A meta-analysis of the academic status of students with emotional/behavioral disturbance.

Emotional/behavioral disturbance (EBD) is characterized by a range of behaviors that adversely affect a child's academic performance and cannot be explained by other sensory or health impairments. Although research has clearly demonstrated that children and youth with EBD tend to exhibit high rates of problem behavior, research on the characteristics of their academic performance has been less clear. This article reports the results of a meta-analysis of the academic status of students with EBD. The overall effect size was -.64, which indicated that students with EBD had significant deficits in academic achievement. An examination of moderators (subject area, setting, and age) indicated that students with EBD performed at a significantly lower level than did students without disabilities across academic subjects and settings; greater deficits were not observed in older students with EBD (i.e., those more than 12 years old), as compared to younger students. Implications and areas for future research are discussed.

**********

Emotional/behavioral disturbance (EBD) is characterized by a range of problems that adversely affect a child's academic performance and cannot be explained by intellectual, sensory, or other health factors (Individuals with Disabilities Education Act, 1997). These behaviors can include a number of internalizing and externalizing characteristics that inhibit a child's ability to build and maintain successful social relationships with peers, teachers, and adults. Children and adolescents with EBD are overwhelmingly male, behaviorally disruptive, noncompliant, verbally abusive, and aggressive. Because their behaviors are so disruptive and irritating, these children often arouse negative feelings in others, alienating schoolmates and adults and ultimately robbing these children of the benefits of learning opportunities (Kauffman, 2001). Inevitably, these behaviors significantly impair a child's ability to succeed in school and in society.

Although research has unequivocally demonstrated that children and youth with EBD exhibit high rates of problem behavior, research on the characteristics of their academic performance remains uncertain (Kauffman, Cullinan, & Epstein, 1987). Data on identification, academic outcomes, graduation rates, absenteeism, employment status, and criminality among children and youth with EBD has suggested that educating them is a complex, confusing, and often daunting task for educators, related services personnel, and family members alike (Smith & Coutinho, 1997). More specifically, the preponderance of studies on academic performance have indicated that students with EBD perform 1 to 2 years below grade level (Trout, Nordness, Pierce, & Epstein, 2003), with academic difficulties emerging at an early age and persisting throughout their schooling (Coutinho, 1986; Rosenblatt & Rosenblatt, 1999; Wagner, 1995). This should be expected, because current criteria for identification of EBD require that students show a deficit in academic achievement. Compared to peers from other high-incidence disability groups, children with EBD evince lower reading and math scores, lower graduation rates, and higher rates of course failure and grade retention, and are less likely to attend postsecondary school (Kauffman, 2001; Wagner, 1995). In 1998-1999, 50.6% of students age 14 and older with EBD had dropped out of high school (U.S. Department of Education, 2001). Furthermore, it has been estimated that 70% of students with EBD will be arrested within 3 years of leaving school, continuing a pervasive pattern of failure that becomes difficult to correct (U.S. Department of Health and Human Services, 1999).

Low academic performance and maladaptive behavior patterns are highly related. Evidence suggests that a reciprocal relationship between school failure and social failure emerges early in life (Brier, 1995; Kauffman, 2001), although the mechanism of the relationship is unknown. The prevalence of academic difficulties among children with EBD is uncertain. Ruhl and Berlinghoff (1992) suggested that between 33% and 81% of children with behavioral disorders have academic difficulties. A causal relationship between behavioral problems and academic underachievement, however, has yet to be determined (Hinshaw, 1992). Nonetheless, researchers have demonstrated that academic failure is one of the most powerful predictors of problem behavior and social failure (Maguin & Loeber, 1996; Morrison & D'Incau, 1997). Conversely, researchers have also demonstrated that academic success is associated with a decrease in problem behavior (Gottfredson, Gottfredson, & Skroban, 1996).

To further our understanding of the academic status of students with EBD, researchers have traditionally relied on conventional narrative reviews of the literature (e.g., Epstein, Kinder, & Bursuck, 1989; Trout et al., 2003). These reviews have found that students with EBD perform significantly below norms on standardized achievement tests (e.g., Coutinho, 1986) and achieve lower scores in mathematics than in reading (e.g., Schroeder, 1965; Stone & Rowley, 1964) and that older children with EBD are farther behind their same-age peers than are younger children with EBD (Coutinho, 1986). Although these qualitative reviews have presented compelling evidence of the academic deficits experienced by children with EBD, they do not provide precise quantitative estimates of the magnitude of these deficits.

A quantitative review of the research on the academic performance of students with EBD compared to same-age, nondisabled peers or norm groups would extend our knowledge in this important area. Previous investigations have focused on grade-equivalent scores (e.g., Rosenblatt & Rosenblatt, 1999; Scruggs & Matropieri, 1986); however, these scores do not provide information on whether performance is within average limits, allow for normative comparisons, or allow for comparisons across studies. One statistical method for analyzing a body of research to compare group performance is meta-analysis. Meta-analytic reviews go beyond traditional narrative reviews in the sense that they are more systematic, more explicit, and rely on quantitative methods of analysis (Rosenthal, 1984). Because of these features, meta-analytic reviews are considered to provide more thorough, comprehensive, and precise summative evaluations that entail greater objectivity than narrative reviews (Rosenthal, 1984). The purpose of this study was to use a meta-analytic approach to quantitatively estimate the magnitude of difference in academic performance between students with EBD and their same-age, nondisabled peers or norm groups. The use of meta-analysis allows for a precise estimate of achievement difference, which allows for normative comparisons, provides a metric that is consistent across age levels, and allows for comparisons across studies. Moreover, meta-analysis is consistent with the American Psychological Association's guidelines calling for the use of effect sizes, which allows for an evaluation of the practical significance of differences (as opposed to statistical significance).