Improve Abstract Reasoning in Computer Introductory Courses

Due to the elevated programming courses' failing rate in our department (45%) an action research was initiated. As part of this action research, that was performed during four semesters several course structures and learning tactics were examined. The evaluation methodology was simple and based only on the percentage of failing students. The success achieved was attributed to two main factors (1) using a visualization environment (Micro-world) for the whole duration of the course, which helped in understanding the more complex and abstract issues, and (2) using individual assignments that enforced better learning habits and development of individual algorithmic thinking. The paper describes the various attempts, as well as the final structure, that reduced the failing students by over 77%.