Abstract:

This paper presents the findings of a study of Game-Based Learning (GBL) practitioners (N=12). The practitioners, who were a mix of academics, researchers and industry practitioners with experience of GBL research and / or implementation, were asked open questions in a survey about the effectiveness of GBL, the benefits that virtual reality (VR) bring to GBL, how the use of learning analytics dashboards (LADs) can help learner and educator, the mapping of learner interactions to taxonomies of learning, and whether universal design should be embedded at the heart of GBL. The questions also addressed the process by which GBL solutions are developed, with a focus on iterative prototyping and how GBL analytics can assist in the formative evaluation of a GBL solution when combined (triangulated) with other approaches (such as observations of learners playing a GBL solution). To assist the respondents, a case study was presented in the form of a game play video (of a VR-based GBL solution that teaches graph theory) with commentary by the GBL developer along with screenshots of learning analytics visualizations from a study of learners (N=20) carried out by the author.

A thematic analysis was carried out on the responses. On a question by question basis, meaning units were identified and categorized, and the categories were organised into themes. This led to several findings per question. The findings include: GBL is effective, but must be well designed; VR is an exciting, largely unexplored medium with many positive pedagogical benefits, but it has disadvantages such as cost and inaccessibility; LADs assist learners to be more autonomous and help with their executive functions, but they need to be carefully designed with actionable information and need to use visual aids to help comprehension; mapping learner interactions to a taxonomy (for the purpose of visualization) can support learners’ executive functions, such as assessing performance and progress, and setting goals – the gathering of learner interaction data has potential for automating the prediction of learner performance, which could lead to earlier intervention; an iterative approach to GBL development allows for greater inclusion of stakeholders, continuous improvement and more complex solutions – this agile approach in GBL development is under way and following the lead of other industries; using visualizations as part of formative evaluation can identify issues of usability, flow and balance, as well as measuring effectiveness; the practitioners agreed that universal design should be embedded at the heart of any model for GBL design and development, but were largely unaware of specific frameworks for universal design in education, such as the Universal Design for Learning (UDL) framework; when evaluating GBL solutions, a mixed methods approach was seen as best practice with triangulation then becoming possible to identify more complex issues.

The findings of the practitioner study will be of interest to all designers and developers of GBL solutions. They provide guidance on many aspects of modern GBL, such as the benefits of VR, the use of agile development processes, the integration of learning analytics, and mixed methods evaluation that allows for triangulation of findings from multiple data sets.

Keywords:

Game-based learning, learning analytics, virtual reality, agile development, mixed methods research, universal design for learning.