Allocating students to projects is a commonplace task in computing education. These decisions underpin student-supervisor allocation, the formation of tutee and capstone groups, and pair programming. These allocations play a critical role for individual learner outcomes and the success of collaborative interventions. For example, imbalance in either gender, ethnicity, or nationality can negatively impact learner outcomes. Despite the critical importance of these allocation choices, we see little consensus on how these are implemented. The allocation task can be challenging and time-consuming for instructors of even moderately-sized classes, and the fairness implications can be difficult to assess. Inadvertently, an instructor may allocate in a way that amplifies existing biases or disproportionately harms those from disadvantaged or protected groups. From students' perspectives, a lack of transparency on the allocation process may also lead to issues of trust. The Working Group will undertake a study of allocation practices by bringing together educational and ML literature to develop and evaluate the fairness of allocation methods, and develop educator guidelines to promote pedagogically grounded allocation practices.

This paper presents a novel 'Socketed' approach to assessment design in computing education, aiming to reconcile the need for individual student assessment with the benefits of collaborative learning. By providing a skeleton project to be combined with individualised components, students experienced a tailored yet unified assessment structure. Initial feedback shows improved comprehension and peer collaboration, and suggests mitigation of pressures towards academic dishonesty. Motivating context criteria and design considerations are provided to assist educators in implementing this in their own teaching.

Parsons problems are well-regarded activities for scaffolding beginner programmers. We present a technique for employing Parsons problems in a cohort of professionals aiming to upskill in programming and data analytics. Recognising the distinct learning needs of this demographic and the limitations of traditional coding exercises in assisting their learning, we developed a series of Parsons problems focused on both foundational programming constructs, and specialised topics within the domain of data analytics.

This paper describes experiences in teaching Python programming as part of a large-enrolment first-year subject, that is a foundation for an undergraduate major in data science and it is also taken by a wide variety of non-majors across the arts, sciences and business fields. The paper focuses on some central design decisions about the content, sequence, approach, and tool support, and we reflect on how they have worked out and what we changed, as we have taught the subject to about 2500 students over four offerings (including two which were entirely on-line due to the pandemic). Particular aspects that are the focus in this paper include: teaching programming through a sequence of common patterns/idioms for data exploration and analysis, rather than in the language-feature focus that is traditional in programming classes; read-then-write-until-correct presentation for each pattern; explicit presentation of a notional machine for execution, in stages as more language complexity is experienced; starting with core Python before then covering use of libraries such as pandas, matplotlib and scikit-learn.

This Research Full Paper analyses a dissonance in the application of accepted educational theory to development and research on laboratories in education. Most existing research has tended to conflate multiple constituent factors, resulting in an inability to appropriately determine which specific factors are the cause of observed variations in learning outcomes, and an inability to reliably interpret the external validity of the research outcomes.We begin with an outline of the supporting educational theories that underpin our understanding of laboratory education. We then identify notable examples in the literature where multiple variables, such as the user interface and access modes, have been conflated into laboratory `supertypes' and are not investigated independently. These supertypes reflect the most common variable combinations: `proximal laboratory, hands-on interface', `remote laboratory, rich contextual interface', and `simulated laboratory, decontextualized interface', referred to as hands-on, remote, and simulated/virtual respectively.We then investigate the impacts of this conflation - in the context of examples, this conflation limits the generalisability of previous results and obscures potentially rewarding avenues of future research. Finally, we present methods by which research may be conducted to investigate technically simpler directions of interface decoupling, the results of which may motivate the investigation of more challenging pairings.