Simulations for Performance Assessments that Report on Knowledge and Skills (SPARKS)

Concord Consortium, Concord MA

Investigators

Abstract

This project is a three year project that is a collaboration between the Concord Consortium and CORD and that extends earlier ATE funded work on assessment of competencies with electronic circuits and test equipment. The new assessments provide finer grain analysis about student understanding and misconceptions, thus enabling specific and well targeted individual interventions. Each assessment challenges the user to accomplish some task, such as making a measurement or troubleshooting a circuit. The computer monitors the student and generates reports for use by the student, the instructor, or both. The project aims to improve learning by providing timely and informative feedback on students? progress, as inferred from their performance on realistic tasks. The PI has found that students? scores on question-and-answer tests are not reliable predictors of their ability to perform a cognitively demanding task, as shown by their performance on simulation-based assessments. In particular, the PI has shown that there must be a serious reconceptualization of the assessment tools used to evaluate students? learning if they are to constitute reliable measures of the targeted knowledge and skills. By extension, the project investigates the central question of how to evaluate whether a student has learned something.