Web-CAT Automated Grader

The Web-CAT Grader was inspired by the Virginia Tech Curator. It is a web-based, automated submission and grading system for programming assignments.

The Web-CAT Grader supports traditional models of automated program grading, but also supports grading of assignments where students do their own testing. It helps encourage test-driven development (also called test-first coding), where students write small unit tests for each piece of code they add. Web-CAT allows a student to submit his or her test cases along with the solution, and grades on test validity and test completeness as well as code correctness.

The current version of the Web-CAT Grader has processed over 10,000 student submissions during the Fall 2003 semester, serving approximately 280 students spread across nine sections of two courses. This is the pilot semester for the project at high volumes but with a limited number of courses. Experimental results from an upper division course in Spring 2003 are briefly described in some of the publications below. A journal submission to ACM's Journal of Educational Resources in Computing that provides full details of the experimental evaluation carried out in the spring is currently under review.

The most important result obtained in the spring was a 45% reduction in the estimated bug density of student code (bugs per thousand non-commented source lines of code).

Papers

M.S. Thesis

 

copyright © 2004 Virginia Tech, ALL RIGHTS RESERVED
Last modified: January 28, 2004, 2:57:46 pm EST, by Stephen H. Edwards <edwards@cs.vt.edu>