An Exploratory Non-experimental Design of Self-Assessment Practice

Work thumb

Views: 60

Abstract

The present study asks whether repeated self-assessment can shape performance, self-assessment accuracy, and confidence in students whose past entails an instructor-centered education. Eighty-eight classes were assigned to either a control condition or a self-assessment practice condition. In the latter, students estimated their final class grades at the start and at the end of the semester, as well as grades before and after the midterm and the final test. Subjective confidence ratings accompanied estimates. At the end of the semester, the practice condition exhibited higher class and final test grades. The accuracy of students’ predictions also improved as knowledge of test and class demands increased over time. Poor performers made the most optimistic estimates but with the lowest confidence. The accuracy of their estimates also showed the largest gains from knowledge of test and class demands. These results suggest that not only repeated self-assessment may benefit performance, but also those who benefit the most from knowledge of class and test demands may be those whose performance is less than satisfactory.