This project expands and fully implements Web-CAT, the Web-based Center for Automated Testing. Web-CAT is an innovative automated grading platform that focuses on how well students test their own code, rather than assessing the correctness of a program's output. Web-CAT encourages students to practice comprehension, analysis, and evaluation skills by forcing them to continually articulate their own understanding of how their code should behave. Students practice writing down hypotheses about how their code should behave, and then experimentally validate this understanding by executing their tests. Web-CAT has been formally evaluated at one institution and has been successfully adopted by three others. This project has four objectives:
To expand Web-CAT from a proof of concept to a state where it can be distributed and applied widely, both by end-users who wish to automate their assignment grading practices, and by educational researchers who wish to build on this platform to develop and evaluate their own automated assessment techniques.
To develop educational materials for instructors who want to use Web-CAT, including a comprehensive teaching manual capturing best practices, example assignments, how-to's on adapting existing assignments for automated assessment, and workshops for adopters.
To develop a community of educators who want to use, extend, and support Web-CAT and the teaching strategies it enables. The community is supported by a wiki and a mailing list, together with birds of-a-feather-style meetings at professional events such as the SIGCSE Symposium.
To evaluate the effectiveness of Web-CAT in a diverse set of contexts.
