Disproportionality of English Learners with Emotional and/or Behavioral Disorders: A Comparative Meta-Analysis with English Learners with Learning Disabilities

* Disproportionality of English learners in special education has been a longstanding (Dunn, 1968) and ongoing (Ford, 2012) concern for almost half a century. The Council for Children with Behavioral Disorders (Skiba, Albrecht, & Losen, 2012) noted that, "despite court challenges, abundant research, and policy initiatives, racial and ethnic disproportionality [including English learners] persists as a critical and unresolved problem in the field" (p. 2). The topic has played a large role in the history of behavior disorders; however, the focus has often been on disproportionate representation of African Americans (Skiba, PoloniStaudinger, Callini, Simmons, & Feggins-Azziz, 2006).In contrast, a paucity of research has specifically examined disproportionality of English learners with emotional and/or behavioral disorders (EBD; Artiles, Bal, Trent, & Thorius, 2012). Although English learners currently represent approximately 21% of all students in U.S. schools (Aud et al., 2010), little is known about their representation in the EBD category. Donovan and Cross (2002) reported that English learners were generally underrepresented in the EBD category based on their synthesis of research conducted in the last century (prior to the year 2001). However, no study has examined patterns in the current century. To address this gap, we examined representation statistics of English learners with EBD in the post-No Child Left Behind (NCLB) education era (2002 to the present) to (a) identify the extent of disproportionality in the EBD category; (b) determine whether disproportionality was variable across educational contexts, be they individual schools, school districts, or state; and (c) compare the results for students with EBD with those for students with learning disabilities (LD).Disproportionality in Special EducationDisproportionality refers to "the extent to which membership in a given group affects the probability of being placed in a specific disability category" (Oswald, Coutinho, Best, & Singh, 1999, p. 198). For example, we would say that representation of English learners is disproportionately high if the probability of English learners being identified as having a LD is significantly greater than the expected probability given census statistics. Overall, findings from research examining disproportionality for culturally diverse students, including English learners, provide mixed signals and inconsistent findings. The picture is particularly murky when one combines results across many decades and many categories of disability and varying ethnic groups. Then, the picture appears mixed and often contradictory (Chinn & Hughes, 1987; Hopstock & Stephenson, 2003; Hosp & Reschly, 2003; Oswald et al., 1999; Sullivan, 2011).Two National Research Council (NRC) reports on disproportionality, separated by two decades (Donovan & Cross, 2002; Heller, Holtzman, & Messick, 1982), found some evidence of overrepresentation of both culturally diverse and English learners in special education. Although they differed in their conclusions and inferences drawn from the data, both expert syntheses concluded that overrepresentation was not consistent across all disability categories or ethnic minorities. Furthermore, patterns of disproportionality appeared mediated by educational context (school, district, state) in some unspecified fashion.Early litigation (Larry P v. Riles 1972/1974/ 1979/1984) cited test bias against English learners as a possible cause of disproportionality in special education. Later research focused on professional practices, including teacher bias (Riccio, Ochoa, Garza, & Nero, 2003) and all aspects of the assessment process (Loe & Miranda, 2002) as possible causes. Others contend that disproportionality may be a function of poverty, not just English learner status (Hosp & Reschly, 2003; MacMillan & Reschly, 1998; Oswald, Coutinho, Best, & Nguyen, 2001 ). …