This research aims to elaborate a tool for diagnosis of dyslexia taking into account the most representative theoretical hypotheses concerning etiology of dyslexia. Research on causes of dyslexia has led to multiple theories (Expertise INSERM, 2007) : phonological theory (Ramus et al., 2003), visuo-attentional hypothesis (Valdois et al., 2004), cerebellar theory (Nicolson et Fawcett, 1999), theory of auditory deficit (Tallal, 1980), theory of magnocellular deficit (Stein, 2001). This way, we have included in the same experimental protocol a set of tasks measuring various cognitive capacities (memory, attention, phonological, morphological, motor, visuo-attentional, auditory). This protocol was used with 91 typical children and 35 dyslexic children attending a specialized consultation for dyslexia. From this experimentation, we have extracted a predictive model, based on a limited number of tasks in order to discriminate in 30 minutes as well as possible between dyslexic and non-dyslexic. Moreover, we made cognitive profiles for each child, that corresponds to the performance in every task. This cognitive profile will be helpful to give indications for a deeper assessment aimed to provide an individualized and adapted remedial program and rehabilitation treatments. Following of our research, a software called Alex is actually in a validation period. It will constitute a tool for the health professionals in charge of children who experienced learning disabilities
