Teaching Innovation Awards Winner

University of Exeter

Dr. Sally Rogers, Dr. Andrew Griffiths, Dr. Bonnie Fraser, Dr. Bryony Williams, Ali Hudson
Headshot of Dr Sally Rogers
Headshot of Dr Andrew Griffiths
Headshot of Dr Bonnie Fraser
Headshot of Dr Bryony Williams
Headshot of Ali Hudson


56% of students enrolled on our first year Genetics module have not completed Mathematics A level (or equivalent), and “Maths Anxiety” is a recognised phenomenon in Biology undergraduates. Consequently, many students are reluctant to embrace statistics in data analysis and lack the confidence to choose appropriate tests. We have also found that within modular degrees, statistics are covered within core skills modules and students frequently “silo” their knowledge, which becomes a further obstacle to applying what they have learned in one module to another. Group work is common in undergraduate modules, meaning students that are not confident with maths can avoid engaging with this area. We wanted to address these issues within the context of our module. Our aim was to design a set of genetics problems that students work through individually, practicing step-by-step analyses and application of statistical tests. Importantly, we wanted to provide real-time feedback, ensuring students learn from the experience as well as providing a summative assessment point. 

We worked with LearnSci to incorporate their new unique dataset generation tool into a bespoke digital worksheet that focuses on data and statistical analyses. To our knowledge, the Smart Worksheets have not been used in this way before for complex statistical questions, and therefore this worksheet represents a significant innovation. This module represented a further challenge, in that students on different programmes are taught to use different statistical software, potentially meaning differences in calculation/approach. We had to carefully “stress test” the questions to make sure they were robust to such differences. The final product carefully combined multiple-choice questions and reporting of test statistics performed on unique datasets by students from different departments and with varying levels of maths competencies. 

350 students completed the worksheet, with an average score of 85% . This is a significant increase from the average score achieved on similar questions last year (74%). There were no significant differences in results between 1) students with or without Mathematics A level, 2) students on different programmes and 3) Home and International students, demonstrating we were successful in developing an assessment which benefited all students equally. Importantly, the majority of students reported positive outcomes with regard to the effectiveness of the worksheet in improving statistical skills and confidence (see figure 1). In addition, marking time for staff was significantly reduced. 

Questions asked to the students, and average scores. 1. 'Provided me the opportunity to apply maths skills and conduct calculations' received average 4. 2. 'Improved my confidence in using statistical tests' received average score 3.5. 3. 'Improved my confidence in statistics software' received average score 3.9. 4. 'Was effective in improving my statistical skills' received average score 3.5.v
Figure 1. Response from 97 students surveyed regarding their experience of the Genetics LearnSci digital worksheet assessment. Students were asked whether they disagreed or agreed with the statements shown, on a scale of 1 to 5, using Mentimeter for anonymity.

Academics teaching on the related core skill modules have reported higher engagement with the statistical resources from their students, demonstrating how the Genetics worksheet has promoted the application of knowledge across modules. Providing students an opportunity to engage with, and apply, maths and statistics early in their degree is crucial for developing confidence in these key employability skills. 

More generally, the progression to utilising unique datasets and Smart Worksheets for statistics powerfully illustrates the flexibility of these approaches beyond their current applications. Potential problems arising from variation in answers from students across different departments were successfully resolved, demonstrating that this novel type of Smart Worksheet design can be deployed at scale across a broad range of programmes.