Work in progress-improving feedback using an automatic assessment tool

Abstract

Students of computer science freshman year usually develop assembler programs to learn processor architecture. Homework exercises are done on paper, while those in lab sessions are solved with the aid of programming tools. Students perceive theory and lab as different subjects, so they donpsilat use lab tools to test their theory solved problems. Moreover, during lab sessions, students often tend to ask for the teacherpsilas guide and advice instead of using the debugging tools because these are new and unfriendly for them, and do not offer a quick and clear feedback. In this paper we present an automatic and friendly assessment tool, SISA-EMU, with a novel feature: exercise driven feedback with teacherpsilas expertise. It provides correctness information and clues to help the students solve their most common mistakes for each individual problem (and not typical generic debug information) without the physical support of a teacher. SISA-EMU is currently in pre-deploy phase via a Moodle learning platform and we will have first evaluation results by the end of the current term.