UpGrade: Learning Analytics for Transparent Thesis Assessment

Project Description

Assessing master’s theses is a critical task in higher education. Even with detailed assessment rubrics, lecturers often have little insight into how their grading compares to that of colleagues or across programs. This lack of transparency can lead to uncertainty and inconsistencies in assessment.

This project aims to address this challenge by developing an interactive visual analytics dashboard for assessing master’s theses at Utrecht University. The dashboard will be built based on an anonymised dataset of ICS master’s theses from 2018 onwards, including variables such as study program, assessors, final grades, rubric scores, and open comments. All data is fully anonymised and handled in accordance with Utrecht University’s ethical and GDPR data protection guidelines.

Using these data, the dashboard will visualise grade distributions, rubric scores, trends over time, and differences between programs and assessors. Topic modelling on the open comments will reveal trends in the qualitative feedback. A combination of the visualisations and topic modelling will help lecturers reflect on their grading practices and support transparency, fairness, and consistency in thesis assessment.

The dashboard will be evaluated with lecturers to examine its usability, usefulness, and contribution to confidence in grading.

Contact Information

Privacy statements

Laptop with a visual dashboard, a stack of theses, and a rubric sheet
Image generated by ChatGPT