Evaluating evaluation: a study carried out on 6th grade Natural Sciences’ written tests

Abstract

Introduction: Several authors say that tests are the main and, sometimes, the sole mean of evaluating students. If these instruments are frequently used during the evaluation process, than it is essential that the teachers formulate them with technical quality, as on its application often depends the students’ obtained grade and, consequently, their future. It was the importance of tests as a (still) privileged tool in the evaluation process and the chance to know its quality that took us to develop this thesis. Objectives: 1) To determine if the teacher-formulated tests are conceived to evaluate the specific skills required for scientific literacy; 2) To determine whether those tests evaluate the achievement of the educational goals proposed by Bloom’s taxonomy; 3) To analyze the summative evaluation sheets for the 6th grade of Natural Sciences, in what concerns their technical requirements. Methods: Recurring to a process of convenience sampling, we used a sample of 24 summative evaluation sheets concerning “Human Reproduction and growth” unit and we performed a content analysis. Results: The data brought on by this study states that the largest percentage (90%) of the tests’ questions is part of the substantive knowledge skills’ category. Items that evaluate other kinds of skills are almost non-existent in the tests; 90% of all of the tests’ questions concern the domains of “knowledge” (50%) and “understanding” (40%) – that asserts that the evaluation practices are often reduced to mere memory exercises. Conclusion: So the results obtained by this study testify the need to improve the process of initial and continuous training of teachers under evaluation.