Design and Evaluation of a Digital Educational Escape Game for AI Literacy

Abstract

Developing competencies in the field of artificial intelligence (AI) represents a central task of modern education while preparing young people for an increasingly digitalized world. In this context, practice-oriented, interactive, and motivating learning formats are gaining importance. Educational escape games (EEGs) have already proven to be a promising approach in other educational contexts, as they combine subject-related learning objectives with structured, suspense-based game scenarios and can therefore support both cognitive and affective learning. This paper presents the design and evaluation of the digital EEG “AI Chaos”, which was developed within the framework of a doctoral research project for students in grade 8 and above. The browser-based game was implemented using Genially. The didactic design includes three thematically structured game rooms featuring different game mechanics and varied puzzles, as well as a final meta-puzzle. The game aims to convey fundamental concepts and principles of AI learning mechanisms in a comprehensible and motivating way, while fostering media literacy and critical judgment regarding opportunities and challenges related to AI. A quasi-experimental pre-post questionnaire design with supplementary observation protocols was used for the evaluation, in which a total of n = 181 students from various types of schools participated. In addition to subject-related knowledge, motivational and emotional aspects such as the experience of flow as well as attitudes toward AI were measured. The results show a significant increase in participants’ knowledge after playing the game. In addition, the learners reported a strong flow experience, with the dimensions of social interaction and perceived competence showing particularly high scores. Attitudes toward AI remained largely stable over the short intervention period. The strongest predictor of cognitive knowledge gain was the participants’ prior knowledge. The activity was rated very positively by the students, who particularly highlighted teamwork, autonomous activity, and the enjoyment of solving puzzles. Overall, the results indicate that (digital) EEGs represent an effective and accessible method for teaching complex future-oriented topics such as AI. Their strengths lie in promoting learning motivation and emotional engagement within the learning process.