Adaptação de teste de leitura The Reading Decision Test para o português do Brasil em crianças do 1º ao 5º ano do Ensino Fundamental

• Ano de defesa: 2011
• Autor(a) principal: Pinheiro, Elayne Cristina Morais [UNIFESP]
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de São Paulo (UNIFESP)
• Palavras-chave em Português: Reading; Screening test; Adaptation; Dyslexia; Translating; Child; Leitura; Teste de triagem; Adaptação; Dislexia; Tradução; Criança
• Link de acesso: http://repositorio.unifesp.br/handle/11600/8874

Screening tests in reading are essential for teachers and professionals working on the field of education, as these instruments provide a fast and economic measure of reading disabilities. So, this study aimed at adapting a screening test for reading, The Reading Decision Test (RDT), to Brazilian Portuguese. This test is made up of three subtests (letter, words and sentences) with increasing difficulty levels. Each subtest had two equivalent forms (F1 and F2). This research was carried out in two stages: (I) translation of the instructions and adaptation of the subtests to Brazilian Portuguese; and (II) assessment of the adaptation of the Brazilian version of the RDT. Stage I has two phases: 1) translation of the instructions of the RDT and adaptation of the word and sentence subtests; 2) analyses of the stimuli of the subtests by specialist judges. Stage II was made up of three studies, with schoolchildren from the 1st to 5th year: Study 1) RDT tested on 40 children with adequate school performance, to assess the adaptation of the items and the necessity of changes; Study 2) test application in 80 children, half with and half without learning problems, with the same aim as study 1, as well as to compare the performance of these children on the RDT in relation to other tests already adapted to the Brazilian population; Study 3) administration of the RDT in 230 randomly selected children (n=46 from each year), in order to assess the initial psychometric properties of the subtests. From studies 1 and 2, it was possible to substitute the items, by means of difficulty index analyses (DI) and item-total correlation (I-T). In study 2, a significant difference between the groups with and without learning problems was observed on all three subtests (p<0.05). The letter and words subtests showed moderate correlation (p<0,05) with the other tests already adapted to the Brazilian population, but the sentences subtest showed a weak correlation (p<0.05). In study 3, the analysis of the items in each subtest showed a significant effect for schoolyear. There was a significant statistical difference (p<0.01) on the recognition of both types of items in the words (real words and pseudowords) and sentences subtest (true and absurd sentences) in F1 and F2. The reading speed analyses showed a significant difference (p<0.05) in the number of right answers, in each school year, in both forms of all three subtests. The intra-class correlation (ICC) showed a moderate correlation between F1 and F2 on the letter (ICC=0.68) subtest, and a strong correlation on the words (ICC=0.90) and sentences (ICC=0.90) subtests (p<0.05). In conclusion, all three subtests of the Brazilian version of the Reading Decision Test show adequate initial psychometric properties, and this may be considered as the final version of the test.