Formal instruction vs informal exposure. What matters more for teenagers’ acquisition of English as a second language?

ABSTRACT

This paper provides a comparative analysis of the drivers of English Language Competence (ELC) in a representative sample of European adolescents. School factors – such as English instruction time and English onset – are found to play an important role, especially in countries whose official languages are more distant from English. However, schools do not fully compensate for disparities in ELC. Gender, parental education and parental socioeconomic status are strongly associated with students’ ELC. Particularly, girls and children from more privileged social backgrounds show systematically higher ELC, even net of school factors and reading ability in their own country language. The role of family background is stronger in countries with languages that are more distant from English, suggesting that family resources are needed more when the skills are more difficult to acquire elsewhere. Finally, informal English exposure through media and cultural products is strongly and positively associated with ELC. This holds true in countries with both high and low linguistic distance from English, suggesting that schools should promote more informal English learning to increase overall ELC and reduce social disparities.

KEYWORDS:

• English language acquisition
• formal instruction
• informal learning
• Europe
• family background

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1. A growing global phenomenon is the practice of teaching academic content in a foreign language, such as English. This instructional approach, also known as Content and Language Integrated Learning (CLIL) (Sanjurjo, Blanco, and Fernández-Costales 2018) or, when specifically using English, English Medium Instruction (EMI) (Dearden 2014) is based on the use of a foreign (English) language to teach academic subjects in countries where the first language of the majority of the population is not English. Unfortunately, the information on the adoption of such practices by schools is not rich enough in the SURVEYLANG data to allow us to study their association with students’ ELC (Araujo and Dinis da Costa 2013).

2. In the Flemish Community of Belgium, and in Spain, additional schools beyond the number of schools required for the ESLC standard design were included in order to represent specific regions. Therefore, for these two cases, two different samples were created. We did not consider these extra schools since our analysis is concerned with cross-country differences and not regional variation within countries.

3. The model was run separately by country and included gender, age, migration background, parental education, ISEI, and school grade.

4. The basic idea underlying the approach used is to compare pairs of words that have the same meaning in two different languages according to their pronunciation. The average similarity across a specific set of words is then taken as a measure for the linguistic distance between the languages (see for more details and applications: Isphording and Otten 2013; Wichmann, Holman, and Brown 2018).

5. A possible channel through which social disparities can manifest is supplementary private tutoring (Bray 2017). Additional analyses (available upon request) show that, indeed, children of highly-educated parents and children of parents with high socio-economic status receive both enrichment and remedial private lessons outside school at significantly higher rates than students from lower social backgrounds. These social differences in private tutoring are particularly pronounced in countries where the spoken language is more distant from English.

6. Unfortunately, the country of birth information was removed from the standard file, providing only an aggregated variable which does not allow us to study heterogeneities between migrants.

7. Countries are classified in two groups based on whether their country language distance score is below or above the median English language distance (P50 is BG = 0.36).

Additional information

Notes on contributors

D. Azzolini

Davide Azzolini is a research fellow at the Research Institute for the Evaluation of Public Policies of the Bruno Kessler Foundation (Italy) and an affiliated scholar at the Urban Institute (Washington DC, USA). He is interested in public policy analysis and evaluation in the fields of education and immigrant integration.

S. Campregher

Sabrina Campregher has a Ph.D in “General Pedagogy, Social Pedagogy and General Education”. She is the representative of school language policies for the Education and Culture Department of the Autonomous Province of Trento (Italy) and adjunct Professor for the University of Trento where she teaches language awareness and cooperative learning. She is the author of books and articles about foreign language learning and acquisition.

J.E. Madia

Joan E. Madia is a PhD student in Sociology at Nuffield College, University of Oxford and an associate researcher at the Institute for the Evaluation of Public Policies (FBK-IRVAPP) in Italy. His main research interest lies at the intersection between educational inequality, the evaluation of public policies and quantitative methods.