EVIDENCE ON THE INTERNAL STRUCTURE OF ENEM AND THE COMPETENCIES OF NATURAL SCIENCES

Abstract

In this work, the internal structure of ENEM is investigated, with special attention to the Natural Sciences test. Microdata from 2014 and 2015 were analyzed. Factor analysis suggests that one factor is sufficient to describe the variance of the 175 items. The two-factor model suggests systematic heterogeneity, in the two years, between the sub-factors, especially in the Mathematics and Languages and Codes tests. On the other hand, the average of the 4 tests proved to be collinear with the general factor. The Natural Sciences test showed considerable internal multidimensionality, which could be a characteristic of scientific literacy. The Natural Sciences competencies described in the Reference Matrix were consistent with the empirical behavior of the items, suggesting that the internal structure of the test reflects its theoretical foundations.