On Monday 22 March 2021, the RSS International Development Section held a meeting titled 'Engaging university students in statistics through innovative approaches – with examples from Africa'.
The speakers were Roger Stern of the University of Reading & Stats4SD, UK; Danny Parsons from the University of Reading and IDEMS International, UK; and James Kaleli Musyoka of the University of Maseno, Kenya.
The speakers presented a variety of innovations aimed at engaging university students in statistics. These were initially used in a 'Statistics Problem Solving' course for MSc Mathematical Science students at AIMS Cameroon (African Institute for Mathematical Sciences). A broad collection of statistical problems was introduced including aspects of design, collection, manipulation and organisation of data through to analysis and reporting. Students had hands-on experience with statistical games leading to simulated data and real data from diverse application areas. Randomised, automated assessment was used to support student mastery of concepts. The course proved to be an eye opener, both for students with a BSc in statistics and those with no statistics ackground. Some of these innovations are being scaled to more students on undergraduate courses, as well as supporting e-learning courses in statistics.
Three main points stood out in this talk. The first was the format of the course which was atypical of statistical training, particularly as it is done at African universities. The students were not given formulas and 'nice' data to analyse using standard techniques. On the contrary, they were exposed to messy data, the need to estimate, the fact that there could be no single wholly correct design for an experiment; they were expected to engage with the whole gamut of statistics from design of data collection to reporting and feedback; they had real world problems to try to solve.
The second was the use of STACK (System for Teaching and Assessment using a Computer algebra Kernel). This free open source package was developed at the University of Birmingham (UK) and is now hosted by the University of Edinburgh. It is essentially a web-based assessment system developed for mathematics, but here it has been adapted for use in statistics teaching. It allows for algebraic inputs, intelligent feedback, and randomisation of questions and can be made interactive.
The third point was the feedback from these innovations having been put into practice at AIMS Cameroon and the University of Maseno. It has to be said that although feedback was mainly positive, some students (particularly those with a good theoretical degree) found that the lack of certainty took some getting used to. At Maseno, the benefits heavily outweighed the challenges: James Musyoka explained that with 1000 students on his Descriptive Statistics course, and with no assistants, providing the expected individual weekly feedback was just not possible. Using STACK, students could get real-time feedback and develop their understanding.
Download the slide presentations (PDF)
Phil Crook, IDS Meetings Secretary