Luque, J. L. 1 , Ortiz, A. 1 , Cobo, M. A. 2 , Giménez, A. 1 , López-Pérez, P. J. 3 , López-Zamora, M. 1 & Sánchez, A. 2
1 Malaga University (Spain)
2 Consejería de Educación (Andalusia, Spain)
3 European University of Atlantic (Spain)
We investigated the predictive power of early linguistic and cognitive skills, and genetic risk on Spanish literacy acquisition and learning difficulties within a prevention-oriented RTI model.
Participants were two different cohorts of around 900 Spanish children ages, from 4 to 7 and 5 to 8. Participants were assessed three times per year during four years, in individual sessions lasting 15-30 minutes. To assess genetic risk we surveyed parents. At the age of 4 and 5, the assessment sessions were designed following a multiple risk perspective. Measures included tasks in phonological awareness, rapid naming, verbal memory, vocabulary, morphology, syntax, oral comprehension, visual-attentional span, executive function, motor coordination and some early mathematical skills. At the age of 6, additional measures were included: word and pseudoword reading, reading speed text, reading comprehension and orthographic decision.
We applied a Support Vector Machine Regression (SVMR). At the third year, predictions showed better balance between sensitivity and specificity than the reference studies. Sensitivity index reached 80%, 90% and 95%, at 5, 6 y 7 ages. Under certain school conditions, indexes shown higher percentages indeed. Since we used an internet platform to deliver and record the data, the procedure is ready to be transferred to school.