Evidências Desfavoráveis ao Postulado de Cargas Fatoriais Simples do Exame Nacional do Ensino Médio (ENEM)

The National High School Examination (ENEM) has a restricted measurement model characterized by the assumption that the marker items in each domain are exclusively linked to their target domain. Previous studies suggest, through indirect evidence, that this model may not be valid. However, this postulate has not yet been directly assessed. In this study, this assumption was investigated through factor analysis of the items of the ENEM 2011 edition. Two models were tested. The first, called Thurstone's simple structure, represents the measurement model of the ENEM. The second, of crossed loadings, refute this model. The crossed loadings model was the only one that presented a good fit to the data according to all the indices employed. The evidence found is unfavorable for the assumption of simple factor loadings of the measurement model of the ENEM, indicating issues of validity and in the quality of the scores produced.