Triggering evaluation improvements with artificial intelligence in a university’s andragogy didactics and curriculum

Abstract The motivation for this mixed-methods research comes from the experiences of a PhD amid participant observation on courses of a master’s Program in Higher Education from a university in Central America. Natural Language Processing is used on students’ communication exchanges from WhatsApp groups, allowing to gain insights on the overall sentiment during the 11 courses of the program from 2024. Robotic Process Automation is employed to monitor the progress without much manual effort using Python, whose code is exposed as part of the article. The focus is then placed on the micro-curriculum due to its relatively agile characteristics. Further findings trigger a wider exercise, as a second research phase, where students from all the university faculties are invited to voluntarily and anonymously answer a questionnaire based on the exploration from the first phase, with the objective of discovering key opportunities to improve the quality of the curriculum evaluation process, with emphasis on the student perspective. As part of the results, the university strengths are identified, while also recognizing key opportunities: better introduction of the mechanisms offered for suggesting improvements; promotion of open feedback from students to the professors and administrative personnel, including an increased accessibility of the latter; and necessary adjustments to the standard course evaluation form in Moodle, to increase its usefulness. Overall, highlighting the importance of student participation in the curriculum development. The study applies an exploratory to descriptive sequential design, providing the means for simplified reusability and comparability in similar projects with other higher education institutions.