Validation of a digital decoding assessment instrument for Brazilian Portuguese: psychometric properties and insights for similar languages

The early and accurate assessment of decoding is critical for preventing persistent reading difficulties, particularly in large-scale educational systems facing chronic disparities. This study aimed to validate a digital decoding assessment task named Protocolo de Avaliação da Decodificação (PRADE) - Protocol for Decoding Assessment specifically designed for Brazilian Portuguese, a language characterized by high orthographic transparency. The primary objective was to establish the psychometric equivalence of this software-based tool compared to a paper-and-pencil gold-standard test (TDE), while the secondary objective examined its sensitivity to the developmental progression of reading automaticity. A stratified sample of 250 elementary school children (grades 1–5) from both public and private schools was assessed. Results indicated excellent internal consistency (α > .90) and strong convergent validity, with moderate-to-strong correlations (r = .40 to .75) between PRADE and the gold-standard test across all grades. The instrument also demonstrated robust criterion-related validity, effectively differentiating proficiency tiers and teacher-rated academic performance groups. Furthermore, the findings reveal that linguistic variables, such as word length, significantly modulate decoding efficiency (accuracy and speed), even in a transparent orthography. These results reflect the qualitative transition from effortful phonological assembly to autonomous orthographic recognition, aligning with established developmental frameworks. This study validates PRADE as a robust and scalable tool for large-scale monitoring and contributes to a more typologically diverse science of reading by incorporating data from an under-researched, transparent orthographic system.