The predictive validity of five behavioral screening strategies was studied by Pennsylvania State University researchers Kelly S. Flanagan, Karen L. Bierman and Chi-Ming Kam. Flanagan et al. believe that effective and reliable identification of at-risk children is critical to the success of prevention and early-intervention programs. Screening that assesses young children’s aggressive behavior, hyperactive-inattentiveness and low social skills identifies those at increased risk for negative social, academic and behavioral outcomes in third grade.
Research suggests that there are poor outcomes for children who exhibit more than one problem behavior. Aggressive-inattentive children are more likely than aggressive-only or inattentive-only children to show reading delays, poor cognitive ability, poor peer relations or adolescent antisocial and addiction problems. These findings suggest that the effectiveness of early screening strategies might be increased by considering multi-problem profiles rather than focusing on single behaviors. Aggression, hyperactivity-inattention and pro-social skill deficits appear to be intertwined developmentally, and children with elevations in any two areas often show problems in the third. These researchers tried to find out the best screening strategy for identifying young children with the potential for lasting academic, social and behavioral problems.
The goals were to examine the prevalence of behavior problems in first grade and to compare the predictive validity of several screening strategies that use teachers’ ratings of the three behaviors in first grade and school adjustment in third grade. Analyses compared identifying at-risk children on the basis of a aggressive behavior alone, or screening for combinations of behavioral problems.
The instruments used in first grade included the Teacher Checklist, which is a six-item scale measuring aggression; the Attention Deficit Hyperactivity Disorder Checklist with eight hyperactive items; and six inattentive items, and a nine-item Social Health Profile, which includes positive behaviors (friendly, helpful) and appropriate social regulation (controls temper, appropriate expression of needs/feelings). The central goal was to examine the predictive risk associated wit different problem profiles in the first grade.
Outcomes of children’s school adjustment in third grade was measured by the Authority Acceptance scale of the Teacher Observation of Classroom Adaptation-Revised, which assesses a child’s behavioral adjustment with 10 aggressive-disruptive items. In addition, interviews were conducted individually with each child to assess social adjustment. The researchers compared peer ratings of each child’s likability and used language-arts grades as a measure of academic performance.
The three behavior problems were moderately related to each other, and each of these early behavior problems predicted later social, academic and behavior problems. First-grade aggression was the strongest predictor of later behavior problems. However, hyperactivity-inattention and poor social skills were stronger predictors of problems with peer relations and academic performance. All three first-grade behavior problems made unique contributions to the prediction of behavioral adjustment in third grade. Most of the children, however, had problems in more than one area, indicating that these three areas were interrelated. As a group, all the children exhibiting problems in third grade were significantly more disruptive and aggressive than the group who displayed no problems. Of particular importance were children who had been considered at-risk in first grade but showed no problems at third grade. The broadest assessment – screening all three behaviors – incorrectly identified more children as at-risk of behavioral problems. Screening for only one behavior such as aggression, however, missed children who were at-risk for academic problems.
Each of the three problem areas measured in first grade predicted later school difficulties. These teacher ratings were good predictors of later problems. Each of the first-grade behavior problems contributed in different ways to later problems. The presence of any one of the three behavior problems increased the risk for negative school outcomes, yet, there didn’t appear to be much cumulative effect. The broader evaluation did correctly identify more children who were at risk for later problems, but it also incorrectly labeled more children at-risk who did not exhibit any problems in third-grade. When the aggression scale was used by itself, it underestimated the number of girls who exhibited later problems. The broader scale was a significantly more accurate indicator of problems in girls.
These researchers warn educators that the effectiveness of different screening systems may change with children at different ages or with different populations. But they conclude that these inexpensive, easy-to-administer teacher rating scales are valuable tools in identifying children who are at risk for future problems. The multi-problem broad-based screening that included children with problems in any of the three behavioral areas identified more children who had later problems but also incorrectly labeled children at-risk.
“Identifying At-Risk Children at School Entry: The Usefulness of Multibehavioral Problem Profiles”, Journal of Clinical Child and Adolescent Psychology, Volume 32, Number 3, September 2003, pp. 396-407.
Published in ERN October 2003 Volume 16 Number 7