Methods: Education data were taken from public school records for all children enrolled in Minnesota's public schools during the 2005-2006 school year (over 858,000 children). Child welfare data were obtained on children in two age groups: young children in grades kindergarten through second grade, and adolescents in grades 7, 8, and 9. Data on child welfare contacts spanned one year prior to the 2005-2006 school year, and one year after. Education data included primary language spoken in the home (which was used to form immigrant and non-immigrant student groups) and a variety of descriptive data including special education participation, free and reduced-price meal use, disruptions to school enrollment, attendance, and residential mobility. Child welfare data included determined maltreatment events and out-of-home placement.
Results: The study results showed that non-English speaking students were disproportionately less likely to participate in special education overall compared to English speaking students. Among all students who declined to participate in, special education Somali and Spanish-speaking students had higher rates of parental refusal in comparison to other non-English speaking students. Non-English speaking students were much more likely to be residentially mobile than English-speaking students, and students who were mobile had disproportionately less participation in special education compared to students who were non-mobile. Among child welfare age groups, rates of determined child maltreatment of English-speaking and non-English speaking students were very similar. Not considering immigrant status, females were disproportionately more likely to experience maltreatment than were males, and most students involved in child welfare were poor.
Implications: This study provides the first known population-level examination of immigrant students. Further investigation of what situations create a referral for special education assessment are needed to clarify whether gender in general, or if ethnicity and gender together, are correlated with how teachers or others recognize indications of learning challenges. Further research could also uncover whether these results are true elsewhere and for other time frames, and what some of the possible reasons are for these disproportionate levels of involvement. Understanding whether immigrant families truly do not require these supportive services or whether there are under-identified and under-served needs are equally important to policymakers, educators, and practitioners.