Learning Analytics & Admin Tools
Dashboards for engagement metrics, predictive models for at-risk students, and AI-assisted scheduling.
Analytics Dashboard Demo
A simulated learning analytics dashboard showing the kinds of engagement, performance, and behavioural data that AI-powered school platforms can track and surface to educators.
At-Risk Indicator Explorer
Explore the indicators that predictive models use to identify at-risk students. Toggle indicators on and off to see how composite risk scores change, and examine the ethical considerations for each data point.
Ethics Warning: Handle Predictive Models with Care
Predictive at-risk models are powerful tools, but they carry significant risks. They can reinforce existing biases, produce false positives that stigmatise students, and create surveillance dynamics that erode trust. These systems should always augment — never replace — professional educator judgment. No student should be labelled or treated differently based solely on an algorithmic prediction.
Risk Score Calculator
Toggle indicators on/off and adjust their weights to see how a composite risk score changes. This demonstrates how different weighting schemes can dramatically alter which students are flagged.
Key Insight
Learning analytics are fundamentally double-edged. A well-designed dashboard can flag a student who is quietly disengaging weeks before a teacher might notice — enabling early intervention that changes a trajectory. But the same system can create a surveillance dynamic that chills student risk-taking, reinforce biases embedded in historical data, and produce false positives that attach an "at-risk" label to students who are simply different, not struggling. The indicators explored above show how easily weighting decisions can shift who gets flagged. A model trained on attendance data from a school where marginalised students face transport barriers will inevitably over-flag those students. The human teacher remains essential — not as a passive consumer of dashboards, but as the critical interpreter who contextualises data, questions its assumptions, and advocates for each student as an individual rather than a data point.