A third study suggests a link between school tracking and adolescents' levels of academic self-concept (Ireson, Hallam, & Plewis, 2001). A cross-sectional study of 13- to 14-year old students in 45 secondary comprehensive schools in Britain found that students in schools that used tracking in many, but not all of their classes had higher levels of academic self-concept than students in schools that had mostly mixed ability classes and schools in which the majority of classes were tracked. The authors also found that the relationship between tracking and academic self-concept was greater for students' self-concepts in English than those in math or science. Although this finding comes from a cross-sectional analysis, and therefore is limited in its ability to indicate a causal relationship, the authors controlled for a few important variables that might predict whether students enroll in schools using greater or lesser tracking, such as the student's economic status, and also controls for their prior levels of achievement on a number of standardized tests.