Success Story
University of Birmingham Exam Modelling
The University of Birmingham was founded in 1900 by Royal Charter, as a new model for higher education. This was England’s first civic or ‘redbrick’ university, where students from all religions and backgrounds were accepted on an equal basis. The university supports over 34,000 students, with 100,000 online learners and 4,000 international students.
The University was embarking on a £606m investment in the Estate. This included the Aston Web Student Hub, the Library, the Sports Centre, and new residences along with teaching spaces and the Collaborative Training Laboratory (CTL). In parallel, the university wanted to increase student numbers by 5% per year over the following five years. This essentially led to the need to identify the supply and demand on current facilities throughout the estate works, along with the impact on the management of exams and scenario planning of the timetable and scheduling for the new CTL facility to be shared by several faculties and across many courses.
This case study explains the work undertaken in an examination modelling exercise to reflect the 5% increase in student numbers each year with the current year standing at 20,991 students doing 1,506 examinations, leading to a total of 93,375 exam enrolments to be managed. While at the same time losing several large-capacity examination venues. This exercise would provide information on the impact of this and inform contingency decisions regarding possible use of alternative venues and extended exam periods.
The exercise was based on five overall scenarios/questions for the current and future years, accompanied with a general outline of the considerations used by the examination timetabling team in the use of resources and maintenance of the quality of the experience for students.
The exercise was conducted using the EventMAP examination scheduling software application which provided the mechanism for modelling the resources, infrastructure and requirements for the exercise as well as running through multiple scenarios using the auto-scheduling capabilities within the software to effectively produce “what if” scenario examination timetables and report on these for comparison and analysis of the 20,991 students and a total of 93,375 exam enrolments, growing by 5% per year.
The data was also cross referenced against the list of “existing” rooms, plus candidate venues involved in the scenarios at each stage of the physical estate was being altered and venues availability. It also included the possible uses of some larger venues as split facilities to provide optimised space in possible exam venues. Based on the finding that in a five-week period there were not enough large rooms to accommodate the large exams without splitting the venues.
The model also included a “pairing” matrix of the available rooms, to link those which (if required) could be used for the same exam, based on geographical proximity.
Some of the scenarios also explored the extension of exam sessions over additional periods/days, with one question looking at use of evening periods, but excluded the use of Saturdays to minimise the restrictions on the students or buildings.
All constraints, rules and restrictions as outlined in the brief were built into the model, including the practical constraints that students should not have more than five hours of exams per day or more than 13 hours per week, and exams should be spaced apart from each other as much as possible. Avoiding exams on Saturdays for certain students and for the Chemical Engineering building in high demand for practical tests. In addition, within a room with multiple exams running, all exams being of the same duration and minimising the splitting of exams, over a maximum of two rooms.
The result was that decisions could be made on both hire of additional exam rooms or the optimised use of the facilities available at any point in time in estate construction works on the campus. This de-risked the deployment of the resulting exam timetabling exercise and allowed clear communications to be issued to both staff and students on the efficacy of the timetable in place. It also allowed the supply and demand to be modelled for future years to accommodate the 5% increase in students within existing facilities.
