Oppositional defiant disorder

Katherine Coy [1]

Matthew L. Speltz [1,3]

Michelle DeKlyen [1]

Karen Jones [2]

The social--cognitive characteristics of 88 preschool boys with Oppositional Defiant Disorder (ODD) and 80 nondisruptive boys were assessed 3 times over a 2-year period. Three questions were addressed: (1) Do social-cognitive processes (encoding, attribution, problem solving, and response evaluation) distinguish clinic-referred preschool boys from peers without behavior problems? (2) What is the relation between preschoolers' social-cognitive processes and observed problem behavior? (3) Are the social-cognitive processes of clinic boys with ODD influenced by comorbidity with Attention Deficit Hyperactivity Disorder (ADHD)? Boys were presented hypothetical, peer-oriented social dilemmas to resolve. At all 3 assessments, clinic boys were twice as likely as were comparison group boys to generate aggressive solutions. Relative to comparison boys, clinic boys' encoding of social information was less accurate. The groups did not differ in their attributions or response evaluations. Verbal IQ and language skills wer e modestly correlated with problem solving and encoding. Within the clinic group, social--cognitive processes were not affected by ADHD comorbidity and they showed little relation to later diagnostic status or severity of behavior problems.

KEY WORDS: Oppositional Defiant Disorder; social cognition; social information-processing.

The onset of serious disruptive behavior in the preschool and early grade school years can mark the beginning stages of "life-course persistent" antisocial behavior (Moffitt, 1993) and other forms of psychopathology (Speltz, McClellan, DeKlyen, & Jones, 1999). Young children referred to clinics for early disruptive behavior are typically boys who are noncompliant, emotionally reactive, and socially provocative. A diagnosis of Oppositional Defiant Disorder (ODD) is often made, in many cases with cooccurring Attention Deficit Hyperactivity Disorder (ADHD; Rey, 1993). When boys meeting these criteria also show physical aggression, the odds of stable antisocial behavior are considerable: As many as one-third to one-half of the children presenting with this combination of problems may eventually receive diagnoses of conduct disorder (CD) and, as adults, antisocial personality disorder (Loeber, Green, Keenan, & Lahey, 1996; Robins, 1991).

Most researchers agree that disruptive behavior problems in early life stem from multiple, interacting factors (Campbell, 1990; Greenberg, Speltz, & DeKlyen, 1993; Loeber, 1990; Rutter, 1988), but few specific pathways have been delineated. Most studies on the origins of disruptive behavior have focused on parent, family, and community factors (see Campbell, 1995, for a review). Among the various child factors that might contribute, temperament has been the most extensively studied (e.g. Coon, Carey, Corley, & Fulker, 1992; Pullis, 1991; Thomas & Chess, 1984). Less attention has been given to social--cognitive processes, which are the focus of this paper. Although interpersonal sensitivity and emotional reactivity are core presenting features of children with ODD, little is known about how these youngsters think about social relationships with peers and unfamiliar adults.

THEORY AND RESEARCH ON CHILDREN'S SOCIAL-COGNITIVE PROCESSES

Dodge (1986) and Crick and Dodge (1994) have proposed a model in which children are believed to "process" social information in a series of discrete steps. These include the encoding and interpretation of information, the formation of social goals and actions, a decision to act in a particular way, and self-evaluation of the effects and desirability of these actions. In comparison with typical children, children with observable social problems are believed to encode fewer relevant cues, show more biased attributions of others' ambiguous social behaviors (e.g., expecting hostility when none was intended), and generate less effective problem solutions that they nevertheless view positively. This pattern of social cognition is likely to enhance the probability of children's reactively aggressive or defiant behavior, which may be perceived by the child as appropriately defensive. Overtime, the behaviors of these children are likely to alienate others and elicit objective hostility, reinforcing the child's expect ations.

Support for the Crick and Dodge (1994) model has come largely from studies examining the social-cognitive characteristics of children from regular school settings. These studies have compared the social information-processing skills of children rated by teachers or peers to be socially problematic (e.g., receiving relatively high ratings of aggressiveness or social rejection) with the skills of more competent children. Inmost studies, children with social problems have been found to encode fewer social cues (Dodge & Tomlin, 1987) and more often to attribute hostile intentions to others' behavior than have more socially competent peers (Dodge, 1980; Slaby & Guerra, 1988). Children who are socially unskilled, socially rejected, or both, have been shown to access fewer competent and more aggressive and inept ways of solving social problems than have better-accepted children (Ladd & Oden, 1979; Renshaw & Asher, 1983; Spivak, Platt, & Shure, 1976). Socially problematic children also are more likely than their soc ially skilled peers to anticipate positive results from behaving aggressively (Crick & Ladd, 1990; Perry, Perry, & Rasmussen, 1986). Further, there is evidence that social information processing patterns are causally related to the behavior of socially rejected children, as observed in laboratory play groups (Rabiner & Coie, 1989).

APPLICATIONS TO CLINICAL SAMPLES OF DISRUPTIVE CHILDREN

The Crick and Dodge (1994) model has clear application to children with ODD, ADHD, and conduct disorder (CD) or a combination of these, who are less socially skilled, less popular, and more aggressive than their peers (Abikoff, Gittelman, & Klein, 1980; Boulanger & Langevin, 1992; Erhardt & Hinshaw, 1994; Matthys et al., 1995). Several symptoms of ODD (DSM-III-R, American Psychiatric Association, 1987; DSM-IV, American Psychiatric Association, 1994) pertain directly or indirectly to social-cognitive processes. For example, "often blames others for mistakes or misbehavior" implies biased attributions of responsibility for interpersonal conflict. Other ODD symptom criteria such as "easily annoyed," "resentful," and "spiteful and vindictive" suggest a pattern of hostile perceptions and expectations regarding interpersonal relationships. Furthermore, the inattention and impulsiveness of children with comorbid ODD and ADHD may lead to impaired encoding of social cues.

Despite these conceptual links, few studies have examined the social-cognitive characteristics of children with diagnosed disruptive behavior disorders. Dunn, Lochman, and Colder (1997) examined the social- cognitive characteristics of clinic-referred boys aged 6- 15 years with either ODD or CD. They found that boys with CD showed more deficits (e.g., fewer verbal and more aggressive solutions) than did boys with ODD. Matthys, Cuperus, and Van Engeland (1999) examined five groups of 7- to 12-year-old boys including those clinically diagnosed with ODD and CD or both, ADHD alone, and ADHD and ODD/CD. Two comparison groups were also studied: clinic boys with internalizing disorders and boys without psychiatric disorders. Social information processing patterns were elicited from videotaped vignettes of common problematic social situations. Relative to the comparison groups, boys from all of the clinic-referred groups encoded fewer relevant social cues and generated fewer responses to the hypothetical problems; b oth groups of boys with ODD endorsed a greater number of aggressive responses to the problems and stated more confidence in enacting these aggressive responses than did boys in the comparison groups. These differences remained after controlling for differences in boys' verbal intelligence. In one of the only studies of preschool-aged children, Webster-Stratton and Lindsay (1999) examined the Crick and Dodge (1994) model in a sample of 4- to 7-year-olds with and without ODD and CD or both. Consistent with previous studies, clinic group children generated significantly fewer positive solutions and had a lower ratio of positive to negative problem-solving strategies than did comparison group children. Clinic group children were also more likely to make hostile attributions of the actors' intentions in the stories.

These studies have focused on concurrent analyses, and there is little information about the prospective validity of social information processes in relation to the future outcomes of clinic-referred children (e.g., clinical diagnosis or caregiver's reports of behavior problem severity). This is an especially salient issue for disruptive preschool children, among whom a sizeable minority are expected to show diminishing behavior problems over time (Campbell, 1990). It is therefore important to determine whether variables assessed at the time of initial referral and assessment hold predictive value. Clinic-referred preschoolers inclined to formulate aggressive solutions to social conflicts and to interpret ambiguous social behavior as hostile or rejecting might be the most likely to show continuing diagnosis and high levels of behavior problems in the years following their initial diagnosis.

THE CURRENT STUDY

This study examined the social--cognitive characteristics of preschool boys first referred to an outpatient child psychiatry clinic for evaluation of disruptive behavior problems. Clinic boys were diagnosed by a team of mental

health professionals and retained in the study if they had a primary diagnosis of ODD, with or without other Axis I diagnoses. A comparison group of nonproblem peers matched on relevant social and demographic variables was also studied. The social--cognitive processes of boys in both groups were assessed at three time points across a 2-year period. The study was designed to address three questions:

1. Do four of the social information-processing steps delineated by Crick and Dodge (1994)--encoding, attribution, problem solving, and response evaluation--distinguish clinic-referred preschool boys from demographically matched peers? We anticipated that clinic group boys would be more likely than nondisruptive peers to make hostile attributions of ambiguous social behaviors and to report aggressive solutions to social dilemmas. Clinic group boys were also expected to evaluate aggressive actions more positively than were comparison group boys, and to encode social cues less accurately.

One factor that may influence group differences in social cognition between disruptive and typical children is the relation between general verbal ability and the ability to participate effectively in social--cognitive assessment tasks. Children and adolescents with disruptive behavior problems have been shown in a number of studies to perform more poorly than matched comparison children on any task that is administered orally, requires language mediation, and calls for a verbal response or a combination of these (e.g., Lynam, Moffitt, & Stouthamer-Loeber, 1993; Moffitt & Lynam, 1995). There is emerging evidence that this distinction applies to preschool-aged children as well (Cohen, Davine, Horodezky, Lipsett, & Isaacson, 1993; Heller, Baker, Henker, & Hinshaw, 1996; Speltz, DeKlyen, Calderon, Greenberg, & Fisher, 1999). Little is known about the extent to which fundamental verbal skills (e.g., comprehension, expressive vocabulary, verbal reasoning) might affect preschoolers' performance of social--cognitiv e tasks. The expected poor performance of disruptive preschoolers on these tasks may simply reflect a general limitation in their verbal abilities, rather than a more specific deficit in how they conceptualize, plan, or evaluate social behavior. We therefore examined the relations between social--cognitive variables and three aspects of verbal ability: verbal intelligence, expressive language, and receptive language. Verbal measures showing significant correlation with social-cognitive variables were covaried in subsequent analyses of group differences.

2. What is the relation between preschoolers' social--cognitive processes and their observed problem behaviors? We addressed two aspects of this question: First, we examined the concurrent associations between the four social--cognitive steps (discussed earlier) and parent and teacher reports of externalizing and internalizing behavior problems in the sample as a whole (clinic and comparison groups combined). Second, for clinic boys only, we tested the significance of two social--cognitive variables assessed in the last 2 years of our study (aggressive problem solving and hostile attributions) to predict DSM-III-R diagnosis and caregivers' reports of behavior problems. It was anticipated that clinic boys reporting aggressive problem solutions and hostile attributions would be more likely to show persisting diagnosis and more severe problem behavior 1 and 2 years following an initial clinic assessment.

3. Are the social--cognitive processes of clinic boys with ODD influenced by comorbidity with ADHD? We wanted to determine whether comorbid ADHD in a sample of boys selected for ODD would add further liability to social--cognitive functioning assessed concurrently. It was anticipated that ODD and ODD/ADHD clinic boys would show similar deficits in the Crick and Dodge (1994) social--cognitive steps, with the exception of encoding, which was hypothesized to reveal a greater deficit among boys with ODD and ADHD.

METHOD

Participants

Participants were 172 preschool-aged boys and their families; 92 of the boys were consecutive referrals to an early childhood behavior problem clinic and 80 were normally developing boys without clinically significant disruptive behavior. The boys' ages ranged from 47 to 68 months (M = 57 months) when they were first assessed (Time 1). Participants were evaluated again 1 year (Time 2) and 2 years (Time 3) following the initial evaluation. Mean ages were 69 and 82 months for Time 2 and Time 3, respectively. Mean family socioeconomic level was 2.3 on the 5-point Hollingshead scale (1 = professional, 5 = unemployed). Mothers identified 141 children (82%) as European American; the remainder were identified as African American (3%) or children of mixed ethnicity (15%). Children of mixed ethnicity typically had a European American parent and an Asian American, African American, or Native American partner. In 117 families, boys lived with both biological parents; 55 families were headed by single mothers. Four clin ic group boys were excluded from the analyses reported in this paper, because of missing or incomplete social-cognitive data at all three time points. See Table I for a description of the remaining 88 clinic group and 80 comparison group participants (N= 168).

Recruitment

Clinic Families

The Preschool Families Clinic was established as part of an outpatient child psychiatry clinic within an urban children's medical center. Among the clinic boys, 67% were referred by a health care professional (physician, psychologist, nurse), 23% by an educator (preschool teacher or daycare staff), and 10% were self (parent) referrals. This group differed from boys seeking services from similar clinics in that families received $50 reimbursement for completion of the assessment; they agreed to be available for follow-up assessments 1 and 2 years later; and parents were asked questions about their sons' behavior, out-of-home care, and home environment, in order to determine eligibility. Criteria for clinic group membership in the larger study included: (1) unsolicited referral for disruptive, oppositional aggressive behavior, or a combination of these; (2) age approximately between 4 and 5.5 years; (3) a T score of 65 or above on the parent version of the Child Behavior Checklist Aggression scale (CBCL; Achen bach, 1991); (4) a primary diagnosis of ODD, with or without other Axis I disorders (e.g., anxiety, encopresis/enuresis); (5) participation in out-of-home daycare or preschool (in order to obtain the reports of a nonfamilial, adult caregiver); (6) in cases of single parent families, that the mother was the primary caregiver in residence; and (7) child did not have mental retardation, autism, previously identified speech and language disability, fetal alcohol syndrome, seizure disorder, head injury or other neurological disorder, or disabling medical condition (e.g. cerebral palsy). Only boys were included in the sample, for three reasons: (1) ODD occurs primarily among males; (2) girls with an early onset of disruptive behavior are less common than boys and therefore more difficult to recruits; and (3) research suggests that the dynamics leading to disorder in girls may differ from those for boys (Greenberg et al., 1993).

Comparison Families

Comparison families were recruited through posters placed in pediatricians' offices, daycare centers, public libraries, and social service facilities, inviting parents of preschool boys to participate in a "developmental study." A total of 820 potential comparison-group families responded. [4] After participating in a telephone screening interview, parents completed child behavior checklists and family information forms to confirm eligibility (e.g., absence of significant problems) and to permit case matching to enrolled clinic families. Eligible comparison families were then called, and a visit was scheduled. Comparison boys in the larger study were case-matched to clinic subjects by age, ethnicity, family structure, and socioeconomic status (SES). [5] they received a T score below 65 on the CBCL Aggression scale, were characterized by parents and teachers as typical, and received no clinical diagnoses. There were no differences between the clinic and comparison groups on any case-matching variable, number of children in the home, maternal age, or maternal employment status. The groups were significantly different on maternal education; mothers of comparison boys were more likely to have achieved a higher level of education than were the mothers of clinic-referred boys. However, correlations between maternal education level and social-cognitive variables were not significant (p [greater than] .05); the highest correlation obtained was less than .20. Maternal education, therefore, was not covaried in subsequent analyses of group differences in social cognition.

Measures

Family Information Form

This questionnaire was used to elicit information about family demographics, structure and composition, child developmental history, child care arrangements, family medical and psychiatric history, and previous professional contacts. DSM-III-R diagnoses were determined following a diagnostic interview, review of behavior checklist information from parents and teachers, and observations of boys during neuropsychological testing and parent--child interactions. (DSM-IV was not yet available.) A consensus diagnosis--based on all available information--was assigned during a team meeting of mental health professionals, including the interviewing clinician and at least one of the authors. The diagnostic interview used the Disruptive Behavior Disorder (DBD) portion of the parent version of the Diagnostic Interview Schedule for Children (DISC; Fisher, Wicks, Shaffer, Piacentini, & Lapkin, 1992). As the DISC was designed for children aged 6 years and older, the questions were modified in several ways to better accommo date preschool children: (1) references to activities or events specific to older children (e.g., homework) were eliminated; (2) parents were prompted to consider whether behaviors such as activity level and short attention span were excessive for the child's age; and (3) examples of child behaviors and context were revised to reflect age-appropriate possibilities (e.g., noncompliance when asked "to put blocks away"). Team decision rules required clear evidence for the functional impairment of symptoms (as in DSM-IV) and distinctions between normal and atypical development were emphasized (e.g., ADHD was evaluated in light of the expected high rate of motor activity and distractibility in 4-year-old boys). In assessing for CD, the symptom "often initiates physical fights" required a pattern of deliberate confrontation and initiation of physical fighting" the occasional shoving or hitting that occurs commonly in preschools was not counted, nor was the hitting of a parent or sibling during a tantrum. All other disorders were assessed by first asking parents screening questions taken from the parent version of the DISC and then, when indicated, following up by reviewing all relevant DSM-III-R criteria. A random selection of 30 cases at Time 1, 27 cases at Time 2, and 24 cases at Time 3 were coded by a second clinician, who watched videotapes of the original interviews. Agreement on diagnoses was 96% (K = .87) at Time 1, 99% (K = .93) at Time 2, and 96% (K = .82) at Time 3; on number of DBD symptoms present, exact agreement was 75% at Time 1, 78% at Time 2, and 75% at Time 3.

At Time 1, the following diagnoses were given to the 88 clinic boys participating in this study: 33 had ODD alone; 24 had comorbid ODD and ADHD; 13 had ODD and another Axis I disorder (e.g., an anxiety disorder); 18 were diagnosed with ODD, ADHD, and another Axis 1 disorder. Only three clinic cases were found at Time 1 to meet full criteria for CD (although the hierarchical criteria of DSM-III-R exclude an ODD diagnosis when CD is present, we recorded both for descriptive purposes).

Speltz et al. (1999) examined the diagnostic status of the full sample of 92 clinic boys 2 years later (Time 3). Among the 79 clinic boys returning for the Time 3 assessment, 76% met criteria for one or more DSM-III-R Axis I disorders. Among those receiving a follow-up diagnosis (n = 60), 30% had ODD alone, 17% had ADHD alone, and 27% had ODD comorbid only with ADHD. Only two cases met criteria for CD, one with comorbid ADHD. There were eight cases of anxiety disorder; all occurred in the presence of ODD, ADHD, or combined ODD/ADHD. Among the five cases available at follow-up with an anxiety disorder at intake, only two accounted for the eight anxiety disorder cases in the sample at follow-up (i.e., six new cases of anxiety disorder were found). Six boys had a mood disorder at follow-up; of these, all but one had an anxiety disorder, ODD, ADHD, and combined ODD/ADHD, or a combination of these.

Child Behavior Checklist (CBCL; Achenbach, 1991)

The CBCL was used to quantify maternal observations of their boys' problem behavior. The CBCL includes 113 problem behavior descriptions rated by parents on a 3-point scale (0 = not true, 2 = very true). It provides standard scores for Total Problems, Internalizing and Externalizing Problems, and several subscales, including Aggression. The CBCL is a well-standardized measure with extensive reliability and validity data (see Achenbach, 1991).

Conners' Teacher Rating Scale (CTRS-39; Conners, 1989)

The CTRS-39 is a 39-item symptom list completed by a child's teacher or daycare provider. It yields six factor scores, including ratings for Conduct Problems and Anxious--Passive Behavior. Adequate test-retest stability (.72-.9 1), internal consistency ([alpha] = .6 1-.95), validity data, and norms have been established in a number of studies (Conners, 1989). Each participating boy's primary teacher or daycare provider was asked to complete the CTRS-39 at each time point.

Measures of Social Cognition

The changing ages and capabilities of the children in our study necessitated the selection of three different but similar measures of social cognition: (1) The Challenging Situations Task (CST; Denham, Bouril, & Belouad, 1994); (2) Cartoon Stories (Dodge, 1980); and (3) Videotaped Stories (VTS; Dodge & Coie, 1987). All three measures employed the presentation of hypothetical social situations ("stories"), to which children were asked to respond through standardized questions. A more demanding, visually complex, and age-relevant measure was used at each advancing age point; for example, the CST, which uses simple line drawings and "what" questions, was given at Time 1, and the VTS, which features videotaped scenarios of school-aged children and "why" questions, was given at Time 3. Only the VTS allowed for the assessment of four of the steps of social information processing described in the Crick and Dodge (1994) model (e.g., encoding, interpretation, problem solving, and evaluation). Among the several differ ent stories included in each measure, we selected stories that (a) depicted a provocative incident and a negative outcome for the participating child, and (b) featured ambiguous social cues (i.e., the cues did not clearly indicate whether a child deliberately caused the negative outcome). Provocations in which the cues are ambiguous have been found to be the most difficult for children with social problems (Dodge, McClaskey, & Feldman, 1985). Table II summarizes the social--cognitive variables that were assessed at Times 1, 2, and 3.

Two stories from the CST (Time 1) met criteria for provocative and ambiguous content. Each story was read to the child as he looked at associated line drawings. After each story, the child was asked, "What would you do if (describe scenario)?" The examiner then offered the child four different problem-solving solutions from which to choose, each reflecting one of the following types of responses: aggressive, dependent on authority, avoidant, or cooperative. For example, after reading a story in which one child was observed to hit another, the examiner asked, "What would you do if someone hit you? Would you: hit him back?, tell the teacher?, go play somewhere else?, or tell him, 'That's not nice'?" A percentage score was used to summarize children's responses to both stories (i.e., percentage of stories to which an aggressive response was given: 0, 50, or 100%).

The Cartoon Stories task administered at Time 2 included four stories in which one child provoked another. Two measures were taken from these stories: aggressive problem solving and hostile attributions. Stories were read to children as they looked at associated cartoon figures. Children's attributions about the intentions of story characters were assessed immediately after the story was read by asking, "Why do you think (describe provocation)?" The examiner categorized responses as "accidental" or "hostile." The child's score was the percentage of stories to which a hostile attribution was given (0,25,50, or 100%).

Interrater reliability was assessed on 25% of cases coded by the first author. Percentage agreement among coders was 92% ([kappa] = .81, range = .64-.93). Aggressive problem solutions were elicited by asking children, "What would you do about (describe provocation)?" immediately following the attribution question. Responses were assigned to one of six categories ("don't know," "nothing," "ask why or ask again," "command," "adult punish," or "retaliate"), and the proportion of stories to which the child indicated that he would retaliate or would have the teacher punish the other child were recorded. Interrater reliability was assessed as just described; percentage agreement was 82% (mean [kappa] = .76, range = .66-.84).

The VTS administered at Time 3 included five stories meeting the criteria outlined earlier (negative out-come and ambiguous social cues). Four measures of social cognition were taken: encoding of relevant information, hostile attributions, aggressive problem solving, and evaluation of the desirability of aggressive solutions. To assess encoding, the examiner asked the child to describe what happened in the story. Responses were rated on a scale from 1 (fully relevant) to 3 (fully irrelevant). Scores were averaged across the five stories to yield a single score for encoding. To assess VTS attributions, children were asked, "Why did the other kid(s) act this way?" The examiner categorized responses as "hostile," "benign," "ambiguous," or "don't know." If the child did not respond or made an uncodeable response, he was prompted once by asking, "Was the other kid(s) being mean or not being mean?" Scores were the proportion of stories to which the child initially gave a hostile response or, following a prompt, en dorsed the "being mean" option. VTS aggressive problem solving was determined by asking, "What would you do if this happened to you?" immediately following the attribution questions. Children were allowed one response each. Children's responses were assigned by the examiner to 1 of 14 categories: "verbal aggression," "physical aggression," "inept," "ask," "tell," "trade/bribe/bargain," "share/take turns," "authority," "general nice," "general assertive," "don't know/irrelevant," "authority-intervene," "do nothing," or "other." Scores were the proportion of stories to which children offered verbally or physically aggressive problem solutions.

After asking the preceding questions (about encoding, attributions, and problem solving) children's VTS evaluations of aggressiveness were assessed by showing them three different conclusions to each story: aggressive, passive, and assertive responses to the provocateur by the child who in the story was provoked. The videotape was stopped after each version and children were asked, "Do you think that's a good thing or a bad thing to say or do?" Once the child made this dichotomous choice (good or bad), the examiner queried further by asking "was it very good (or bad) or a little good (or bad)"? The examiner summarized the child's responses by assigning them to 1 of 4 categories, which represented a scale ranging from 1 (very bad) to 4 (very good). Scores taken from the aggressive versions of the five stories were averaged to yield an overall evaluation of aggression score.

Interrater reliability was assessed in the same manner: percentage agreement was 77% for encoding ([kappa] = .59, range = .55-.72), 97% for hostile attributions ([kappa] = .80, range = .76-.84), and 80% for aggressive problem solutions ([kappa] = .73, range = .57-.87).

Measures of Verbal Ability

The Wechsler Preschool and Primary Scale of Intelligence-Revised (WPPSI-R; Wechsler, 1989) was used to estimate children's verbal intelligence. Two sub-tests from the WPPSI-R Verbal scale (Comprehension and Arithmetic) were used to create a verbal IQ score, constituting one of the brief forms of the WPPSI-R recommended by Sattler (1992). Expressive vocabulary was estimated by giving the Expressive One-Word Picture Vocabulary Test-Revised (EOWPVT-R; Gardner, 1990), which is a picture-naming task. It has established norms and good reliability data; for 4-year-olds, Gardner reported a split-half reliability of .89. It correlates significantly but moderately with other measures of verbal ability (e.g., the Peabody Picture Vocabulary Test-Revised and WPPSI-R Vocabulary subtest). Receptive vocabulary was assessed with the Peabody Picture Vocabulary Test-Revised (PPVT-R; Dunn & Dunn, 1981), a well-established measure with published norms, reliability, and validity data; the manual indicates reliabilities of .70 (sp lit-half) and .78 (test-retest) for 4-year-olds. In the PPVT-R, a word is read aloud and the subject is asked to point to the picture (from an array of four) it represents.

Procedure

Time 1

Once eligible participants were identified, questionnaire packets were sent and a clinic visit was scheduled. During the first visit, parents participated in the DISC interview and children were administered the CST, the verbal ability tests, and other measures not reported here. Teachers and daycare providers were given the CTRS-39 to complete shortly after the clinic visit. On the basis of the information collected during the Time 1 assessment, parents were given diagnostic feedback and recommendations for treatment. The Preschool Families Clinic offered only evaluations and referrals to other services; no treatment was provided. However, families' subsequent treatment participation was monitored.

Time 2

Clinic visits were scheduled 1 year later, at which time parents, teachers, and child care providers again completed behavior checklists, the DISC was readministered to parents, and the Cartoon Stories were administered to children.

Time 3

Clinic visits were again scheduled, at which time parents, teachers, and child care providers completed behavior checklists, the DISC was readministered to parents, and the VTS was administered to children.

Preliminary Analyses

Parent and child participation in mental health treatment and educational services for behavior problems between the Time 1 and Time 2 visits were assessed during a telephone interview with mothers just prior to the Time 2 visit. Among returning clinic families, 55% (n = 44) had participated in one or more interventions in the year following the Time 1 evaluation and 45% (n = 36) had participated in the year following the Time 2 evaluation. In order to determine whether participation in treatment was significantly related to any of the social-cognitive variables, a one-way (treated vs. untreated) multivariate analysis of variance (MANOVA) was performed on seven social--cognitive variables that were measured across the 3 years of this study. No significant effect of treatment was found on these variables. Thus, participation in treatment was not considered further in this study.

RESULTS

In order to maximize power for cross-sectional group comparisons, all boys with complete social--cognitive data at every time point were included, even those who were missing data at other time points due to attrition, temporary drop-out (e.g., families not participating at Time 2 returning at Time 3), or incomplete testing (child too disruptive or fatigued). The following numbers of clinic and comparison group cases were analyzed at each time point, respectively: Time 1, 84 and 80; Time 2, 79 and 75; Time 3, 71 and 67. Complete data were available at all three time points for 62 clinic and 64 comparison group cases.

Do Social--Cognitive Processes Distinguish Clinic and Comparison Group Boys?

Table III shows the means and standard deviations of percentage and rating scores for all dependent variables by time point (percentage scores reflect the proportion of stories for which an aggressive or hostile response was given by the child). Table IV shows the number of clinic and comparison boys who generated an aggressive solution for one or more stories at each time point.

Table V shows the correlations between measures of verbal ability (WWPSI-R Verbal IQ, EOWPVT-R, and PPVT-R) and all social--cognitive variables for the entire sample. Analyses of covariance (ANCOVA) were used to evaluate group differences on the social--cognitive variables showing significant correlation with one or more of the verbal measures in Table V. None of the verbal ability covariates was significant (group differences are therefore reported without further consideration of the verbal ability measures).

Analysis of variance (ANOVA) of Time 1 aggressive problem-solving scores revealed that clinic-referred boys were more likely to generate aggressive solutions at Time 1 than were comparison boys, F(1, 163) = 10.80, p [less than] .005. A one-way multivariate analysis of variance (MANOVA) conducted for Time 2 variables indicated that clinic group status was significantly related to children's social--cognitive processes, F(2, 148) = 7.57, p [less than] .005. Follow-up univariate tests indicated that clinic-referred boys were more likely to generate aggressive solutions to problems, F(1, 149) = 13.54, p [less than] .001, than were comparison boys. A one-way MANOVA conducted at Time 3 also indicated that clinic group status was significantly related to children's social cognitive processes, F(4, 129) = 7.57, p [less than] .01. Clinic-referred boys were more likely to generate aggressive solutions to problems, F(1, 132) = 7.83, p [less than] .01, and they had lower ratings of encoded relevant social information th an had comparison boys, F(1, 132) = 6.22, p [less than].05.

What is the Relation Between Social--Cognitive Processes and Observed Problem Behavior?

Concurrent Relations

Table VI shows the bivariate correlations between two social--cognitive variables (aggressive solutions and hostile attributions) and concurrent CBCL and CTRS-39 scores for the clinic group. Statistically significant but modest correlations (0.17--0.26, all ps [less than] .05) were found between aggressive problem solutions and caregiver reports of behavior problems (6 of 12 correlations involving aggressive problem solutions were significant). With the exception of Time 3 parent CBCL externalizing reports and encoding, 0.19, p [less than] .05, no other social-cognitive variable showed significant association with observed problem behavior.

Predictive Relations

The predictive significance of three social cognitive variables (Time 1 and Time 2 aggressive problem solutions and Time 2 hostile attributions) was examined in relation to five outcome measures taken at Time 3: CBCL externalizing and internalizing T scores, CTRS-39 Conduct Problem and Anxious-Passive sub-scale T scores, and an 8-category variable representing DSM-III-R diagnostic status at Time 3 (including "no diagnosis" and all single and comorbid presentations of ODD, ADHD, and "other diagnoses").

The Time 3 CBCL and CTRS-39 scores of clinic boys giving aggressive or hostile responses to stories at Times 1 and 2 were compared to those of clinic boys not giving such responses. No significant differences were indicated by t-test analyses (all ps [greater than] .05). Chi-square analyses were conducted to test the association between Time 3 DSM-III-R status and each of the three social-cognitive predictors (presence/absence of one or more aggressive or hostile stories at Times 1 and 2). Presence/absence of hostile attributions at Time 2 was significantly associated with Time 3 diagnosis, [x.sup.2] (7) = 15.472, p [less than] .05: clinic group boys without hostile attributions at Time 2 were more likely to be free of diagnosis or to have ODD alone at Time 3 than were clinic group boys reporting one or more hostile attributions.

Are the Social-Cognitive Processes Influenced by Comorbidity With ADHD?

Clinic boys with diagnoses at Time 1 of ODD alone (n = 46) were compared on all seven of the social cognitive variables shown in Table II with boys having comorbid ODD and ADHD (n = 42). No significant differences between these groups were revealed by MANOVA analysis. To test the a priori hypothesis that boys with ODD/ADHD would encode fewer relevant social cues than would boys with ODD alone, the univariate test of this difference was examined. This hypothesis was not confirmed, F(1, 60) = 2.64, p [greater than] .10.

DISCUSSION

The primary goal of this study was to determine whether four of the social information-processing steps delineated by Crick and Dodge (1994)--encoding, attribution, problem solving, and response evaluation--would distinguish clinic-referred preschool boys with ODD (with and without ADHD) from typical peers without significant behavior problems. The findings most clearly support an association between problem-solving and clinic referral. Clinic group boys were more likely to provide an aggressive solution to hypothetical, peer-oriented social problems than were comparison group boys. This effect was indicated by both the average number of stories to which boys provided aggressive "endings" and by the number of boys in each group who gave an aggressive response to a story. These findings are consistent with the Matthys et al. (1999) study in which groups containing school-aged boys with ODD or CD, with and without ADHD, endorsed a greater number of aggressive responses to hypothetical problems than did boys in two comparison groups.

The discrepancy between groups in aggressive solutions was remarkably stable across the 2-year duration of this study. At all time points, about twice as many clinic boys as comparison group boys generated aggressive solutions to the story tasks (see Table V). The consistency of this finding across multiple assessments, multiple stories within assessments, and different story presentation methods (e.g., line drawings, cartoons, and videotaped stories) suggest that there is robust association during this age period between clinic referral and boys' inclination to think about aggressive solutions to peer social problems. It is less clear whether there is, at this age, a relation between boys' reports of aggressive solutions and their observed social behavior with peers. The present study is limited in its ability to address this issue, as the parent and teacher measures of child problem behavior that were used (i.e., CTRS-39, CBCL) are sparse in their inclusion of peer-explicit items. Moreover, items pertainin g to aggression often do not specify the recipient (adult or peer, as in the CBCL's "hits others"). Nevertheless, we examined the correlations between aggressive problem solving and behavior checklist scores (Table IV) and found that concurrent relations between aggressive problem solutions and observed problem behaviors were modest. No correlation exceeded .30. Webster-Stratton and Lindsay (1999) reported correlations of a similar nature and magnitude. In post hoc analyses, we also examined correlations between aggressive problem solving and specific items from the CTRS-39 that pertain to peer relationship difficulties (e.g., CTRS-39 items "disturbs other children" and "does not get along with peers"). Although several were statistically significant, none exceeded .25 and, when examining these relations separately for clinic and comparison groups, similar results were obtained. This suggests that for both clinic and nonclinic preschool boys there is a modest relation between child reports of aggressive respo nses to hypothetical situations and caregivers' reports of observed aggressive behavior, although clinic group boys clearly show higher levels of both.

No other step measured in the Crick and Dodge (1994) model was a compelling discriminator of clinic and nonclinic boys. Although there was a statistically significant group difference in encoding at Time 3, the narrow range of the encoding scale and the magnitude of this difference (about 0.6 SD) limit the clinical meaning of this finding. Future studies would benefit from the use of a more sensitive story-recall task than the one used here, which relied on a single item ("tell me what happened in the story") and a 3-point response scale. Regarding the attribution step, clinic boys were no more likely to attribute hostile intentions to ambiguous peer behaviors than were comparison group boys. This was evident at both Times 2 and 3. Unlike aggressive problem solving, there were relatively high rates of hostile story interpretations among the boys in this sample (see Table V). Nearly half of all the boys in the study attributed "hostile" intent to ambiguous peer behaviors, regardless of their referral status. This finding contrasts with those reported by Webster-Stratton and Lindsay (1999), who found that the preschoolers in their clinic group (average age about 5.5 years) were somewhat more likely than comparison children to assign hostile intentions to the ambiguous, provocative behaviors of story figures. This inconsistency might be due to gender (27% of the Webster-Stratton and Lindsay sample were girls) and the attribution task itself, or both. In the present study, children's unstructured responses to open-ended questions were coded for attribution, whereas the Webster-Stratton and Lindsay procedures required the child to select one of two alternative explanations for what happened in the story (intentional vs. accidental provocation). The "forced choice" method may be more sensitive to clinic-comparison group differences at this age than the open-ended format used herein, a possibility that requires further study.

With respect to response evaluations, almost all the boys in both the clinic and comparison groups said that aggressive actions were "very bad" or "bad." The transparency of this question may have made it particularly vulnerable to social desirability effects (e.g., not wanting to tell the examiner that aggression is a good solution).

Effect of ADHD Comorbidity

As the peer relationship problems of school-aged children with ADHD are well known (Erhardt & Hinshaw, 1994; Neel, Jenkins, & Meadows, 1990), we wanted to determine whether preschool boys meeting criteria for both ODD and ADHD would show a more problematic level of social-information processing than boys with ODD alone. As encoding of information would, in theory, benefit from better attention and vigilance, we anticipated that the ODD/ADHD group would perform worse on this measure. However, no effects were attributable to the presence of ADHD in this clinic sample, and it may simply be that the additional diagnosis of ADHD in preschool boys selected for ODD does not provide sufficient diagnostic contrast to affect the measures used. Similarly, Matthys et al. (1999) found little diagnostic distinctiveness on social-- cognitive variables in their school-aged sample of boys with various combinations of ODD, ADHD, and CD.

Predictive Significance of Social--Cognitive Tasks for Clinic Boys

Within the clinic group, social--cognitive task performance had little prognostic significance. Clinic boys distinguished by the presence or absence of aggressive problem solving (Time 1 and 2) or hostile attributions (Time 2) showed roughly equivalent levels of externalizing and internalizing behaviors 1 and/or 2 years later (Time 3). This was evident in both mothers' and teachers' behavior checklist reports. The lack of association between social cognition and later behavior problems may be due in part to constrained variance in behavior checklist scores within the clinic group, as well as the limitations noted earlier in instrumentation (e.g., inadequate measurement of peer interactions by the checklists we used). Future research with clinic-referred children of this age would be better served by using a peer-specific outcome measure (e.g., direct observations of peer interactions, sociometric measures).

Modest prediction to DSM-III-R diagnosis was offered by the attribution measure at Time 2. Clinic group boys reporting negative interpretations of peer behaviors were more likely than those reporting positive or neutral interpretations to receive a DSM diagnosis 1 year later. They were also the least likely to have a diagnosis of only ODD, as compared to various other diagnoses or co-morbid conditions; this may indicate that the tendency to make hostile attributions is associated with a more severe or complex diagnostic pattern. However, this finding is extremely tentative and requires replication.

Limitations of Different Measures Across Time

As noted earlier, three different measures of social cognition were used to accommodate the age range and capabilities of the children in this longitudinal study. We anticipated that the content and visual/verbal processing demands of the Videotaped Stories Task (which contains scenarios featuring older elementary school children) would have been inappropriate for the youngest boys at Time 1. Conversely, the Challenging Situations Task was felt to be insufficiently relevant and challenging for most boys at Time 3, when many were first graders. Although all three measures employed similar formats, there were several procedural variations (e.g., in response options) that precluded analyses of stability in the constructs assessed. This is a significant problem as, to our knowledge, there are no data pertaining to the continuity of individual preschoolers' responses to repeated social-cognitive assessments over the short-term (test-retest reliability) or long-term (stability). Future longitudinal studies need to address this issue by developing a single social-cognitive measure that can be used to span the preschool to grade school years.

Limitations of Generalizability

The nature of the present sample is likely to have influenced the results in a number of ways. Potential group differences may have been diminished by screening out youngsters with previously identified mental retardation and language, neurological, or certain medical disorders. Such children constitute a small but important segment of the population with early onset disruptive behavior disorders. On the other hand, clinic referral selects for more severe and complicated cases than would an epidemiological study, thus possibly inflating group differences. Generalization is limited by the relatively low-risk (i.e. middle-class, with low rates of domestic violence and divorce) nature of the sample. These findings may also not generalize to girls with behavior problems.

Conclusions

The results of this study suggest that aggressive problem solving is a robust discriminator of preschool boys with and without clinically significant disruptive behavior. During multiple assessments across a 2-year age range, clinic group boys were twice as likely as nondisruptive boys to generate aggressive solutions to hypothetical social dilemmas. Boys' responses to this largely verbal task were only modestly correlated with general verbal abilities, suggesting that social-cognitive story tasks are, at this age, capturing relatively specific mental processes related to the conceptualization and planning of social behavior than are tests of general language and verbal reasoning. However, as verbal skills were assessed only at Time 1, this conclusion is tempered by potential instability in verbal development overtime. A more definitive examination of this issue in future longitudinal studies will require assessments of both verbal ability and social cognition at every time point.

The study found only modest associations between social-cognitive measures and caregivers' reports of concurrent and future behavior problem severity. This finding may be due to inadequate peer information in the measures we used to quantify caregivers' observations of child behavior. Alternatively, it may be due to the magnitude of the association between observed peer interactions and social-cognitive task performance as measured by this task. Dodge and Somberg (1987) found that social cognitions were less predictive of peer group behavior under laboratory conditions than under conditions designed to stimulate feelings of threat to the self (and therefore arousing affect). This may be a fruitful avenue for future research.

(1.) University of Washington School of Medicine, Seattle, Washington.

(2.) Children's Hospital and Regional Medical Center, Seattle, Washington.

(3.) Address all correspondence to Matthew Speltz, Department of Psychiatry and Behavioral Sciences, University of Washington, Box 359300-CL-08, Seattle, Washington 98105; e-mail: mspeltz@u.washington.edu.

(4.) Of the 820 families who responded, 250 did meet inclusion criteria, 101 exhibited behavior problems according to either parent or teacher reports, 140 voluntarily withdrew (primarily because of unwillingness to make the required time commitment), and 249 were not included because we were unable to case-match them with an enrolled clinic group family.

(5.) There were 12 eligible clinic families for whom exact matches could not be found because of statistically infrequent combinations of matching variables (e.g., it was difficult to find exact matches on all four matching variables when remarried parents had differing ethnic backgrounds). In order to maximize power, we included these clinic families in the present analyses if their sons had complete social--cognitive data at one or more time points. These cases did not differ from matched clinic cases on major dependent variables and their inclusion did not alter the equivalence of the clinic and comparison groups on matching variables (see Table I).

REFERENCES

Abikoff, H., Gittelman, R., & Klein, D. F. (1980). Classroom observation code for hyperactive children: A replication of validity. Journal of Consulting and Clinical Psychology, 48, 555-565.

Achenbach, T. (1991). Manual for the Child Behavior Checklists/4-18 and 1991 Profile. Burlington, VT: University of Vermont, Department of Psychiatry.

American Psychiatric Association. (1987). Diagnostic and statistical manual of mental disorders (3rd ed., Rev.). Washington, DC: Author.

American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders (4th ed.). Washington, DC: Author.

Boulanger, M. D., & Langevin, C. (1992). Direct observation of playgroup therapy for social skills deficits. Journal of Child and Adolescent Group Therapy, 2, 227-236.

Campbell, S. B. (1990). Behavior problems in preschool children. New York: Guilford Press.

Campbell, S. B. (1995). Behavior problems in preschool children: A review of recent research. Journal of Child Psychology and Psychiatry, 36, 113-149.

Cohen, N. J., Davine, M., Horodezky, N., Lipsett, L., & Isaacson, L. (1993). Unsuspected language impairment in psychiatrically disturbed children: Prevalence and language and behavioral characteristics. Journal of the American Academy of Child and Adolescent Psychiatry, 32, 595-603.

Conners, C. K. (1989). Manual for Conners' Rating Scale. North Tonawanda, NY: Multi-Health Systems.

Coon, H., Carey, G., Corley, R., & Fulker, D. (1992). Identifying children in the Colorado Adoption Project at risk for conduct disorder. Journal of the American Academy of Child and Adolescent Psychiatry, 31, 503-511.

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.

Crick, N. R., & Ladd, G. W. (1990). Children's perceptions of the outcomes of social strategies: Do the ends justify being mean? Developmental Psychology, 26, 612-170.

Denham, S. A., Bouril, B., & Belouad, F. (1994). Preschoolers' affect and cognition about challenging peer situations. Child Study Journal, 24, 1-21.

Dodge, K. A. (1986). A social information processing model of social competence in children. In M. Perlmutter (Ed.), Minnesota symposium on child psychology (Vol. 18, pp. 77-125). Hillsdale, NJ: Erlbaum.

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

Dodge, K. A., & Coie, J. D. (1987). Social-information-processing factors in reactive and proactive aggression in children's peer groups. Journal of Personality and Social Psychology, 53, 1146-1158.

Dodge, K. A., McClaskey, C. L., & Feldman, E. (1985). Situational approach to the assessment of social competence in children. Journal of Consulting and Clinical Psychology, 53, 344-353.

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

Dodge, K. A., & Tomlin, A. (1987). Cue utilization as a mechanism of attributional bias in aggressive children. Social Cognition, 5, 280-300.

Dunn, L. M., & Dunn, L. M. (1981). Peabody Picture Vocabulary Test, Revised. Circle Pines, MN: American Guidance Service.

Dunn, S. E., Lochman, J. E., & Colder, C. R. (1997). Social problem-solving skills in boys with conduct and oppositional defiant disorders. Aggressive Behavior, 23, 457-469.

Erhardt, D., & Hinshaw, S. P. (1994). Initial sociometric impressions of attention-deficit hyperactivity disorder and comparison boys: Predictions from social behaviors and from nonbehavioral variables. Journal of Consulting and Clinical Psychology, 62, 833-842.

Fisher, P., Wicks, J., Shaffer, D., Piacentini, J, & Lapkin, J. (1992). Diagnostic Interview Schedule for Children--Second Edition (DISC 2.3). Washington, DC: National Institute for Mental Health.

Gardner, M. F. (1990). Expressive One-Word Picture Vocabulary Test--Revised, Novato, CA: Academic Therapy.

Greenberg, M. T., Speltz, M. L., & DeKlyen, M. (1993). The role of attachment in the early development of disruptive behavior problems. Development and Psychopathology, 5, 191-213.

Heller, T. L., Baker, B. L., Henker, B., & Hinshaw, S. P. (1996). Externalizing behavior and cognitive functioning from preschool to first grade: Stability and predictors. Journal of Clinical Child Psychology, 25, 376-387.

Ladd, G. W., & Oden, S. L. (1979). The relationship between peer acceptance and children's ideas about helpfulness. Child Development, 50, 402-408.

Loeber, R. (1990). Development and risk factors of juvenile antisocial behavior and delinquency. Clinical Psychology Review, 10, 1-41.

Loeber, R., Green, S., Keenan, K., & Lahey, B. B. (1996). Which boys will fare worse? Early predictors of the onset of conduct disorder in a six-year study. Journal of the American Academy of Child and Adolescent Psychiatry, 34, 499-509.

Lynam, D., Moffitt, T. E., & Stouthamer-Loeber, J. (1993). Explaining the relations between IQ and delinquency: Class, race, test motivation, school failure, or self-control? Journal of Abnormal Psychology, 102, 187-196.

Matthys, W., Van Loo, P., Pachen, V., De Vries, H., Van Hooff, J., & Van Engeland, H. (1995). Behavior of conduct disordered children in interaction with each other and with normal peers. Child Psychiatry and Human Development, 25, 183-195.

Matthys, W., Cuperus, J. M., & Van Engeland, H. (1999). Deficient social problem-solving in boys with ODD/CD, with ADHD, and with both disorders. Journal of the American Academy of Child and Adolescent Psychiatry, 38, 311-321.

Moffitt, T. E. (1993). The neuropsychology of conduct disorder. Development and Psychopathology, 5, 135-151.

Moffitt, T. E., & Lynam, D., Jr. (1995). The neuropsychology of conduct disorder and delinquency: Implications for understanding antisocial behavior. In D. Fowles, P. Sutker, & S. Goodman (Eds.), Psychopathology and antisocial behavior: A developmental perspective (pp. 233-262). New York: Springer.

Neel, R. S., Jenkins, Z. N., & Meadows, N. (1990). Social problem-solving behavior and aggression in young children: A descriptive observational study. Behavioral Disorders, 16, 39-51.

Perry, D. G., Perry, L. C., & Rasmussen, P. R. (1986). Aggressive children believe that aggression is easy to perform and leads to rewards. Child Development, 57, 700-711.

Pullis, M. (1991). Practical considerations of excluding conduct disordered students: An empirical analysis. Behavioral Disorders, 17, 9-22.

Rabiner, D., & Coie, J. (1989). Effect of expectancy inductions on rejected children's acceptance by unfamiliar peers. Developmental Psychology, 25, 450-457.

Renshaw, P. D., & Asher, S. R. (1983). Children's goals and strategies for social interaction. Merrill-Palmer Quarterly, 29, 353-374.

Rey, J. M. (1993). Oppositional defiant disorder. American Journal of Psychiatry, 150, 1769-1778.

Robins, L. (1991). Conduct disorder. Journal of Child Psychiatry and Psychology, 32, 193-212.

Rutter, M. (1988). Studies of psychosocial risk. New York: Cambridge Press.

Sattler, J. M. (1992). Assessment of children (3rd ed.). San Diego, CA: Author.

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

Speltz, M. L., DeKlyen, M., Calderon, R., Greenberg, M. T., & Fisher, P. A. (1999). Neuropsychological characteristics and test behaviors of boys with early onset conduct problems. Journal of Abnormal Psychology, 108, 315-325.

Speltz, M. L., McClellan, J., DeKlyen, M., & Jones, K. (1999). Preschool boys with oppositional defiant disorder: Clinical presentation and diagnostic change. Journal of the American Academy of Child and Adolescent Psychiatry, 38, 838-845.

Spivack, G., Platt, J. J., & Shure, M. B. (1976). The problem-solving approach to adjustment. San Francisco, CA: Jossey-Bass.

Thomas, A., & Chess, S. (1984). Genesis and evolution of behavioral disorders: From infancy to early adult life. American Journal of Psychiatry, 14, 1-9.

Webster-Stratton, C., & Lindsay, D. W. (1999). Social competence and conduct problems in young children: Issues in assessment. Journal of Clinical Child Psychology, 28, 25-43.

Wechsler, D. (1989). Wechsler Preschool and Primary Scale of Intelligence-Revised (WPPSI-R). San Antonio, TX: The Psychological Corporation.

(a.)1 = professional, 5 = unemployed.

(b.)1 = [less than]8 years education, 7 graduate school.

(c.)Sum of standard scores (M = 10; SD = 3) for two WPPSI-R verbal subscales (Comprehension and Arithmetic).

(*.)p [less than] .001.

(a.)Proportion of stories for which an aggressive solution was given.

(b.)Proportion of stories for which a hostile attribution was given.

(*.)Group difference, p [less than] .05.

(**.)Group difference, p [less than] .01.

COPYRIGHT 2001 Plenum Publishing Corporation
COPYRIGHT 2001 Gale Group