Visual analytics could help universities to better manage budgets, and student dropout rates

The #Feesmustfall movement is driving an increased focus on cost-effective tertiary institution management, says Murray de Villiers, manager of the academic program for Middle-East, Africa and Turkey at SAS Institute.

Johannesburg, 07 Jun 2016
Read time 3min 50sec
Murray de Villiers
Murray de Villiers

Forecasting budget allocations and managing available resources have become increasingly important. This is particularly true in light of the economic climate and the #Feesmustfall movement, which are driving an increased focus on cost-effective tertiary institution management, says Murray de Villiers, manager of the academic program for Middle-East, Africa and Turkey at SAS Institute.

"Unless universities are able to forecast with greater accuracy and manage resources optimally, protest action could undermine the quality of education at our institutions of higher learning. It has become crucial for universities to budget accurately and allocate resources such as staff and equipment where the maximum value can be derived," he says.

Traditionally, universities must look to historical data and request specific report modelling in a bid to understand factors impacting university management and trends shaping future needs. The ability to uncover underlying trends and forecast more accurately is expected to support better decision-making by university management, De Villiers believes.

Visual analytics solutions, he says, should be designed to support university management by integrating all available administration and student data and enabling highly visual data interrogation and forecasting. This could help to reduce student dropout rates and - crucially - help universities to manage available resources and funds allocation more effectively.

A Proof of Concept of South Africa's first visual analytics system for tertiary institution management is currently being conducted at a major local degree granting institution, De Villiers says. It is being implemented by SAS in collaboration with a SAS partner and the University Management Information Systems team.

The visual analytics solution is designed to support forecasting and management by integrating all available data, including data from student smart cards, which are used across campuses for facilities access and payments. De Villiers explains that these smart cards deliver a wealth of information about students' social and academic behaviour, which can be analysed to determine general patterns around poor student performance and dropout rates.

"The primary goal of universities is to help students graduate in the recommended timeframe, and to meet the government's required throughput guidelines, so it is very important that they understand which factors lead to poor academic performance and dropping out altogether," he says. In an environment where less than half of students graduate within the prescribed period, education authorities are anxious to increase pass rates.

"We believe the factors that influence a student's success extend beyond class attendance. They include lack of adequate nutrition, social activity, a lack of adequate preparation and in some cases substance abuse. Also, the high schools students attended impact on their chances of success at university. Our initial view already indicates that students in residences who do not eat regularly tend to get lower marks. There are likely many other corroborating factors impacting on performance too.

"By gathering all available data and analysing the trends, university management would be in a better position to gain a single view of the student and intervene to support those showing early signs of failure. It can also help determine which facilities are well used and which are not, including the optimal deployment of resources."

Although the Proof of Concept is limited to university administration data and information sourced from student cards, De Villiers notes that more valuable insights could be derived if systems such as this could be extended to include background information on students from sources such as schools and social media.

"The Proof of Concept we are putting in place will allow rectors, deans and registrars to sit in the boardroom and use their mobile devices to visualise reports, drill down into the reports, and interrogate the data with 'what if' questions on the fly," says De Villiers.

The Proof of Concept is expected to start delivering useful results within the next three to six months, says De Villiers, after which there is potential for it to be rolled out across South Africa's other tertiary institutions and even to serve as a foundation for an integrated education analytics system in future.

For more, download the SAS Solution Brief [link to] on gaining insight into students to improve outcomes across the student lifecycle.

Editorial contacts
ITWeb Tracy Burrows (011) 807 3294
Login with