A new study published on arXiv has examined the usability of a self-service learning analytics (SSLA) tool designed to let non-technical educators build custom data indicators without writing code — and found actionable paths for improvement.

The paper, titled 'Usability Evaluation and Improvement of a Tool for Self-Service Learning Analytics' (arXiv:2603.24321), centers on a tool called the Indicator Editor, described as a no-code, exploratory platform that guides users through a structured workflow to implement learning analytics indicators. The research team employed an iterative evaluation methodology that combined qualitative user studies, usability inspections of high-fidelity prototypes, and a workshop-based evaluation conducted in a real educational environment.

The workshop study involved 46 students and used three standardized measurement instruments: the System Usability Scale (SUS), the User Experience Questionnaire (UEQ), and the Net Promoter Score (NPS). These tools provided quantitative benchmarks against which the Indicator Editor's performance could be assessed and compared to broader usability norms.

Based on their findings, the researchers derived specific design implications targeting three key areas: workflow guidance, system feedback, and information presentation. The authors argue that while SSLA tools hold significant promise for democratizing data-driven decision-making in education — offering user control and transparency — their real-world adoption hinges critically on usability.

The study situates itself within a growing pattern in educational technology research: the recognition that technical capability alone does not ensure uptake. Tools built for non-technical stakeholders must meet those users where they are, or risk becoming unused infrastructure. The paper's iterative approach, moving from qualitative exploration to prototype inspection to authentic classroom testing, reflects methodological rigor increasingly expected in HCI-adjacent education research.

The full paper is available at https://arxiv.org/abs/2603.24321. Note: As an arXiv preprint, this work has not yet undergone formal peer review.