Abstract
The aim of this study was to use path modelling to establish how rapid automatized naming (RAN), verbal short-term memory (VSTM), letter-sound connection (LSC), phoneme blending (PHB), and Raven tasks predict reading in Finnish and German. Students (N = 769) from Finland, Germany, and Italy (German-speaking children from South Tyrol) were followed from first grade until the end of second grade. Firstly, in all countries, LSC was found to be the strongest predictor for reading in first grade. Secondly, Finnish students’ word-reading skills were better than those of German and Italian students throughout the follow-up period, but word-reading level in first grade predicted word-reading level after one year only for Italian and German students. Thirdly, rapid automatized naming (RAN) and verbal short-term memory (VSTM) predicted reading skills in each orthography and country with a different power and at different phases, implying that the educational system also has a role in predicting reading skills.