Work

Automated grading assistance for student programs

Year: 1994
Type: article
Abstract: OverviewStudents are ill-served when grading fails to detect errors in their solutions to programming assignments; they will assume that their development process was adequate when in fact it was not.Yet assignments even in introductory courses can be too complex to evaluate thoroughly by conventional, manual means.Some instructors have used various sofiware tools to assist in the task of evaluating student programming assignments.The ability to collect students' work in a consistent form and test it automatically against a wide range of data can enhance students' skills in program design, coding, analysis, and testing, and can free the instructor from much tedium, allowing more time for more rewarding activities.But the use of such tools is far from widespread, despite the impossibility of thorough, valid, reliable assessment without them.No repository of such systems exists, nor do such systems themselves exist for all popular instructional computing platforms.Automating the testing process is trivial in theory ("for each student, for each test case, produce the result"), but difficult in application.Practically usable systems must be reasonably secure from student tampering, scalable to large classes, flexible enough to accommodate different forms of assignments, and robust in the face of bugs in the students' programs and in the lab environment.
Language: English
Open Access status: bronze
OpenAlex ID: W4245511663
FWCI:
Cites:
Cited by: 10
Related to: 10