Variables affecting early foreign language learning

Even though early bilingual foreign language learning has been promoted in Europe for more than a decade now, programs for young learners are still scarce, and more empirical investigation into L2A in such contexts is needed.
This paper presents evidence from a longitudinal European project (ELIAS: Early Language and Intercultural Acquisition Studies) in 10 bilingual preschools (L2 English). First results (Kersten et al. 2010) showed that the quality of L2 input, as well as the duration and the intensity of L2 contact, present strong predictors for early grammar and lexical comprehension.
For the current study, the data of 389 children and 21 preschool teachers were reanalyzed using multilevel modeling and structural equation modeling to shed light on the interplay of additional learner-internal and learner-external factors affecting early L2A. The focus of the study was to investigate to what extent receptive L2 vocabulary and grammar are affected by learner-internal variables such as age, gender, and language background, as well as external factors such as the socioeconomic background, early literacy activities in the family, L2 contact duration, L2 contact intensity, and the quality of the L2 input, which was operationalized using the IQOS rating scheme (Input Quality Observation Scheme, Weitz et al. 2010). The IQOS quantifies verbal and non-verbal L2 scaffolding techniques. Data were elicited using the BPVS II (Dunn & Dunn 1997) and the ELIAS Grammar Test (Kersten et al. 2010), a parent questionnaire, as well as participant observation and videography for the program variables.
The time nested data structure was analyzed statistically using linear multilevel modeling and structural equation modeling. Preliminary results show a strong increase of children’s receptive L2 vocabulary and grammar test results. These changes are predicted by the duration and the intensity of contact time to the L2 and the amount of scaffolded L2 input. Additionally, the quality of the L2 input predicts the rate of L2 grammar comprehension, and early literacy activities in the family predict receptive L2 vocabulary changes. In contrast, no statistical effects were found for the direct effects of children’s socioeconomic as well as language background, and gender. Further moderator and mediator analyses concerning the interplay of the different variables will be carried out. We will discuss why certain variables might affect receptive L2 grammar and vocabulary to a different extent.