Roland Dante

Predicting End-of-course Test Using Teacher-made Test: Using Data to Drive Instructions Roland Dante

Predicting how students will perform on standardized tests is relevant to all educators. This work examines the relationship of students? performance on teacher-edited benchmark assessments administered to their performance on the state?s end-of-course test. Teachers have information on student achievement within specific topics based on tests, quizzes, homework, and other assignments. However, the best way to use this information in predicting high-stakes testing is unknown. Providing teachers with guidance on how to use this information would be extremely valuable. This research determines how well the data from periodically-administered tests may predict scores on the state?s end-of-course test. The school implemented an intervention process that involves frequent assessment of the skills being taught. Students were taught and re-taught, tested and re-tested until they mastered the curriculum. It is my hope that the research reported here will help decision-makers on education with evidence-based guidance on how to use student data to drive instructions.