Abstract
This study aimed at the development of the Reading Subtest of the second edition of the School Performance Test (TDE-II). This study consisted in items selection, analysis by experts, and validation in which a sample of 661 students from the first to the ninth grades of Elementary School participated. The Subtest was divided in two versions: version A for students from the first to the fourth grades and version B for students from the fifth to ninth grades. Exploratory factor analysis indicated unidimensionality with one factor explaining 57% of the shared items variance for version A and 38% for version B. Item Response Theory analysis indicated appropriate levels of discrimination and difficulty for measuring the construct. Results suggest sources of validity in relation to test content, and internal structure. In terms of reliability, both versions covered the range of ability with high precision, thus suggesting that the Subtest presents good psychometric properties for reading evaluation.
Keywords:
cognitive neuropsychology; reading; psychometric; evaluation
