As participation in higher and further education increases and technology advances there is an increasing interest in the use of automated systems to support teaching and learning. One area of interest is the automated assessment of students’ work for formative and summative purposes.

The Database Systems module in computer science has an annual enrolment of approximately 250 to 300 students. With such a large class the amount of time available per student is limited. Historically this has meant that the module has been assessed entirely by written examination, with only formative laboratory exercises.

In the 2002/03 session an automated marking system, CourseMarker, was deployed to support the formative assessment, and in the 2003/04 session this was extended to include summative assessment based on an automatically assessed programming assignment and a supporting written report.

In order to evaluate student’s attitudes to the introduction of the coursework surveys were conducted over two years through questionnaires. The questionnaires included fixed questions about participation, difficulty, and the usefulness of the coursework, and open-ended questions to determine why some students choose not to do formative coursework and general attitudes towards the coursework component of the module. Comparative results of these surveys are difficult because of changes made between the two cohorts, but do not show any clear effect of the introduction of summative coursework on the participation in formative work.

In addition, the logs maintained by CourseMarker were analysed. This analysis shows that students are not exploiting the advantages of the automated assessment system. With the system that was implemented students could make multiple submissions to identify areas where their programs were not working according to the specification. Most students, however, made only one or two submissions, and did not make any submissions before the day that the assignment was due. This observation is supported by anecdotal evidence of student attendance in laboratory sessions, and suggests that the introduction of automated assessment should be accompanied by guidance as to how to make best use of the systems that are being used.
While the methods of automated assessment trialled only apply directly to computer programming assignments, student comments raise some more general issues about the use of automated assessment. In particular the automated assessment system used in this case exposed some of the mechanisms of assessment, which caused some students to compare it to their previous experiences of assessment and, in some cases, automated assessment. In addition several students commented that the use of an automated tool to mark their work was unfair since it is rather inflexible.