Atualização do teste de desempenho escolar (TDE) : subtestes de leitura e escrita

• Ano de defesa: 2012
• Autor(a) principal: Knijnik, Luiza Feijó
• Orientador(a): Stein, Lilian Milnitsky
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Pontifícia Universidade Católica do Rio Grande do Sul
• Programa de Pós-Graduação: Programa de Pós-Graduação em Psicologia
• Departamento: Faculdade de Psicologia
• País: BR
• Palavras-chave em Português: PSICOLOGIA COGNITIVA; DESEMPENHO ESCOLAR - TESTES; PSICOLOGIA EDUCACIONAL - TESTES; PSICOLOGIA INFANTIL; ENSINO FUNDAMENTAL - ASPECTOS PSICOLÓGICOS
• Área do conhecimento CNPq: CNPQ::CIENCIAS HUMANAS::PSICOLOGIA
• Link de acesso: http://tede2.pucrs.br/tede2/handle/tede/816

Two papers constitute this dissertation: one theoretical and the other empiric. Study I brings a national and international bibliographic survey on the School Achievement Test (TDE) (Stein, 1994). TDE was published 18 years ago and there was interest in finding out if/how this instrument was being used by the scientific community, inasmuch as possible critics to the test. Data from 222 publications were analyzed, mostly scientific papers (129). The high quality of these papers was attested by the Qualis Capes classification, being 41% of the journals classified as A1 Qualis. A growing number of publications using TDE through the years were reported, reflecting both the growth of Brazilian scientific publication and the increase in interest for the test. This study revealed the importance of this instrument in Brazilian educational evaluation scenario. The purpose of study II was to explore the quality of TDE subtests reading and writing, through an Item Response Theory analysis. This study used the databases of five researchers from four Brazilian states. Data from 1.850 participants were analyzed, between 1st to 6th grades, from both public and private schools. Results indicated that the writing subtest is measuring precisely medium levels of ability and less satisfactory high and low levels.These data indicate that students with low and high ability within the continuum of the latent trait may not have been well differentiated and that this subtest could be improved by adding more difficult items as well as easier items, so that more information in these areas could be provided. Results of the reading subtest indicate that it is discriminating adequately low and medium levels of skill. Thus, more difficult items could be added to this subtest so that children with higher ability could be better discriminated. Conclusions are that an items review and adding of new items is highly desirable, as much as new standards for TDE