The present study examines the cross-language transfer of lower-level reading processes (reading fluency and comprehension, morphological awareness) in L1 German to L2 French across different learner groups in primary school. To investigate measurement validity and then low-achieving readers, different models are tested drawing on a structural equation modelling approach.
Language transfer studies have shown that L1 language skills have an influence on L2 Language learning (Cummins, 2007; Sparks et al., 2009) and that cross-linguistic transfer facilitates reading in a foreign language (Kim & Piper, 2019; Koda, 2005). Therefore, the present study examines the cross-language transfer of lower-level reading processes in L1 German to L2 French in grade 3 across different learner groups. Lower-level reading processes as defined by Khalifa & Weir (2009) are already well established among normal readers in 3thgrade in a shallow orthography like German (Seymour, Aro & Erskine, 2003). Therefore, cross-linguistic transfer might be facilitated for lower-level reading processes in normal readers, but it remains unclear if and to what extend readers with dyslexia differ in their L2 lower-level reading processes in French. 366 predominantly German-speaking Swiss primary school children were tested after one year of learning French at school. The test battery contained of word reading fluency, letter knowledge, morphological awareness, reading comprehension and verbal short-term memory. The data were analysed using structural equation modelling in Mplus (Muthén & Muthén, 1998-2017). The analysis included two stages. First, the construct validity of the French test was examined by conducting a cross-validation multi-factorial approach. In a second step, the final model was used to investigate cross-linguistic influences. The final model yields a bifactor model with a general underlying factor for reading and 3 group factors. The first group factor is associated with letter knowledge and the second and third factor are related to morphological awareness (measured by a morpho-orthographic and sentence segmentation task). The model indicates that after one year of instruction, normal readers are successfully applying L2 lower-level reading processes. Already in this early stage, dyslexic pupils are falling behind in FL lower-level reading processes. Overall, the findings give evidence of transfer in lower-level reading processes.