Conference grading of computer programs

Abstract

Two methods of grading computer science programs in a CSI course were evaluated in terms of (1) benefit to the student, (2) advantages for the instructor and (3) plagiarism. The two methods compared were conference grading and grading using written comments. Results indicate that students thought the grading was more fair when graded by one-on-one conference than when graded with written comments. The time required by the instructor was nearly the same. More was communicated orally in this time than in writing. There was less plagiarism in programs graded by conference, and the students whose programs had been graded by conference performed better on a programming test than those who had their programs graded by written comments. There was no significant difference between the two groups on the written examinations.