Experimenting matters Learning and assessing science skills in primary education

The research in this dissertation aims to investigate the acquisition of students’ science skills in grades 5 and 6 of primary education in the Netherlands. In most primary science classes, science skills are mainly taught by way of conducting investigations. However, prior research indicates that explicit instruction and separate skills training may be more effective.
In this dissertation, four studies are discussed. In the first study, an instructional framework was developed based on a categorization of science skills into thinking skills, science-specific skills and metacognitive skills. This instructional framework was used to develop lessons using systematic instruction aimed at the development of these different skills. The second study describes the development and psychometric quality of the measurement instruments in order to examine the acquisition and transfer of science skills. Two paper-and-pencil tests, three performance assessments and two questionnaires were used for this purpose. In a third study, the effects of two experimental conditions were evaluated, following an experimental pretest-posttest design: a condition with explicit instruction and a condition in which all aspects of explicit instruction were absent. Students in both conditions received an 8-week intervention and were compared to students in a baseline condition who followed their regular science curriculum. The fourth study addresses the use of performance assessments as a diagnostic tool for science teachers. In general, the results indicate that the measurement instruments can be used to reliably measure science skills. Findings also show that explicit instruction facilitates acquisition and transfer of science skills.