Assessing the Feasibility and Impact of Using Online Problem-Solving in Computer Science

Ramapo College Of New Jersey, Mahwah NJ

Investigators

Abstract

Computer Science (31) This project develops four problem-generators on programming languages topics. The problem generators are designed to generate problems as parameterized instances of a problem template. Their performance and use is evaluated over four semesters. The problem generators are expected to promote active learning among students and to improve the problem-solving skills of students. Problem-based learning improves long-term retention. Researchers have advocated the use of self-paced exercises, practice to build problem-solving skills, and the use of frequent, graded assignments in a course. Faculty are increasingly turning to the use of technology to provide practice problems to students. They are developing programs to generate problems, and making them available to students for practice. But, such programs that provide students with frequent, self- paced exercises have been infrequently attempted for computer science topics, because problems in computer science are not always quantitative, and they often depend on the structure of arbitrary computer programs. This project investigates the potential for use of problem-generations in computer science topics in light of the unique constraints of the computer science curriculum.