22

This longitudinal study examined data from 2,767 children, monitored over two years, who were assessed in early literacy at four-years-old (HabLit4a) and then in reading at six-years-old (HabLec6a). HabLit4a was evaluated using three tasks: letter-name knowledge, phonological skills, and vocabulary. Meanwhile, HabLec6a was assessed based on four tasks: phoneme segmentation, word reading, listening comprehension, and reading comprehension. The impact of three sociodemographic variables—parental education level (PEL), family income (FI), and parental literacy practices (ParLitPr)—and two individual variables—age and gender—were also examined. Results from a structural equation analysis show a strong link (0.68) between HabLit4a and HabLec6a. All variables directly affect performance at four-years-old and PEL has the highest impact (0.25) for HabLit4a. The effect of age on skills favors older children and this is more pronounced in girls. The five variables also have an indirect effect on HabLec6a via HabLit4a: 0.17 for PEL and age. In addition to the expected results, and in line with what has been observed in other languages, the current study also distinguishes between two socioeconomic variables, namely NEP and RF: the results show the level of parental education has the greatest impact on the reading level reached at the end of Grade 1.