Predicting dyslexia in Arabic-speaking children: Developing instruments and estimating their psychometric indices.

Abstract

Dyslexia is a reading disability that is characterized by when an individual has trouble in rapid and accurate word decoding. This study developed, piloted and assessed the validity and reliability of three instruments: Rapid Automatized Scale (RANS), Arabic Reading Ability Scale (ARAS) and Phonological Awareness Scale (PAS), on a sample of 700 students (aged 8-9 years). Four groups (n = 30) were formed based on the participants' results of the three instruments; Double Deficit (DD), Rapid Automatized Naming Deficit (RAND), Phonological Awareness Deficit (PAD) and No Deficit groups. Content validities of the instruments were supported using published reports; though educational experts further revised RANS. It found a significant inverse correlation between the PA test score and RAN (mistakes and time) score (r = -.44; p < .001), and a significant positive correlation between RAN mistakes and RAN time (r = .47; p < .001). Acceptable internal reliability of the RANS was demonstrated by a Cronbach's alpha test coefficient of α = .85 (>.70; acceptable). High inter-rater reliability tests were observed for the three instruments (r ≥ .86, p < .001). The three instruments can predict reading difficulties and dyslexia in Arabic-speaking populations.