Elevating the student voice and weaving qualitative feedback into transformative action.
Transforming our response to student feedback
Universities receive thousands of student comments annually containing rich feedback detailing their learning experience. Responding to this feedback at a community level has proven challenging for universities, due to the volume and variety of comments. Further, there has been a challenge pinpointing a causal link between university learning programs and particular feedback from students. Historically, a lack of nuanced insight has constrained innovation in univeristy teaching and student experience practice, leaving teachers and teaching leaders without the clarity need to drive meaningful change – and contributing to declining student sentiment, particularly in larger universities.
The Student Experience AI Dashboard was created to transform the way universities listen to their students. Its primary aim is to elevate the student voice by turning thousands of open‑ended comments from the QILT Student Experience Survey into clear, actionable insights. By leveraging advanced coding and ethically‑managed AI, the tool empowers educators and leaders to understand what matters most to students and to act on it swiftly.
Our dashboard exemplifies responsible AI innovation in higher education analytics and demonstrates how universities can meaningfully engage with student feedback to drive continuous improvement. It moves beyond traditional score-based dashboards by harnessing natural language processing and machine learning to analyse large-scale open-ended survey comments. Demographic filters enable navigation across student cohorts, supporting the delivery of program-level insights and actionable, policy-aligned recommendations, all while preserving data privacy.
The vision behind the Student Experience AI Dashboard
Watch the recording below of Deputy Vice‑Chancellor (Academic), Professor Kris Ryan, as he discusses the vision and impact of the Student Experience AI Dashboard project.
Amplified student voice
Brings the richness of students’ own words into the spotlight – moving beyond numerical scores to surface authentic perspectives that matter to teachers and teaching leaders. This enables educators to respond to student needs with greater empathy, relevance, and impact.
Inclusive, timely insight generation
Reveals what’s working and where change is needed through clear, data-informed insights. Teaching leaders can use this evidence to guide program improvements, strengthen pedagogical approaches, and target support where it will most enhance learning.
Evidence-driven teaching decisions
Rapidly transforms thousands of student comments into structured, actionable themes – providing equitable representation of all voices, including those less likely to be heard. Teachers and teaching leaders can access this feedback in real time to adapt strategies and address issues when they matter most.
Transferrable, ethical innovation
Built on accessible, widely available technologies and guided by ethical AI principles, this approach can be readily adapted across disciplines and institutions. Teaching leaders can champion its adoption to advance student experience analytics and foster evidence-informed practice on a broader scale.
AI-assisted thematic coding
Applies semi-supervised, few-shot classification with human expert oversight to generate consistent and comparable themes across disciplines.
Interactive plotly dashboard
Allows users to filter data by faculty, program, year level, topic, or student background, with clickable heatmaps enabling in-depth cohort exploration.
Targeted summarisation tool
Generates clear bullet-point summaries for selected programs, cohorts, or themes, amplifying student insights.
RAG-powered recommendation engine
Matches summarised insights with UQ’s Policy and Procedure Library to produce policy-aligned actions – highlighting areas of excellence and those requiring improvement.
The Student Experience AI Dashboard follows a 4-step, human‑centred, data‑driven approach:
- Data collection: Thousands of open-ended comments from the national QILT Student Experience Survey are collected as the foundational data source.
- AI-assisted thematic coding: A locally trained AI model transforms traditional unsupervised topic modelling into a semi-supervised, few-shot classification pipeline. A human expert oversees the process to ensure the accuracy and relevance of the topics. Run in parallel, this approach delivers consistent, comparable themes and topics across faculties and programs – surfacing what works, identifying concerns, and highlighting improvement opportunities in minutes rather than months.
- Interactive exploration: A custom open-source Plotly dashboard enables users to filter data by faculty, program, year level, topic, or student background, making it easy to explore feedback for specific cohorts and support timely, targeted decision-making.
- Targeted summarisation: Users can select a specific program, student cohort, or set of topics to generate clear, bullet-point summaries that capture key insights and amplify the student voice.
- Recommended actions: The Program Pulse dashboard uses Retrieval-Augmented Generation (RAG) with UQ’s Policy and Procedure Library to instantly highlight programs that stand out either for excellence or areas needing improvement and generate actionable, policy-aligned recommendations that drive institutional change.
Overview
Our dashboard provides an interactive view of Student Experience Survey feedback, using AI to classify comments into 5 themes: Skills Development, Peer Engagement, Teaching Quality, Student Support and Services, and Learning Resources. Users can filter by faculty, program, stage, level, or citizenship, with real-time updates on selection statistics and comment counts for Best Aspects and Need Improvements. Interactive heatmaps display thematic percentages by group, enabling drill-down exploration for deeper insights. This AI-driven, filterable format allows rapid identification of strengths and improvement areas across student cohorts.
Subtopic analysis
The Subtopic Analysis tab enables deeper exploration within the 5 main themes. Users select a theme to view ranked bar charts showing subtopics by frequency, highlighting key areas needing improvement alongside positive aspects. Comparative heatmaps allow analysis across faculties, programs, or other groups, with all filters from the Overview retained for consistency. This view pinpoints specific strengths and concerns within each theme, supporting targeted, data-driven action.
Feedback table
The Feedback Table provides 2 interactive tables for detailed student comments – one for Best Aspects and one for Need Improvements. Users can filter by topics, subtopics, demographics, and more, or use a slider to select top subtopics and apply them instantly. Search boxes in each column refine results further, and tables can be exported to Excel for offline analysis. This setup enables targeted exploration of feedback while maintaining access to the original comment text.
Summaries and recommendations
The AI Summaries and Recommendations tab generates 4 concise, filter-driven briefing papers: Best Aspects, Needs Improvement, Summary of Findings, and policy-aligned Recommendations. Users adjust filters and Top-N sliders before clicking AI Summary & Recommendations, with results ready in under 2 minutes. Subtopic counts help refine inputs before regenerating. Outputs remain available for the session and can be exported as a polished Word document for executive review. Recommendations are aligned with UQ Policies and Procedures, ensuring insights are actionable and governance-ready.
Program pulse
The Program Pulse dashboard offers a real-time view of how programs perform relative to peers across 5 themes, using z-scores to highlight excellence (⊕) and underperformance (⊖). Users can filter by faculty, study area, level, and more, then adjust controls for minimum responses, z-score thresholds, and sorting. Heatmaps show Best Aspects (teal) and Need Improvements (red) side by side, enabling quick comparison of strengths and weaknesses. Switching between tabs provides both perspectives, helping identify standout programs and those needing attention, supporting evidence-based, targeted action for program quality improvement.
Contact
If you are interested in adopting or collaborating on this initiative, please reach out below.