E-assessment of computer programming
Rhys Gwynllyw, James Edward Smith · UWE Research Repository (UWE Bristol) · 2018
This paper demonstrates how we have used Dewis, an algorithmic open source e-assessment system, to automatically assess programming skills, in particular, in the C programming language. Teaching and assessing programming skills is challenging; prior to the implementation of this automatic assessment system, computing assessments were marked manually and this proved unpopular with students and academics due to the delay in marking and providing feedback. This new approach enables students to submit their computer code online through a link on their Virtual Learning Environment. From a student’s perspective the marking process takes a matter of seconds before the student is provided with a mark and feedback. A number of pre submission and post-submission checks are performed on the student’s supplied code. These are essential to ensure that the student’s code satisfies certain operational and security requirements before running on the system. Typically, the e-assessment system executes the student’s supplied code a number of times with different input sets and the student’s code is evaluated based on their code’s resulting output. Prior to execution, the student’s code may be augmented by Dewis specific code to facilitate deeper analysis of the student’s code. The analysis of the code’s output enables the system to respond to the student’s submission with ‘intelligent feedback’. This feedback explains to the student, where appropriate, reasons for their submission not scoring full marks. There are also in-built detectors for ‘common student errors’ which, when triggered, further enhances the intelligent feedback. In their first few weeks at UWE Bristol, students are presented with a number of mini-tasks involving assessing C competencies with Dewis. These tasks are purely formative, support is given to students in computer lab sessions and students may submit as many times as they wish. Later on in their first year, students are given more significant programming projects (e.g. the n-Queens problem, path finding problem), the assessing of which is summative. Results show that this innovative work is making a positive impact on students.