Enhancing student project selection and allocation in higher education programmes

Abstract

Towards the end of higher education programmes of study, particularly in engineering, a substantial written projector dissertation is often included. The increased use of such projects is based on sound notions of increasing student independence, the development and demonstration of deeper learning, and an opportunity to showcase how to synthesise and use a range of knowledge and skills. However, there are a number of constraints, which we identify, that limit the potential for these positive benefits to be realised. In particular, a variety of methods are used to allocate supervisors, each of which has associated difficulties. In this paper we report upon a shift in approach, designed to impact positively upon the identified constraints by shifting the emphasis towards a broader topic selection. This is achieved by acknowledging the interaction of various elements in the selection of projects, such as potential supervisor, previous student experience, and subject matter. To make the allocation process as transparent as possible, we use an interactive web application to collect student preferences. The application gives immediate feedback to students on the current popularity of keywords, allowing students to make informed strategic choices. Finally, instead of determining the allocation manually, we use a simulated annealing algorithm to seek a global optimum. Results suggest that this is an appropriate and effective approach, leading to less dissatisfaction amongst students and staff and having a positive effect on student performance.