Overidentification of learning disorders among language-minority students Implications for the standardization of school achievement tests

This German prevalence study examined disproportionate representation of language-minority students among children identified with learning disorder (LD) according to ICD-10 (WHO, 1992). Most German school achievement tests used in LD diagnostics do not provide separate norms for language-minority students, and thus do not take these children's second language status into account when evaluating their academic performance. Although this is likely to result in an LD overidentification of language-minority students, little is known about the magnitude of this effect. Therefore, we compared the estimation of LD prevalence between native German speaking students (n = 566) and language-minority students (n = 478) when pooled versus group-specific achievement norms were used for LD classification. Three important findings emerged from our study: Firstly, and as expected, significant disproportionality effects occurred under pooled norms. Specifically, the likelihood of being diagnosed with LD amounted to 1418 % among native German speakers and nearly doubled to 25-30 % among language-minority students. Secondly, disproportionality varied as a function of LD subtype: Whereas no disproportionate representation was revealed for arithmetic LD (F81.2), overidentification of language-minority students was found for verbal LD subtypes (namely, reading disorder [F81.0], spelling disorder [F81.1], and mixed disorder of scholastic skills [F81.3]). Thirdly, disproportionality effects were absent when group-specific norms were used for LD classification that controlled for second-language issues. Challenges that have to be met when testing language-minority students for LD are discussed.