Abstract
Language has been conceptualized as both a measure as well as a predictor of integration among immigrants and their children. However, the relationship between language spoken at home and different educational outcomes remains poorly understood. Many studies indicate that nurturing students' first languages is positively associated with their learning at school. Other research suggests that one of the reasons why children of immigrants tend to perform worse at school is due to speaking a language other than that of instruction at home. In order to shed further light on the role of language choices at home for education, we examine both the correlates of language use at home as well as the relationship between this and reading scores and educational expectations. We differentiate between three language use groups: those who mainly use the language of origin at home, those who only use the language of instruction at home, and those who use both of these. We analyze these relationships using data from the Programme for International Student Assessment (PISA). In order to examine country differences, we place a special focus on two immigrant-origin groups that are present in significant numbers in a number of European countries: children with origins in Turkey and former Yugoslavia. These two groups have also been identified as being at major educational disadvantage across Europe. Our results suggest that continuing to (mainly) use the language of origin at home is more prevalent among children from socioeconomically more disadvantaged families, but is supported by more socioeconomically advantaged and more diverse school environments. In the majority of countries studied, switching to the language of instruction is associated with higher reading scores but not with higher educational expectations than continuing to speak mainly the language of origin at home. These relationships are to a large extent confounded (or in some cases potentially mediated) by family factors such as socioeconomic status and school-related factors such as school's socioeconomic composition. We conclude by highlighting the role that linguistically responsive pedagogies and a positive school climate can play for the education of all young people but in particular newly-arrived immigrants.