Bilingual education using a Content and Language Integrated Learning (CLIL) approach is widespread in secondary education throughout Europe and also found further afield. In many contexts CLIL seems to select or attract the more able and more academically-inclined pupils, or only be available to pupils in higher academic secondary streams. Positive effects of CLIL for target language proficiency development may be due in part to this cognitive or academic selection effect. Can the target language skills of pupils with lower scholastic attainment – a group which, in some contexts, has less access to CLIL programs - also benefit from the CLIL approach?
The current two-year longitudinal quasi-experimental research, part of a larger study, focused on the development of oral proficiency skills of three cohorts of 603 pre-vocational pupils in 25 classes in the Netherlands in both CLIL and non-CLIL programs. Pre-vocational secondary education in the Netherlands serves approximately fifty percent of the total pupil population, including a large percentage with a minority-language background, and consists of the least academic streams. Despite the lack of explicit school-based selection procedures for prevocational pupils’ participation in CLIL, there were significant differences in favor of the CLIL groups in the initial levels of English oral proficiency, fluency, and Willingness to Communicate. Furthermore, the CLIL pupils showed significantly more growth than the non-CLIL control group in Speaking proficiency, but not for Speaking fluency or Willingness to Communicate. This positive result for the CLIL group did not appear to be moderated by pupil background variables. Despite the small effect sizes found, these results are encouraging for the Manuscript further development of CLIL provision for pre-vocational pupils in the Netherlands and elsewhere, and indicate that despite the cognitive challenges, the CLIL approach can have a positive effect on foreign language proficiency of pupils in less academic educational streams.
Doi to the final published version: doi.org/10.1075/aila.22020.den