Ainsley Miller (University of Strathclyde)
Anxiety Around Learning R in First Year Undergraduate Students: Mathematics vs Biomedical Students.

R is an open-source statistical programming language that is becoming more widely used in industry. It is also becoming the standard for teaching statistics due to its flexibility, replacing software programmes such as Minitab and SPSS.

The main driver for reform within Scottish statistical undergraduate modules is the creation of the Scottish Qualification Authority’s (SQA) Higher Applications of Mathematics course which has statistics as one of its core topics. The course which launched in Scottish high schools from August 2021 will see R introduced into the Scottish high school curriculum for the first time. This therefore facilitates the need for R to be introduced into Higher Education courses at an earlier stage.

To that end, in academic year 2021/22 we introduced R Studio into our first year introductory statistics module. This class is taken by students studying Mathematics, and those studying Biomedical Sciences. Both cohorts were surveyed in order to assess their anxiety and enjoyment of learning how to use R, with a goal of assessing any differences between the two groups of students. We will also examine attainment within R for the perceived experts (mathematics students) and non-experts (biomedical students).

We found that there was no association in software anxiety at the start of the course, both cohorts were equally anxious/not anxious about learning R. However, as the course progressed the mathematics students reported lower levels of anxiety compared to the biomedical students. Furthermore, the mathematics students seemed to enjoy the module more than the biomedical students, thus needing further investigation into enjoyment vs anxiety.

Craig Alexander (University of Glasgow)
Teaching R for online distance learning – moving from the lab to the laptop

With the rise in popularity of online distance learning programmes and MOOCs within recent years within statistics, and the rise in popularity of online learning platforms such as DataCamp, there is a requirement to effectively teach programming courses online. The school of mathematics and statistics at the University of Glasgow launched an online distance learning masters in data analytics in 2017 which contains several courses focussed on programming, including R.

In this talk, I will discuss how this course was designed, in terms of the course content, delivery style and assessment structure and look at the differences when compared to an on-campus version of an R programming course. I will also discuss different changes made to the course over the years, those of which have worked well and those that weren’t quite as successful.

In particular, I will look at the inclusion of more interactive features of R such as RMarkdown and Shiny to the course content, teaching tools to aid live coding sessions and structuring assessments for online learning, and suggestions to replicate the interactivity of the in-person lab setting. Throughout the talk, I will contrast the changes with in-person teaching approaches, looking at their effectiveness from the perspective of both the teacher and learner. I will conclude with some future proposals to build on the current structure of the course.

Andy Field (University of Sussex)
Using learnr to create interactive tutorials for R (and RStudio): from the pinnacle to the pit

This talk describes the lessons I have learned through developing and teaching with interactive tutorials for R/RStudio written using the learnr package (Borges & Allaire, 2017).

Part 1: the pinnacle. The first half of the talk highlights the capabilities of learnr and its potential for creating an immersive interactive learning experience through the use of embedded code chunks (with solutions), video and quizzes. I will illustrate these features and discuss how I used learnr in conjunction with flipped classroom teaching to facilitate large group teaching. I will reflect upon the benefits of the approach for formative assessment, inclusivity, active and peer-based learning.

Part 2: the pit. It didn’t all go smoothly. The second half is a journey through various pits of despair into which I unwittingly fell. I will reflect upon the deployment of learnr tutorials, the potential for learnr tutorials to create a disconnect with R/RStudio, problem areas for teaching data science using learnr, whether using code solutions hinders learning, and whether the flipped classroom fosters helplessness for less confident students.

Throughout the talk I will highlight some areas of potential good (and not so good) practice when using learnr to teach statistics and R/RStudio. I will conclude with suggestions for teaching workflows that aim to maximise the chances of experiencing the pinnacle and not the pit.

Mark Andrews and Russell Turk (Nottingham Trent University)
Transitioning from SPSS to R teaching in a large psychology department

Statistics is a major part of the teaching curriculum in the field of psychology. For the past few decades, the most widely used software package for statistics teaching in psychology has been SPSS. Beginning in 2019, the department of psychology in Nottingham Trent University, which is the largest psychology department in the UK, transitioned all of its statistics teaching from SPSS to R. Each year, we now teach statistics using R to around 2,500 undergraduate students and this involves over 150 members of staff. In this talk, we will discuss the practical and other challenges we encountered in this transition, what worked well, what did not work well, our plans for the future. In particular, we will describe how we modified our curriculum to take advantage of statistics teaching opportunities provided by R, how we create bespoke teaching tools to facilitate teaching and learning R, and how both students and academic teaching staff generally perceived and engaged with R. We hope that our experience will be useful for other departments who are considering a similar transition to teaching with R.