Supporting educational choices: possibilities and ethical concerns of data- and algorithm-driven technology

Thesis event information

Date and time of the thesis defence

Place of the thesis defence

Linnanmaa, auditorium L2

Topic of the dissertation

Supporting educational choices: possibilities and ethical concerns of data- and algorithm-driven technology

Doctoral candidate

Master of Psychology Egle Gedrimiene

Faculty and unit

University of Oulu Graduate School, Faculty of Education and Psychology, Learning and Learning Processes

Subject of study

Educational Psychology

Opponent

Professor Teresa Cerratto-Pargman, Stockholm University

Custos

Professor Hanni Muukkonen, University of Oulu

Visit thesis event

Add event to calendar

Supporting educational choices: possibilities and ethical concerns of data- and algorithm-driven technology

Data- and algorithm-driven technologies such as artificial intelligence and learning analytics dashboards offer new possibilities to support individuals through their educational paths and crossroads. This dissertation focuses on users of such technologies, examining reported benefits, challenges, and emerging ethical concerns.

The results revealed high expectations and a wide range of decisions in educational settings targeted by data- and algorithm-based technologies. Five key areas of possible support were identified by users in the context of study path choice: provision of career information, research and analysis of the information, diversification of ideas on possible career paths, direction and decision support, and self-reflection. Diversification of ideas about alternative career paths emerged as a key need for the users. Regarding career decision facilitation, the provided support lacked alignment with the optimal decision-making process suggested by theory. Especially contextual information needed to make decisions was lacking. The results also reveal multiple ethical considerations, specifically limitations of the transparency ideal from the user perspective, tendency towards confirmation bias, and concerns for overreliance on AI guidance.

For practice, results highlight the vulnerability of potential users, which should be considered and counterbalanced with accountability by practitioners, developers, and policy makers. Findings also provide further direction for developing data- and algorithm-driven tools to support individuals in their educational crossroads.
Last updated: 11.2.2025