LAUSR

Abstract This study investigates how the proportion of SEN students in regular classes is related to the student-level and class-level cross-curricular competences. The data (N = 5368) come from a large-scale, longitudinal assessment study conducted on students at the beginning and end of lower secondary education in a Finnish metropolitan area. The results of the multilevel regression models showed that students in regular classes with SEN students performed on average lower than students in classes without SEN students, and that the proportion of students with SEN in class weakly predicted negatively the ninth-grade test scores. Furthermore, SEN students seemed to perform at the same level regardless of the proportion of other SEN students in class. However, students without SEN in classes with SEN students performed slightly lower than their peers in classes without SEN students in the ninth-grade assessment even when the initial differences related to placement were taken into account.