The study develops the 'Ostrakinda' toolkit for teaching morphologies of ancient Mediterranean cities. The methodology consists of a triple structure involving design, play, and analysis phases. In the design phase, the level of interest and actual demands of students regarding game-based learning (GBL) and ancient cities were explored through pre-experiments. The play phase involved experimental studies conducted in courses and workshops, where both physical and digital toolkits were tested by a total of 331 students. The analysis phase aimed to evaluate the impact of the game by systematically addressing five research questions.
The results of the paired t-test confirm a significant difference in scores before and after playing the game, thus validating the effectiveness of the proposed strategy. Additionally, Chi-square (X2) tests revealed significant relationships between the form of play and the level of abstraction in the game outputs. Individual and group experiences during gameplay were associated with various interactions, and group experiences particularly fostered competitive and collaborative learning. The extensive analysis using the GBL Design Scale, comprising ten critical factors (GBL-Fn), showed that the challenge factor (F8) had the weakest values, while fantasy (F5) and narrative (F6) had the highest values in the correlation matrix.
Overall, the study contributes methodologically and contextually to the existing literature, offering valuable findings for the future development of similar strategies in architectural education. However, it is important to acknowledge the limitations of this interdisciplinary study, particularly with regards to technical, cognitive, and archaeological aspects. Nevertheless, the clear and effective structure of the game holds promise as a valuable tool for learning.