatment status, F(10, 516) = 2.67, p = .003, hp2 = .05. The omnibus MANCOVA was followed up with a series of univariate analyses of covariance (ANCOVA) meant to more closely investi- gate the relationship between maltreatment status and social information processing, emotion regulation, and peers’ perceptions. Planned contrasts were designed to examine differences between the physically abused group and non-maltreated children, as well as differences between children with a history of other forms of maltreatment and non-maltreated children.

Child Maltreatment and Externalizing Behavior

In order to test the first hypothesis that maltreated children would exhibit more aggressive and disruptive behavior than non-maltreated children, an ANCOVA was conducted with maltreatment status as the independent variable, peer nominations of aggression and disruptive behavior as the dependent variable, and children’s age, gender, and race/ethnicity entered as covariates. Results indicated that both race, F(1, 261) = 6.50, p = .01, hp2 = .02, and maltreatment status, F(2, 261) = 5.41, p = .005, hp2 = .04, were significantly associated with peer nominations of aggression and disruptiveness, with minority subjects evidencing higher scores than non-minority subjects. Planned contrasts further indicated that physically abused children were rated by peers as showing significantly more aggression and disruptiveness than were non-maltreated children (mean difference = .41, p < .01, d = .46, CI = .18 to .74) or other maltreated children (mean difference = .40, p < .01, d = .44, 95 percent CI = .12 to .67). Children experiencing other forms of maltreatment did not differ significantly from non-maltreated comparison children (mean difference = .01, NS). Although expectations that the general experience of maltreatment would be associated with higher scores for aggressive/disruptive behavior were not met, results confirmed our hypothesis that peers would perceive physically abused children, in particular, as more aggressive and disruptive than non-maltreated children.

Child Maltreatment and Social Cognition

In order to evaluate whether or not maltreated children, relative to non-maltreated children, evidenced a bias toward identifying the intentions of hypothetical others as hostile, we examined the children’s responses to ambiguous vignettes, as well as their tendencies to misidentify clearly non-hostile as hostile. A univariate ANCOVA was conducted with maltreatment status as the independent variable, the proportion of

12 Michael Teisl and Dante Cicchetti

hostile interpretations of ambiguous vignettes as the dependent variable, and children’s age, gender, and race/ethnicity entered as covariates. Results indicated a significant main effect of gender, F(1, 261) = 4.33, p = .04, hp2 = .02, but no main effect of maltreatment status, F(2, 261) = .19, ns, in predicting the tendency to attribute hostility to others in ambiguous situations. Females appeared more likely than males to make such attributions. A similar analysis with the proportion of hostile interpretations of clearly non-hostile vignettes as the dependent variable found a significant main effect of age, F(1, 261) = 30.62, p < .001, hp2 = .11, with younger children more likely to misperceive hostile intent. Only marginally significant differences were noted based on maltreatment status, F(2, 261) = 2.35, p = .10, hp2 = .02. Thus, it appeared, based on initial analyses, that neither physically abused nor other maltreated children were more likely to develop hostile attribution biases than were non-maltreated children.

The next set of analyses sought to determine whether or not the experience of child maltreatment was associated with other examples of information processing typical of aggression, such as the tendency to access aggressive responses to hypothetical provo- cation. A univariate ANCOVA was conducted with maltreatment status as the inde- pendent variable, the proportion of aggressive responses accessed across all vignettes on the intention-cue task as the dependent variable, and with children’s age, gender, and race/ethnicity entered as covariates. Results indicated significant main effects for both age, F(1, 261) = 47.13, p < .001, hp2 = .15, and race/ethnicity, F(1, 261) = 4.18, p = .04, hp2 = .02, with younger children and children from minority groups generating more aggressive responses. Maltreatment status was also found to be significantly associated with the tendency to access aggressive responses, F(2, 261) = 3.21, p = .04, hp2 = .02. Planned contrasts revealed that both physically abused children (mean difference = .09, p = .03, d = .33, CI = .05 to .61), as well as children with histories of other forms of maltreatment (mean difference = .08, p = .03, d = .30, CI = .02 to .58) were significantly more likely to access aggressive responses than were non-maltreated children. Thus, consistent with our hypotheses, maltreated children, in general, were found to access more aggressive responses than non-maltreated children, although it did not appear that physically abused children were uniquely at risk. However, using Bonferroni corrections based upon multiple comparisons, both comparisons fell just short of reaching the required significance level of p < .025, indicating that these results should be interpreted with caution.

Child Maltreatment and Emotion Regulation

Analyses were conducted to examine the third set of hypotheses, which predicted that children who had been maltreated would receive lower ratings of emotion regulation from camp counselors than would non-maltreated comparison children. A univariate ANCOVA with maltreatment status as the independent variable, q-sort ratings of emotion regulation as the dependent variable, and children’s age, gender, and race/ ethnicity entered as covariates provided support for the hypothesis, F(2, 261) = 5.50, p = .005, hp2 = .04. Planned contrasts indicated that both physically abused children (mean difference = -4.62, p < .001, d = -.49, CI = -.77 to -.21), as well as children with a history of other forms of maltreatment (mean difference = -2.96, p = .03, d = -.31, CI = -.59 to -.03) evidenced poorer emotion regulation as compared to non-maltreated comparison children. However, after employing Bonferroni correc- tions, only the contrast between physically abused and non-maltreated children remained significant (Table 4).

Abuse and Aggressive/Disruptive Behavior 13

T ab

le 4.

M ea

n s

by M

al tr

ea tm

en t

S ta

tu s

M al

tr ea

tm en

t S

ta tu

V ar

ia bl

e N

M M

(S D

) PA

M (S

D )

O M

M (S

D )

T ot

al M

(S D

) R

an ge

A /D

be ha

vi or

-. 12

(. 90

) .2

9 (.

90 )

-. 10

(. 91

) .0

3 (.

90 )

-1 .3

2 to

2. 50

H I

(a m

b) .5

9 (.

32 )

.5 9

(. 32

) .6

2 (.

32 )

.6 0

(. 33

) .0

0 to

1. 00

H I

(n on

) .3

1 (.

28 )

.3 9

(. 28

) .3

1 (.

28 )

.3 3

(. 28

) .0

0 to

1. 00

A gg

re sp

on se

s .5

0 (.

26 )

.5 9

(. 26

) .5

8 (.

27 )

.5 6

(. 26

) .0

0 to

1. 00

E m

ot io

n re

gu la

ti on

53 .0

6 (9

.4 4)

48 .4

4 (9

.4 1)

50 .1

0 (9

.4 7)

50 .5

3 (9

.4 3)

23 .5

0 to

66 .0

A /D

be ha

vi or

= ag

gr es

si ve

/d is

ru pt

iv e

be ha

vi or

;H I

(a m

b) =

ho st

il e

in te

rp re

ta ti

on s

of am

bi gu

ou s

vi gn

et te

s; H

I (n

on )

= ho

st il

e in

te rp

re ta

ti on

s of

no n-

ho st

il e

vi gn

et te

s; ag

g re

sp on

se s

= ag

gr es

si ve

re sp

on se

s; N

M =

no n-

m al

tr ea

tm en

t; PA

= ph

ys ic

al ab

us e,

O M

= ot

he r

m al

tr ea

tm en

14 Michael Teisl and Dante Cicchetti

Analyses of Mediation

In order to evaluate whether or not the experience of maltreatment influenced chil- dren’s social behavior by way of its impact on the development of social cognition and emotion processes, a series of hierarchical regressions was performed. Mediation was evaluated using the procedures suggested by Baron and Kenny (1986), as well as tests suggested by MacKinnon, Lockwood, Hoffman, West, and Sheets (2002). Given that only physically abused children were previously found to differ significantly from non-maltreated comparison children on nominations of aggression and disruptiveness, children experiencing other types of maltreatment were initially excluded from analy- ses regarding mediation.

Firstly, given the reduced sample, linear regression analyses were conducted in order to verify relationships between physical abuse status and the dependent variables of interest. Peer nominations of behavior problems and all proposed mediators, including incorrect ratings of hostility, were examined based upon our initial hypotheses regard- ing physical abuse. Results of these analyses indicated that, after controlling for age, gender, and race/ethnicity, physical abuse status continued to significantly predict peer nominations of behavior problems (b = .22, p = .004), with physically abused children more often nominated as aggressive and disruptive. Physical abuse status also signifi- cantly predicted three of the proposed mediators, with physically abused children evidencing more incorrect ratings of hostility (b = .14, p = .05), more aggressive responses (b = .15, p = .03), and poorer emotion regulation (b = -.23, p = .003).

Next, in order to establish a relationship between each mediator and peer nomina- tions of aggressive and disruptive behavior, three hierarchical regression analyses were conducted, controlling for children’s age, gender, and race/ethnicity. Results indicated that incorrect ratings of hostility (b = .24, p = .003) and emotion regulation (b = -.56, p < .001) each accounted for significant portions of the variance in peer nominated behavior problems, while aggressive response accessing did not, (b = .09, ns). Thus, the tendency to access aggressive responses was excluded from further analyses of mediation, as this variable failed to meet the requirements outlined by Baron and Kenny (1986).

An additional hierarchical regression was subsequently performed to assess for mediation by both cognitive and emotional processes (see Table 5). Children’s age, gender, and race/ethnicity were again controlled for by entry at Step 1, with incorrect ratings of hostility and q-sort ratings of emotion regulation entered at Step 2. Physi- cal abuse status was entered at Step 3 and no longer significantly predicted aggres- sive and disruptive behavior. Both incorrect ratings of hostility and emotion regulation, however, remained significant predictors, with the tendency to misper- ceive hostility and poorer emotion regulation associated with more nominations of problem behavior.

Next we tested the significance of the intervening variable effects using methods suggested by MacKinnon et al. (2002). We calculated z" = ab/se[ab] for each mediator and found that the indirect pathways through both emotion regulation (z" = 2.87, p < .001) as well as social cognition (z" = 1.52, p < .01) were significant. We further sought to understand the proportion of the overall effect accounted for by each media- tor, and thus divided the effect of each variable by the total effect of the model, using absolute values as suggested by Alwin and Hauser (1975). Results revealed that information processing errors accounted for 10 percent of the overall effect, whereas emotion regulation accounted for 55 percent.

Abuse and Aggressive/Disruptive Behavior 15

T ab

le 5.

M u

lt ip

le M

ed ia

to r

M od

el of

th e

P hy

si ca

lA b

u se

— A

gg re

ss iv

e/ D

is ru

p ti

ve B

eh av

io r

R el

at io

n sh

V ar

ia bl

e b

t M

od el

R 2

F R

2 ch

an ge

F ch

an ge

St ep

1 .0

1 .6

7 .0

1 .6

7 A

ge .0

1 .1

7 R

ac e

.1 1

1. 39

G en

de r

.0 0

-. 02

St ep

2 .3

4 17

.4 7*

** .3

3 42

.1 9*

** A

ge .0

8 1.

15 R

ac e

.0 4

.6 1

G en

de r

-. 02

-. 39

H os

ti le

in te

rp re

ta ti

on s

(n on

) .1

7 2.

45 *

E m

ot io

n re

gu la

ti on

-. 54

-8 .4

8* **

St ep

3 .3

5 14

.8 2*

** .0

1 1.

39 A

ge .0

7 .9

9 R

ac e

.0 4

.6 4

G en

de r

-. 03

-. 47

H os

ti le

in te

rp re

ta ti

on s

(n on

) .1

6 2.

28 *

E m

ot io

n re

gu la

ti on

-. 52

-8 .0

4* **

P hy

si ca

l ab

us e

st at

us .0

8 1.

* p

< .0

5, **

* p

< .0

01 .

16 Michael Teisl and Dante Cicchetti

A similar test was conducted to evaluate potential intervening variable effects in explaining aggressive behavior in the group of maltreated children who had not been physically abused. Children’s age, gender, and race/ethnicity were controlled by initial entry in each regression. While regression analyses indicated no significant differences between non-physically abused maltreated children and the non-maltreated compari- son group in terms of peer nominations of aggressive and disruptive behavior (b = .01, NS), it is still possible to examine indirect effects using MacKinnon et al.’s (2002) methods. However, because neither misperceived hostility (b = .12, NS) nor aggressive response tendency (b = -.02, NS) significantly predicted aggressive and disruptive behavior when comparing the non-physically abused maltreatment group with the non-maltreated comparison group, it was clear that there could be no indirect effect of maltreatment through these variables. Emotion regulation, on the other hand, did significantly predict aggressive and disruptive behavior (b = -.40, p < .001), and the indirect pathway from maltreatment to behavior through emotion regulation also was found to be significant (z" = 2.30, p < .001).

Discussion

The findings reported in this study expand the understanding of processes underlying the development of aggression and disruptive behavior problems often demonstrated by maltreated youth in the context of their peer relationships. As expected, initial analyses indicated that peers rated the subgroup of maltreated children who had experienced physical abuse as more aggressive and disruptive than non-maltreated children. Children with a history of other forms of maltreatment, on the other hand, were not found to significantly differ from non-maltreated children on this account. These findings are consistent with other investigations identifying physically abused children as more likely to exhibit externalizing behavior and to be negatively evaluated by their peers (Manly et al., 2001; Trickett & McBride-Chang, 1995).

Subsequent analyses aimed to elucidate the factors purported to be responsible for the development of these behavioral problems in physically abused children. We examined the role of social cognitive processes in predicting aggressive behavior problems in maltreated children, as deviations in this area have previously been linked to aggression in research with normative samples of children (Dodge, 1980; Dodge et al., 1986), and children with histories of harsh physical discipline and physical abuse (Dodge et al., 1990; Dodge et al., 1995). Initial efforts to identify the presence of a hostile attributional bias in maltreated children found no significant differences in the interpretations of ambiguous scenarios based upon maltreatment status. There was also no significant main effect of maltreatment status on the ten- dency to misidentify non-hostile vignettes as hostile. Furthermore, only marginally significant differences in the tendency to generate aggressive behavioral responses to hypothetical provocations were noted between maltreatment groups. Although overall group differences were not significant, regression analyses indicated that, relative to non-maltreated children, physically abused children were more likely to incorrectly interpret prosocial and accidental scenarios as hostile and were more likely to access aggressive responses to provocation. Only the tendency to misinterpret others’ inten- tions as hostile, however, was related to nominations of aggressive and disruptive behavior from peers.

It is interesting that maltreated children in this study were no more likely to evidence hostile interpretations of ambiguous events than were non-maltreated comparison

Abuse and Aggressive/Disruptive Behavior 17

children. Perhaps this measure simply lacked robustness, as the variable was based on three items and evidenced moderate internal consistency. Alternatively, however, the lack of differences noted might be due to the fact that all participants were drawn from low-income urban neighborhoods where exposure to community violence is typical. As a result, provocative acts from others with ambiguous intent might easily be perceived as deliberately hostile by any of these children. Regardless, physically abused children demonstrated a distinct sensitivity to perceiving hostility, even when it was apparent that the provocateur held no malicious intent. This propensity for actually misperceiving aggression from others represents a considerable error in social infor- mation processing which at least partially explains the heightened risk for aggressive and disruptive behavior in physically abused children. These results suggest that experiences of hostility and violence may lead physically abused children to distort the meaning of even non-threatening cues in the environment, a perceptual style which may have developed as a protective mechanism against future hostility. Although the data cannot provide confirmation, this notion has also been suggested by previous research (Pollak et al., 2000; Rieder & Cicchetti, 1989).

Aspects of emotional development were also included in this study due to the documented association between dysregulated emotion and aggression (Cole & Zahn-Waxler, 1992), as well as the important role that parent–child interactions play in emotional development. Given that children in maltreating families are the recipients of frequently inconsistent and/or insensitive parenting (Rogosch, Cicchetti, & Aber, 1995), these children are considered less likely to form an appropriate understanding of emotion and emotional expression, and are at increased risk for the development of attenuated emotional competence (Denham, Mitchell-Copeland, Strandberg, Auerbach, & Blair, 1997). The failure of caregivers to respond sensitively to children’s emotional needs and to serve as appropriate models for emotional expression and coping with arousal is also likely to contribute to significant disruptions in emotional development, particularly in terms of children’s capacity for emotional self-regulation (Thompson & Calkins, 1996). Although it was expected than any maltreatment expe- rience would result in lower ratings of emotion regulation, only physically abused children were found to be at significant risk in the current investigation.

Because children with a history of maltreatment subtypes other than physical abuse did not exhibit significantly more aggression and disruptive behavior than any non- maltreated children, further examinations of the factors underlying the development of aggressive and disruptive behavior problems focused on physically abused children. A mediation analysis indicated that physical abuse was significantly associated with an increased likelihood for misperceiving hostility from others, easy access to aggressive responses to conflict, and poorer emotion regulation. Only misperceived hostility and poor emotion regulation, however, were found to significantly predict children’s aggressive and disruptive behavior. Furthermore, both of these variables at least partially accounted for the relationship between physical abuse and aggressive and disruptive behavior problems.

Although maltreated children who had not been physically abused did not appear to be more aggressive and disruptive than the non-maltreated comparison group, an indirect effects analysis was conducted to determine whether or not the same cognitive and emotional factors accounting for the behavior problems evident in physically abused children were also operative in explaining acting-out behavior in other mal- treated children. The results indicated that social information processing variables were not related to aggressive and disruptive behavior in maltreated children who had not

18 Michael Teisl and Dante Cicchetti

been physically abused. However, the behavior problems evident in these children were to a certain extent the result of the deleterious impact that maltreatment experiences had on their ability to effectively regulate their emotions. Thus, general emotion regulation deficits appear to confer a risk for behavioral difficulties in all maltreated children. The tendency to misinterpret others’ intentions as hostile, on the other hand, serves as a risk factor for aggressive behavior that is distinctly associated with the experience of physical abuse.

The present investigation confirms the findings of Shields and colleagues, which indicate the considerable importance of emotional processes in the etiology of behav- ioral dysregulation (Shields & Cicchetti, 2001; Shields et al., 1994). The current study also extends the work of Shields et al. (2001) by indicating that the negative expecta- tions that maltreated children develop about others hold differential importance for behavioral outcomes based upon the type of maltreatment experienced. Whereas negative representations and hypervigilance to threat may lead to reactive aggression in physically abused children, they may lead to other defensive reactions, such as fearful avoidance, in other maltreated children who have not been physically abused. Future research should investigate this hypothesis as well as attempt to clarify the specific cognitive and emotional processes underlying this phenomenon.

The results of the current study are consistent with investigations noting that maladaptation in both the cognitive as well as affective domains makes unique con- tributions to the development of problematic relationships with peers (Rogosch et al., 1995), and emphasize the importance of considering multiple domains of functioning in efforts geared toward improving the understanding of the processes underlying development (Cicchetti & Schneider-Rosen, 1984; Sroufe, 1996). These findings high- light the diversity of impact that the experience of child maltreatment can have on various developmental processes, indicating that maladaptation varies across maltreat- ment subtypes. Furthermore, the present investigation underscores the need to consider not only the iniquities inherent to the experience of maltreatment itself, but also the differential effects that maladaptation in multiple domains of functioning may have in mediating the varied pathways leading from disruptions in the early caregiving envi- ronment to later maladjustment.

It is important to note that there are some limitations to the current study. One limitation involves the measurement of errors in social cognition. In this study, we examined two of the six stages outlined by Dodge and his colleagues in their descrip- tion of the social information-processing model of peer relations (Crick & Dodge, 1994; Dodge, 1986). This may explain the relatively minor role of social information- processing patterns found in mediating the relationship between physical abuse and aggressive and disruptive behavior. If we had included other aspects of information processing, such as expectations of positive outcomes of aggression or the encoding of social cues, aggressive patterns of processing might have been more influential in explaining the development of behavior problems in physically abused children.

Another limitation of this study is the manner in which emotion regulation was measured. The q-sort method employed in the present investigation did not allow for an examination of distinct patterns of emotion regulation difficulty. Research inve- stigating patterns of emotion dysregulation as overcontrolled or undercontrolled (e.g., Maughan & Cicchetti, 2002), with a closer examination of particular aspects of affective regulation (i.e., lability, empathy, expression, etc.; e.g., Shields and Cicchetti 1998) will be beneficial in identifying whether or not specific patterns or elements are uniquely associated with the exhibition of externalizing behavior.

Abuse and Aggressive/Disruptive Behavior 19

In summary, the present investigation succeeded in building on the extant literature regarding the development of externalizing behavior problems in children. This study is among the first to address the role that both cognitive and affective processes concur- rently play in predicting children’s aggressive interactions with peers, particularly in considering the development of these processes in maltreated children. Although social cognition certainly played a role in explaining the relationship between physical abuse and children’s aggression, it is clear that maladaptive emotional processes were oper- ating as well. In fact, emotion regulation was a stronger mediator of the relationship between physical abuse and behavior problems, highlighting the important risks that disruptions in the development of emotional processes hold for the successful negotia- tion of peer relationships. These findings stress the need to focus interventions on improving emotion regulation strategies for children with externalizing behavior prob- lems. Future research based upon the work of Lemerise and Arsenio (2000) might extend the findings of the present study by examining how aspects of social cognition and emotion processes may interact in accounting for the problematic behavior dem- onstrated by physically abused children during peer interactions.

The findings presented in this study make a significant contribution toward im- proving the understanding of the processes linking early aversive experiences to later maladaptation, and underscore the importance of considering multiple domains of development in explicating these processes. The relevance of these findings is strength- ened by the conservative nature of comparisons, as maltreated children were contrasted with a particularly high-risk sample of non-maltreated children. It is also notable that children were observed in an ecologically valid summer camp setting where peer interactions were spontaneous and natural. In short, this investigation highlighted the risk that aversive childhood experiences, such as child maltreatment, have for malad- aptation in the cognitive and affective domains, thereby contributing to subsequent difficulties in managing peer relations. Physically abused children, in particular, were noted to be at unique risk for the development of aggression and disruptive behavior problems through the impact of abuse on maladaptive functioning in both cognitive and affective areas. That the relationships among variables of interest in this study developed over one week suggests the seriousness of this risk for physically abused children entering novel social contexts, and underscores the importance of interven- tions targeted at improving the cognitive and emotional functioning of these children.

References

Alwin, D. F., & Hauser, R. M. (1975). The decomposition of effects in path analysis. American Sociological Review, 40, 37–47.

Barnett, D., Manly, J. T., & Cicchetti, D. (1993). Defining child maltreatment: The interface between policy and research. In D. Cicchetti, & S. L. Toth (Eds.), Child abuse, child devel- opment, and social policy (pp. 7–73). Norwood, NJ: Ablex.

Baron, R. M., & Kenny, D. A. (1986). The moderator-mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Per- sonality and Social Psychology, 51, 1173–1182.

Block, J., & Block, J. H. (1980). The California Child Q-set. Palo Alto, CA: Consulting Psychologists Press. (Original work published 1969).

Bolger, K. E., & Patterson, C. J. (2001). Developmental pathways from child maltreatment to peer rejection. Child Development, 72, 549–568.

Calkins, S. D. Gill, K. L., Johnson, M. C., & Smith, C. L. (1999). Emotional reactivity and emotional regulation strategies as predictors of social behavior with peers during toddlerhood. Social Development, 8, 310–334.

20 Michael Teisl and Dante Cicchetti

Cicchetti, D. (1993). Developmental psychopathology: Reactions, reflections, projections. Developmental Review, 13, 471–502.

Cicchetti, D., & Barnett, D. (1991). Toward the development of a scientific nosology of child maltreatment. In W. M. Grove, & D. Cicchetti (Eds.), Thinking clearly about psychology, Volume 2: Personality and psychopathology (pp. 346–377). Minneapolis, MN: University of Minnesota Press.

Cicchetti, D., & Lynch, M. (1995). Failures in the expectable environment and their impact on individual development: The case of child maltreatment. In D. Cicchetti, & D. J. Cohen (Eds.), Developmental psychopathology, Vol. 2: Risk, disorder, and adaptation. Wiley series on personality processes (pp. 32–71). Oxford: Wiley.

Cicchetti, D., & Manly, J. T. (1990). A personal perspective on conducting research with maltreating families: Problems and solutions. In E. Brody, & I. Sigel (Eds.), Family research, Volume 2, Families at risk (pp. 32–71). New York: Wiley.

Cicchetti, D., & Rizley, R. (1981). Developmental perspectives on the etiology, intergenera- tional transmission, and sequelae of child maltreatment. New Directions for Child Develop- ment, 11, 31–55.

Cicchetti, D., & Schneider-Rosen, K. (1984). Theoretical and empirical considerations in the investigation of the relationship between affect and cognition in atypical populations of infants: Contributions to the formulation of an integrative theory of development. In C. Izard, J. Kagan, & R. Zajonc (Eds.), Emotions, cognition, and behavior (pp. 366–406). New York: Cambridge University Press.

Coie, J. D., & Dodge, K. A. (1983). Continuities and changes in children’s social status: A five-year longitudinal study. Merrill-Palmer Quarterly, 29, 261–282.

Coie, J. D., & Kupersmidt, J. B. (1983). A behavioral analysis of emerging social status in boys’ groups. Child Development, 54, 1400–1416.

Coie, J. D., Lochman, J. E., Terry R., & Hyman, C. (1992). Predicting early adolescent disorder from childhood aggression and peer rejection. Journal of Consulting & Clinical Psychology, 60, 783–792.

Cole, P. M., Martin, S. E., & Dennis, T. A. (2004). Emotion regulation as a scientific construct: Methodological challenges and directions for child development research. Child Develop- ment, 75, 317–333.

Cole, P. M., & Zahn-Waxler, C. (1992). Emotional dysregulation in disruptive behavior disorders. In D. Cicchetti, & S. L. Toth (Eds.), Developmental perspectives on depression. Rochester symposium on developmental psychopathology, Vol. 4 (pp. 173–209). Rochester, NY: University of Rochester Press.

Crick, N. R., & Dodge, K. A. (1994). A review and reformulation of social information- processing mechanisms in children’s social adjustment. Psychological Bulletin, 115, 74–101.

Denham, S. A., Mitchell-Copeland, J., Strandberg, K., Auerbach, S., & Blair, K. (1997). Parental contributions to preschoolers’ emotional competence: Direct and indirect effects. Motivation & Emotion, 21, 65–86.

Dodge, K. A. (1980). Social cognition and children’s aggressive behavior. Child Development, 51, 162–170.

Dodge, K. A. (1986). A social information processing model of social competence in children. In M. Perlmutter (Ed.), The Minnesota Symposium on Child Psychology (Vol. 18, pp. 77–125). Hillsdale, NJ: Erlbaum.

Dodge, K. A. (1991). Emotion and social information processing. In J. Garber, & K. A. Dodge (Eds.), The development of emotion regulation and dysregulation. Cambridge studies in social and emotional development (pp. 159–181). New York: Cambridge University Press.

Dodge, K. A., Bates, J. E., & Pettit, G. S. (1990). Mechanisms in the cycle of violence. Science, 250, 1678–1683.

Dodge, K. A., Murphy, R. R., & Buchsbaum, K. (1984). The assessment of intention-cue detection skills in children: Implications for developmental psychopathology. Child Devel- opment, 55, 163–173.

Dodge, K. A., Pettit, G. S., Bates, J. E., & Valente, E. (1995). Social information-processing patterns partially mediate the effect of early physical abuse on later conduct problems. Journal of Abnormal Psychology, 104, 632–643.

Dodge, K. A., Pettit, G. S., McClaskey, C. L., & Brown, M. M. (1986). Social competence in children. Monographs of the Society for Research in Child Development, 51, 1–85.

Abuse and Aggressive/Disruptive Behavior 21

Dodge, K. A., Price, J. M., Bachorowski, J., & Newman, J. P. (1990). Hostile attributional biases in severely aggressive adolescents. Journal of Abnormal Psychology, 99, 385– 392.

Dodge, K. A., & Somberg, D. R. (1987). Hostile attributional biases among aggressive boys are exacerbated under conditions of threats to the self. Child Development, 58, 213– 224.

Eisenberg, N., Fabes, R. A., Guthrie, I. K., Murphy, B. C., Maszk, P., Holmgren, R., et al. (1996). The relations of regulation and emotionality to problem behavior in elementary school children. Development and Psychopathology, 8, 141–162.

Fabes, R. A., & Eisenberg, N. (1992). Young children’s coping with interpersonal anger. Child Development, 63, 116–128.

Farrington, D. P. (1995). The development of offending and antisocial behavior from childhood: Key findings from the Cambridge Study in Delinquent Development. Journal of Child Psychology and Psychiatry, 36, 929–964.

Giovannoni, J. M., & Becerra, R. M. (1979). Defining child abuse. New York: Free Press. Graham, S., Hudley, C., & Williams, E. (1992). Attributional and emotional determinants of

aggression among African-American and Latino young adolescents. Developmental Psychol- ogy, 28, 731–740.

Hart, S. N., & Brassard, M. R. (1991). Psychological maltreatment: Progress achieved. Devel- opment and Psychopathology, 3, 61–70.

Huesmann, L. R., Eron, L. D., Lefkowitz, M. M., & Walder, L. O. (1984). Stability of aggression over time and generations. Developmental Psychology, 20, 1120–1134.

Lemerise, E. A., & Arsenio, W. F. (2000). An integrated model of emotion processes and cognition in social information processing. Child Development, 71, 107–118.

MacKinnon, D. P., Lockwood, C. M., Hoffman, J. M., West, S. G., & Sheets, V. (2002). A comparison of methods to test the significance of the mediated effect. Psychological Methods, 7, 83–104.

Manly, J. T., Kim, J. E., Rogosch, F. A., & Cicchetti, D. (2001). Dimensions of child maltreat- ment and children’s adjustment: Contributions of developmental timing and subtype. Devel- opment & Psychopathology, 13, 759–782.

Maughan, A., & Cicchetti, D. (2002). Impact of child maltreatment and interadult violence on children’s emotion regulation abilities and socioemotional adjustment. Child Development, 73, 1525–1542.

McGee, R. A., & Wolfe, D. A. (1991). Psychological maltreatment: Toward an operational definition. Development and Psychopathology, 3, 3–18.

Moffitt, T. E. (1993). Adolescence-limited and life-course-persistent antisocial behavior: A developmental taxonomy. Psychological Review, 100, 674–701.

Moffitt, T. E., & Caspi, A. (2001). Childhood predictors differentiate life-course persistent and adolescence-limited antisocial pathways among males and females. Development & Psychopathology, 13, 355–375.

Moffitt, T. E., Caspi, A., Harrington, H., & Milne, B. J. (2002). Males on the life-course- persistent and adolescence-limited antisocial pathways: Follow-up at age 26 years. Develop- ment & Psychopathology, 14, 179–207.

Nasby, W., Hayden, B., & DePaulo, B. M. (1980). Attributional bias among aggressive boys to interpret unambiguous social stimuli as displays of hostility. Journal of Abnormal Psychol- ogy, 89, 459–468.

Olweus, D. (1979). Stability of aggressive reaction patterns in males: A review. Psychological Bulletin, 86, 852–875.

Orobio de Castro, B., Slot, N. W., Bosch, J. D., Koops, W., & Veerman, J. W. (2003). Negative feelings exacerbate hostile attibutions of intent in highly aggressive boys. Journal of Clinical Child & Adolescent Psychology, 32, 56–65.

Pollak, S. D., Cicchetti, D., Hornung, K., & Reed, A. (2000). Recognizing emotion in faces: Developmental effects of child abuse and neglect. Developmental Psychology, 36, 679– 688.

Pollak, S. D., & Sinha, P. (2002). Effects of early experience on children’s recognition of facial displays of emotion. Developmental Psychology, 38, 784–791.

Pollak, S. D., & Tolley-Schell, S. A. (2003). Selective attention to facial emotion in physically abused children. Journal of Abnormal Psychology, 112, 323–338.

22 Michael Teisl and Dante Cicchetti

Rieder, C., & Cicchetti, D. (1989). Organizational perspective on cognitive control functioning and cognitive-affective balance in maltreated children. Developmental Psychology, 25, 382– 393.

Rogosch, F. A., Cicchetti, D., & Aber, J. L. (1995). The role of child maltreatment in early deviations in cognitive and affective processing abilities and later peer relationship problems. Development & Psychopathology, 7, 591–609.

Rogosch, F. A., Cicchetti, D., Shields, A., & Toth, S. L. (1995). Parenting dysfunction in child maltreatment. In M. H. Bornstein (Ed.), Handbook of parenting, Vol. 4: Applied and practical parenting. (pp. 127–159). Hillsdale, NJ: Erlbaum.

Shields, A., & Cicchetti, D. (1997). Emotion regulation among school-age children: the development and validation of a new criterion q-sort scale. Developmental Psychology, 33, 906–916.

Shields, A., & Cicchetti, D. (1998). Reactive aggression among maltreated children: The contributions of attention and emotion dysregulation. Journal of Clinical Child Psychology, 27, 381–395.

Shields, A., & Cicchetti, D. (2001). Parental maltreatment and emotion dysregulation as risk factors for bullying and victimization in middle childhood. Journal of Clinical Child Psy- chology, 30, 349–363.

Shields, A. M., Cicchetti, D., & Ryan, R. M. (1994). The development of emotional and behavioral self-regulation and social competence among maltreated school-age children. Development and Psychopathology, 6, 57–75.

Shields, A., Ryan, R. M., & Cicchetti, D. (2001). Narrative representations of caregivers and emotion dysregulation as predictors of maltreated children’s rejection by peers. Developmen- tal Psychology, 37, 321–337.

Shipman, K., & Zeman, J. (2001). Socialization of children’s emotion regulation in mother- child dyads: A developmental psychopathology perspective. Development & Psychopathol- ogy, 13, 317–336.

Slaby, R. G., & Guerra, N. G. (1988). Cognitive mediators of aggression in adolescent offend- ers: 1 assessment. Developmental Psychology, 24, 580–588.

Smith, C., & Thornberry, T. P. (1995). The relationship between child maltreatment and adolescent involvement in delinquency. Criminology, 33, 451–481.

Sroufe, L. A. (1990). Considering normal and abnormal together: The essence of developmental psychopathology. Development & Psychopathology, 2, 335–347.

Sroufe, L. A. (1996). Emotional development: The organization of emotional life in the early years. New York: Cambridge University Press.

Thompson, R. A. (1994). Emotion regulation: A theme in search of definition. Monographs of the Society for Research in Child Development, 59, 250–283.

Thompson, R. A., & Calkins, S. D. (1996). The double-edged sword: Emotional regulation for children at risk. Development & Psychopathology, 8, 163–182.

Trickett, P. K., & McBride-Chang, C. (1995). The developmental impact of different forms of child abuse and neglect. Developmental Review, 15, 311–337.

Acknowledgments

This research was conducted with the support of grants from the National Institute of Mental Health, Office of Child Abuse and Neglect, and the Spunk Fund, Inc.

We would like to acknowledge the efforts of Joan Kaufman, Ph.D. and Jody Todd Manly, Ph.D. in their roles as directors of the research summer camps. We also would like to recognize all of the research assistants as well as the children and families who helped make this work possible. Finally, we wish to thank Carol Ann Dubovsky for all of her support and assistance in managing the data used in this study as well as Fred Rogosch, Ph.D. and Jungmeen Kim, Ph.D. for their feedback regarding statistical procedures used.

Abuse and Aggressive/Disruptive Behavior 23